1	1	machine translation	[machine translation]	26.266666666666666	2	28	26	15
2	2	natural language	[natural language]	25.0	2	26	9	9
3	3	statistical machine translation	[statistical machine translation]	12.67970000576925	3	10	6	3
4	3	machine translation system	[machine translation system]	12.67970000576925	3	10	10	5
5	4	paper	[paper]	10.8	1	109	2	2
6	5	language	[language]	10.764583333333334	1	109	65	48
7	6	experimental result	[experimental results]	10.0	2	11	1	1
8	7	speech recognition	[speech recognition]	8.0	2	9	7	7
9	8	word	[word]	7.87948717948718	1	80	47	39
10	9	parallel corpus	[parallel corpus]	7.0	2	8	7	7
11	10	method	[method]	6.882758620689656	1	70	34	29
12	11	system	[system]	6.788	1	69	28	25
13	12	training datum	[training data]	6.666666666666667	2	8	4	3
14	12	language processing	[language processing]	6.666666666666667	2	8	8	6
15	13	natural language interface	[natural language interface]	6.339850002884625	3	5	2	2
16	14	approach	[approach]	6.081818181818182	1	62	39	33
17	15	statistical machine	[statistical machine]	6.0	2	6	0	0
18	15	word sense	[word sense]	6.0	2	7	6	6
19	15	mt system	[MT system]	6.0	2	7	5	5
20	16	translation	[translation]	5.9709677419354845	1	61	40	31
21	17	user modeling	[user modeling]	5.8	2	7	6	5
22	18	model	[model]	5.7785714285714285	1	59	34	28
23	19	sentence	[sentence]	5.586666666666667	1	57	34	30
24	20	dialogue system	[dialogue system]	5.5	2	7	3	2
25	21	feature structure	[feature structures]	5.333333333333333	2	7	5	3
26	22	machine	[machine]	5.066666666666666	1	53	21	9
27	23	unknown word	[unknown word]	5.0	2	6	2	2
28	23	language model	[language model]	5.0	2	6	4	4
29	23	language understanding	[language understanding]	5.0	2	6	5	5
30	23	user model	[user model]	5.0	2	6	2	2
31	24	bilingual corpus	[bilingual corpus]	4.8	2	6	6	5
32	25	result	[result]	4.771428571428571	1	50	32	14
33	26	different inference type	[different inference types]	4.754887502163469	3	3	0	0
34	26	natural language system	[natural language system]	4.754887502163469	3	3	0	0
35	26	machine translation output	[machine translation output]	4.754887502163469	3	4	3	3
36	26	natural language processing	[natural language processing]	4.754887502163469	3	4	2	2
37	26	result of the experiment	[results of the experiments]	4.754887502163469	4	3	0	0
38	26	set of feature	[set of features]	4.754887502163469	3	4	3	3
39	26	statistical machine translation system	[statistical machine translation system]	4.754887502163469	4	4	4	4
40	26	machine translation mt	[machine translation MT]	4.754887502163469	3	5	4	2
41	27	SMT model	[SMT model]	4.75	2	6	5	4
42	28	significant improvement	[Significant improvement]	4.333333333333333	2	6	5	3
43	29	answer to question	[answers to questions]	4.226566668589751	3	4	4	3
44	30	information	[information]	4.196428571428572	1	43	29	28
45	31	corpus	[corpus]	4.147619047619048	1	43	32	21
46	32	text	[text]	4.085	1	42	23	20
47	33	email .g	[e .g]	4.0	2	5	4	4
48	33	evaluation measure	[evaluation measure]	4.0	2	5	4	4
49	33	language pair	[language pairs]	4.0	2	5	4	4
50	34	user	[user]	3.8000000000000003	1	39	14	14
51	35	sophisticated representation	[sophisticated representation]	3.75	2	5	5	4
52	36	sentence planner	[sentence planner]	3.666666666666667	2	5	4	3
53	37	homophone error	[homophone errors]	3.5	2	5	3	2
54	38	performance	[performance]	3.4823529411764707	1	36	20	17
55	39	language system	[language system]	3.333333333333333	2	5	5	3
56	40	feature	[feature]	3.2888888888888888	1	34	20	18
57	41	grammar	[grammar]	3.194736842105263	1	33	20	19
58	42	type of information	[types of information]	3.1699250014423126	3	2	0	0
59	42	linguistically sophisticated representation	[linguistically sophisticated representation]	3.1699250014423126	3	3	3	3
60	42	natural language input	[natural language input]	3.1699250014423126	3	2	0	0
61	42	machine translation mt system	[machine translation MT systems]	3.1699250014423126	4	2	0	0
62	42	statistical machine translation SMT	[Statistical machine translation SMT]	3.1699250014423126	4	2	0	0
63	42	100 ,000 word	[100 ,000 words]	3.1699250014423126	3	2	0	0
64	42	cue phrase refer expression	[cue phrases referring expressions]	3.1699250014423126	4	2	0	0
65	42	trainable sentence planner	[trainable sentence planner]	3.1699250014423126	3	2	0	0
66	42	new event detection	[new event detection]	3.1699250014423126	3	2	0	0
67	42	correct syntactic structure	[correct syntactic structure]	3.1699250014423126	3	2	0	0
68	42	pronoun resolution in dialogue	[pronoun resolution in dialogue]	3.1699250014423126	4	2	0	0
69	42	expert human translation	[expert human translation]	3.1699250014423126	3	2	0	0
70	42	word segmentation problem	[word segmentation problem]	3.1699250014423126	3	2	0	0
71	43	speech	[speech]	3.1	1	32	12	12
72	44	error correction	[error correction]	3.0	2	4	3	3
73	44	inference type	[inference types]	3.0	2	4	3	3
74	44	word vector	[word vectors]	3.0	2	4	4	4
75	44	attentional state	[attentional state]	3.0	2	4	1	1
76	44	previous work	[previous work]	3.0	2	3	0	0
77	44	translation model	[translation model]	3.0	2	4	4	4
78	44	user utterance	[user utterance]	3.0	2	3	0	0
79	44	polynomial time	[polynomial time]	3.0	2	3	0	0
80	44	parsing strategy	[parsing strategies]	3.0	2	4	2	2
81	44	training speaker	[training speaker]	3.0	2	4	4	4
82	44	target speaker	[target speaker]	3.0	2	4	1	1
83	44	information retrieval	[information retrieval]	3.0	2	4	4	4
84	44	two formalism	[two formalisms]	3.0	2	3	0	0
85	44	computational complexity	[computational complexity]	3.0	2	4	1	1
86	44	Machine Translation	[Machine Translation]	3.0	2	4	4	4
87	45	evaluation	[evaluation]	2.992857142857143	1	31	15	14
88	46	problem	[problem]	2.8913043478260874	1	30	25	23
89	47	experiment	[experiment]	2.791666666666667	1	29	13	12
90	48	understanding result	[understanding result]	2.666666666666667	2	4	4	3
91	48	CSR corpus	[CSR corpus]	2.666666666666667	2	4	4	3
92	48	grammar formalism	[grammar formalism]	2.666666666666667	2	4	4	3
93	48	small corpus	[small corpus]	2.666666666666667	2	4	4	3
94	49	algorithm	[algorithm]	2.5941176470588236	1	27	18	17
95	50	training	[training]	2.5923076923076924	1	27	14	13
96	51	discourse	[discourse]	2.591666666666667	1	27	13	12
97	52	use	[use]	2.555555555555556	1	28	22	9
98	53	correct candidate	[correct candidate]	2.5	2	4	3	2
99	53	important role	[important role]	2.5	2	4	3	2
100	54	mt	[MT]	2.375	1	25	10	8
101	55	error	[error]	2.1750000000000003	1	23	15	12
102	56	parsing	[parsing]	2.1	1	22	9	9
103	56	dialogue	[dialogue]	2.1	1	22	7	7
104	57	task	[task]	2.0909090909090913	1	22	12	11
105	58	subtopic boundary	[subtopic boundaries]	2.0	2	3	3	3
106	58	node-based inference	[Node-based inference]	2.0	2	2	0	0
107	58	human language	[human language]	2.0	2	2	0	0
108	58	large corpus	[large corpus]	2.0	2	3	3	3
109	58	novel method	[novel method]	2.0	2	3	3	3
110	58	linguistic structure	[linguistic structure]	2.0	2	2	0	0
111	58	parse tree	[parse tree]	2.0	2	3	3	3
112	58	English word	[English word]	2.0	2	2	0	0
113	58	answer agent	[answering agents]	2.0	2	3	1	1
114	58	human judgment	[human judgment]	2.0	2	3	1	1
115	58	control structure	[control structure]	2.0	2	3	1	1
116	58	word string	[word string]	2.0	2	3	2	2
117	58	target language	[target language]	2.0	2	2	0	0
118	58	syntax-based statistical machine translation mt	[Syntax-based statistical machine translation MT]	2.0	5	2	0	0
119	58	isomorphic Grammar	[Isomorphic Grammars]	2.0	2	2	0	0
120	58	noun phrase	[noun phrase]	2.0	2	3	3	3
121	58	one sense	[one sense]	2.0	2	2	0	0
122	58	hold multiple candidate for understanding result	[holding multiple candidates for understanding results]	2.0	6	2	0	0
123	58	empirical result	[empirical results]	2.0	2	3	1	1
124	58	wsd model	[WSD models]	2.0	2	3	1	1
125	58	single word	[single word]	2.0	2	2	0	0
126	58	correct one	[correct one]	2.0	2	2	0	0
127	58	input document	[input documents]	2.0	2	3	2	2
128	58	path-based inference	[Path-based inference]	2.0	2	2	0	0
129	58	intentional structure	[intentional structure]	2.0	2	2	0	0
130	58	text understanding	[text understanding]	2.0	2	3	1	1
131	58	HDAG Kernel	[HDAG Kernel]	2.0	2	2	0	0
132	58	lexical entry	[lexical entries]	2.0	2	2	0	0
133	58	learning phrase	[learning phrases]	2.0	2	3	1	1
134	58	oral communication	[Oral communication]	2.0	2	2	0	0
135	58	dialog system	[dialog systems]	2.0	2	3	2	2
136	58	error-correction rule	[error-correction rules]	2.0	2	2	0	0
137	58	sense-tagged datum	[sense-tagged data]	2.0	2	3	1	1
138	58	disjunctive value	[disjunctive values]	2.0	2	2	0	0
139	58	task-oriented dialog	[task-oriented dialogs]	2.0	2	2	0	0
140	58	such phrase	[such phrases]	2.0	2	2	0	0
141	58	two task	[two tasks]	2.0	2	3	1	1
142	58	language translation	[language translation]	2.0	2	3	2	2
143	58	semantic constraint	[semantic constraints]	2.0	2	3	1	1
144	58	translation quality	[translation quality]	2.0	2	3	3	3
145	58	spontaneous speech	[spontaneous speech]	2.0	2	2	0	0
146	58	missing fragment	[missing fragments]	2.0	2	3	1	1
147	58	accord to	[according to]	2.0	2	3	3	3
148	58	equivalent RCG	[equivalent RCG]	2.0	2	2	0	0
149	58	Japanese homophone error in compound noun	[Japanese homophone errors in compound nouns]	2.0	6	2	0	0
150	58	100 word	[100 words]	2.0	2	3	1	1
151	58	lexical substitution	[lexical substitutions]	2.0	2	2	0	0
152	58	combination method	[Combination methods]	2.0	2	2	0	0
153	58	summarization quality	[summarization quality]	2.0	2	3	3	3
154	58	actor paradigm	[actor paradigm]	2.0	2	2	0	0
155	58	taxonomic similarity	[taxonomic similarity]	2.0	2	2	0	0
156	58	important content	[important contents]	2.0	2	2	0	0
157	58	exponential time	[exponential time]	2.0	2	2	0	0
158	58	error rate	[error rate]	2.0	2	3	1	1
159	58	statistical model	[statistical models]	2.0	2	3	2	2
160	58	three strategy	[three strategies]	2.0	2	2	0	0
161	58	new domain	[new domains]	2.0	2	3	1	1
162	58	associative similarity	[associative similarity]	2.0	2	2	0	0
163	58	sentence plan	[sentence plan]	2.0	2	3	2	2
164	58	depend on	[depending on]	2.0	2	2	0	0
165	58	semantic equivalence	[semantic equivalence]	2.0	2	2	0	0
166	58	bilingual text	[bilingual texts]	2.0	2	2	0	0
167	58	word alignment	[word alignment]	2.0	2	3	3	3
168	58	translation process	[translation process]	2.0	2	2	0	0
169	58	defeasible reasoning	[defeasible reasoning]	2.0	2	2	0	0
170	58	competitive result	[competitive results]	2.0	2	2	0	0
171	58	speaker intention	[speaker's intention]	2.0	2	2	0	0
172	58	translation system	[translation system]	2.0	2	3	3	3
173	58	synchronous tag	[synchronous TAGs]	2.0	2	2	0	0
174	58	inference rule	[inference rules]	2.0	2	3	3	3
175	59	learning	[learning]	1.9923076923076923	1	21	14	13
176	60	analysis	[analysis]	1.9833333333333334	1	21	14	12
177	61	datum	[data]	1.975	1	21	10	8
178	62	application	[application]	1.9444444444444444	1	21	14	9
179	63	set	[set]	1.9142857142857144	1	22	20	7
180	64	question	[question]	1.9000000000000001	1	20	8	8
181	65	domain	[domain]	1.8846153846153848	1	20	15	13
182	66	structure	[structure]	1.85	1	20	15	10
183	67	theory	[theory]	1.8	1	19	8	8
184	68	phrase	[phrase]	1.7727272727272727	1	19	14	11
185	69	answer	[answer]	1.6875	1	18	9	8
186	69	document	[document]	1.6875	1	18	9	8
187	70	work	[work]	1.6714285714285717	1	18	9	7
188	71	ambiguity	[ambiguity]	1.6	1	17	6	6
189	71	research	[research]	1.6	1	17	8	8
190	71	paraphrase	[paraphrase]	1.6	1	17	7	7
191	72	SMT	[SMT]	1.5857142857142859	1	17	8	7
192	73	bigram language model	[bigram language model]	1.5849625007211563	3	1	0	0
193	73	unsupervised wsd system	[unsupervised WSD systems]	1.5849625007211563	3	1	0	0
194	73	word contribute match	[words contributing matches]	1.5849625007211563	3	1	0	0
195	73	use rule in addition	[using rules in addition]	1.5849625007211563	4	1	0	0
196	73	use of feature	[use of features]	1.5849625007211563	3	2	2	2
197	73	application of tag	[application of TAGs]	1.5849625007211563	3	2	2	2
198	73	thai-english language identification	[Thai-English language identification]	1.5849625007211563	3	1	0	0
199	73	analysis of context language	[analysis of context languages]	1.5849625007211563	4	1	0	0
200	73	English -speak analyst	[English -speaking analyst]	1.5849625007211563	3	1	0	0
201	73	web-based computer-assisted academic writing	[Web-based computer-assisted academic writing]	1.5849625007211563	4	1	0	0
202	73	output of recognition	[output of recognition]	1.5849625007211563	3	1	0	0
203	73	less extent entailment	[lesser extent entailment]	1.5849625007211563	3	1	0	0
204	73	Penn Treebank WSJ	[Penn Treebank WSJ]	1.5849625007211563	3	1	0	0
205	73	elicit subjective human judgment	[eliciting subjective human judgments]	1.5849625007211563	4	1	0	0
206	73	word-based trigram language model	[word-based trigram language model]	1.5849625007211563	4	1	0	0
207	73	use various tagging strategy	[using various tagging strategies]	1.5849625007211563	4	1	0	0
208	73	word sense disambiguation WSD	[word sense disambiguation WSD]	1.5849625007211563	4	1	0	0
209	73	sum of the number	[sum of the number]	1.5849625007211563	4	1	0	0
210	73	many more question	[many more questions]	1.5849625007211563	3	1	0	0
211	73	specific move in light	[specific move in light]	1.5849625007211563	4	1	0	0
212	73	appropriate pint in mt	[appropriate pints in MT]	1.5849625007211563	4	1	0	0
213	73	automatic evaluation metric	[Automatic evaluation metrics]	1.5849625007211563	3	1	0	0
214	73	range concatenation grammar	[range concatenation grammar]	1.5849625007211563	3	1	0	0
215	73	natural language question	[natural language question]	1.5849625007211563	3	1	0	0
216	73	Syntactic Role System	[Syntactic Role System]	1.5849625007211563	3	1	0	0
217	73	evidence of usefulness	[evidence of usefulness]	1.5849625007211563	3	1	0	0
218	73	confines of syntax	[confines of syntax]	1.5849625007211563	3	1	0	0
219	73	explication of inheritance	[explication of inheritance]	1.5849625007211563	3	1	0	0
220	73	measure of progress	[measure of progress]	1.5849625007211563	3	1	0	0
221	73	find the correct candidate	[Finding the correct candidates]	1.5849625007211563	4	1	0	0
222	73	object-oriented grammar model	[object-oriented grammar model]	1.5849625007211563	3	1	0	0
223	73	utterance of the discourse	[utterances of the discourse]	1.5849625007211563	4	1	0	0
224	73	compound noun component	[compound noun component]	1.5849625007211563	3	1	0	0
225	73	purpose of correction	[purposes of correction]	1.5849625007211563	3	1	0	0
226	73	activity in meeting situation	[activities in meeting situation]	1.5849625007211563	4	1	0	0
227	73	performance use asr output	[performance using ASR output]	1.5849625007211563	4	1	0	0
228	73	subset only 14 .0%	[subset only 14 .0%]	1.5849625007211563	4	1	0	0
229	73	speaker gaze information	[speaker's gaze information]	1.5849625007211563	3	1	0	0
230	73	terminal substring parsing	[terminal substring parsing]	1.5849625007211563	3	1	0	0
231	73	user's informational goal	[user's informational goals]	1.5849625007211563	3	1	0	0
232	73	more linguistically sophisticated representation	[more linguistically sophisticated representation]	1.5849625007211563	4	2	2	2
233	73	factor of 5 speed-up	[factor of 5 speed-up]	1.5849625007211563	4	1	0	0
234	73	two sentence planner	[two sentence planners]	1.5849625007211563	3	2	2	2
235	73	basis of translation example	[basis of translation examples]	1.5849625007211563	4	1	0	0
236	73	piece of work	[piece of work]	1.5849625007211563	3	1	0	0
237	73	name identification capability	[name identification capability]	1.5849625007211563	3	1	0	0
238	73	English sentence structure	[English sentence structure]	1.5849625007211563	3	1	0	0
239	73	challenge for the application	[challenge for the application]	1.5849625007211563	4	1	0	0
240	73	reduction in error	[reduction in error]	1.5849625007211563	3	2	2	2
241	73	polysemous word like sentence	[polysemous words like sentence]	1.5849625007211563	4	1	0	0
242	73	least 99% correct sentence	[least 99% correct sentences]	1.5849625007211563	4	1	0	0
243	73	impact of precision	[impact of precision]	1.5849625007211563	3	1	0	0
244	73	foundation explore English-Chinese bitext	[foundation exploring English-Chinese bitexts]	1.5849625007211563	4	1	0	0
245	73	probabilistic spectral mapping	[probabilistic spectral mapping]	1.5849625007211563	3	1	0	0
246	73	integration of image	[integration of image]	1.5849625007211563	3	1	0	0
247	73	benefit of system combination	[benefits of system combination]	1.5849625007211563	4	1	0	0
248	73	exploit linguistically sophisticated representation	[exploiting linguistically sophisticated representations]	1.5849625007211563	4	1	0	0
249	73	voice interactive system	[voice interactive system]	1.5849625007211563	3	1	0	0
250	73	access to text collection	[access to text collections]	1.5849625007211563	4	1	0	0
251	73	complexity of document	[complexity of documents]	1.5849625007211563	3	1	0	0
252	73	result NE system	[resulting NE system]	1.5849625007211563	3	1	0	0
253	73	need of a publication	[needs of a publication]	1.5849625007211563	4	1	0	0
254	73	rating on average	[rating on average]	1.5849625007211563	3	1	0	0
255	73	only scarce resource	[only scarce resources]	1.5849625007211563	3	1	0	0
256	73	choice of syntactic structure	[choice of syntactic structure]	1.5849625007211563	4	1	0	0
257	73	calculus of equivalence	[calculus of equivalences]	1.5849625007211563	3	1	0	0
258	73	use this novel approach	[using this novel approach]	1.5849625007211563	4	1	0	0
259	73	state of the art	[state of the art]	1.5849625007211563	4	2	1	1
260	73	he/she/man/woman for person ne	[he/she/man/woman for PERSON NE]	1.5849625007211563	4	1	0	0
261	73	instance of a pattern	[instance of a pattern]	1.5849625007211563	4	1	0	0
262	73	domain of sentence condensation	[domain of sentence condensation]	1.5849625007211563	4	1	0	0
263	73	method use expectation	[method using expectations]	1.5849625007211563	3	1	0	0
264	73	view of language definition	[view of language definition]	1.5849625007211563	4	1	0	0
265	73	Natural language processing	[Natural Language Processing]	1.5849625007211563	3	2	2	2
266	73	LNR research group	[LNR research group]	1.5849625007211563	3	1	0	0
267	73	predict top-level subtopic boundary	[predicting top-level subtopic boundaries]	1.5849625007211563	4	1	0	0
268	73	directly from weather datum	[directly from weather data]	1.5849625007211563	4	1	0	0
269	73	multi-site datum collection paradigm	[multi-site data collection paradigm]	1.5849625007211563	4	1	0	0
270	73	natural language generation system	[natural language generation system]	1.5849625007211563	4	1	0	0
271	73	user-defined performance threshold	[user-defined performance thresholds]	1.5849625007211563	3	1	0	0
272	73	CCLINC Korean-to-English translation system	[CCLINC Korean-to-English translation system]	1.5849625007211563	4	1	0	0
273	73	series of experiment	[series of experiments]	1.5849625007211563	3	1	0	0
274	73	source-language dependency parser	[source-language dependency parser]	1.5849625007211563	3	1	0	0
275	73	manually sense-tagged datum	[manually sense-tagged data]	1.5849625007211563	3	1	0	0
276	73	machine translation task	[machine translation task]	1.5849625007211563	3	1	0	0
277	73	chunk of information	[chunk of information]	1.5849625007211563	3	1	0	0
278	73	block selection criterion	[block selection criteria]	1.5849625007211563	3	1	0	0
279	73	backbone of the annotation	[backbone of the annotation]	1.5849625007211563	4	1	0	0
280	73	research in generation	[research in generation]	1.5849625007211563	3	1	0	0
281	73	quality of the datum	[quality of the data]	1.5849625007211563	4	1	0	0
282	73	concept for the ne	[concept for the NE]	1.5849625007211563	4	1	0	0
283	73	basis of a ontology	[basis of an ontology]	1.5849625007211563	4	1	0	0
284	73	February '92 test sentence	[February '92 test sentences]	1.5849625007211563	4	1	0	0
285	73	large training set	[larger training set]	1.5849625007211563	3	1	0	0
286	73	cooccurrence of category	[cooccurrence of categories]	1.5849625007211563	3	1	0	0
287	73	large class of word	[large class of words]	1.5849625007211563	4	1	0	0
288	73	two distinct task	[two distinct tasks]	1.5849625007211563	3	1	0	0
289	73	Air Travel Information System	[Air Travel Information System]	1.5849625007211563	4	1	0	0
290	73	database query application	[database query applications]	1.5849625007211563	3	1	0	0
291	73	parse reduction operate	[parse reduction operating]	1.5849625007211563	3	1	0	0
292	73	respond to the user	[responding to the user]	1.5849625007211563	4	1	0	0
293	73	construction of format	[construction of format]	1.5849625007211563	3	1	0	0
294	73	model TDMT on APs	[model TDMT on APs]	1.5849625007211563	4	1	0	0
295	73	adopt statistical technique	[adopting statistical techniques]	1.5849625007211563	3	1	0	0
296	73	large amount of speech	[large amount of speech]	1.5849625007211563	4	1	0	0
297	73	use of test-score semantics	[use of test-score semantics]	1.5849625007211563	4	1	0	0
298	73	redundancy elimination problem	[redundancy elimination problem]	1.5849625007211563	3	1	0	0
299	73	unsupervised structure finding method	[unsupervised structure finding methods]	1.5849625007211563	4	1	0	0
300	73	noisy channel framework	[noisy channel framework]	1.5849625007211563	3	1	0	0
301	73	variety of application	[variety of applications]	1.5849625007211563	3	2	1	1
302	73	non-literal aspect of communication	[non-literal aspects of communication]	1.5849625007211563	4	2	1	1
303	73	high dimensionality space	[high dimensionality space]	1.5849625007211563	3	1	0	0
304	73	standard statistical model	[standard statistical models]	1.5849625007211563	3	1	0	0
305	73	namely the forward-backward algorithm	[namely the forward-backward algorithm]	1.5849625007211563	4	1	0	0
306	73	progress in the field	[progress in the field]	1.5849625007211563	4	1	0	0
307	73	monolingual parallel corpus	[monolingual parallel corpora]	1.5849625007211563	3	1	0	0
308	73	linguistic generality available	[linguistic generality available]	1.5849625007211563	3	1	0	0
309	73	grammar coverage problem	[grammar coverage problems]	1.5849625007211563	3	1	0	0
310	73	probabilistic translation model	[probabilistic translation models]	1.5849625007211563	3	1	0	0
311	73	decision tree learning	[decision tree learning]	1.5849625007211563	3	2	1	1
312	73	very strong discourse effect	[very strong discourse effect]	1.5849625007211563	4	1	0	0
313	73	wide coverage English grammar	[wide coverage English grammar]	1.5849625007211563	4	1	0	0
314	73	discussion compare the expressiveness	[discussion comparing the expressiveness]	1.5849625007211563	4	1	0	0
315	73	abstract of research article	[abstracts of research articles]	1.5849625007211563	4	1	0	0
316	73	appropriate user modeling	[appropriate user modeling]	1.5849625007211563	3	1	0	0
317	73	framework assess development	[framework assessing developments]	1.5849625007211563	3	1	0	0
318	73	context of use	[context of use]	1.5849625007211563	3	1	0	0
319	73	range concatenation language	[range concatenation languages]	1.5849625007211563	3	1	0	0
320	73	human plausible reasoning	[human plausible reasoning]	1.5849625007211563	3	1	0	0
321	73	arbiter-based combination strategy	[arbiter-based combination strategies]	1.5849625007211563	3	1	0	0
322	73	knowledge of the domain	[knowledge of the domain]	1.5849625007211563	4	1	0	0
323	73	most exist approach	[most existing approaches]	1.5849625007211563	3	1	0	0
324	73	resultant bilingual corpus	[resultant bilingual corpus]	1.5849625007211563	3	1	0	0
325	73	problem with determiner	[problem with determiners]	1.5849625007211563	3	1	0	0
326	73	user in dialogue system	[user in dialogue systems]	1.5849625007211563	4	1	0	0
327	73	task place comma	[task placing commas]	1.5849625007211563	3	1	0	0
328	73	expert in text mining	[experts in text mining]	1.5849625007211563	4	1	0	0
329	73	meaning of a proposition	[meaning of a proposition]	1.5849625007211563	4	2	2	2
330	73	token classification task	[token classification task]	1.5849625007211563	3	1	0	0
331	73	own lexical hierarchy	[own lexical hierarchy]	1.5849625007211563	3	1	0	0
332	73	focus on noun phrase	[focusing on noun phrases]	1.5849625007211563	4	1	0	0
333	73	most text processing software	[most text processing software]	1.5849625007211563	4	1	0	0
334	73	baseline of 54 .55%)	[baseline of 54 .55%)]	1.5849625007211563	4	1	0	0
335	73	much more available resource	[much more available resource]	1.5849625007211563	4	1	0	0
336	73	result suppress the quantifier	[result suppressing the quantifier]	1.5849625007211563	4	1	0	0
337	73	approach to semantics	[approach to semantics]	1.5849625007211563	3	1	0	0
338	73	intelligibility of mt output	[intelligibility of MT output]	1.5849625007211563	4	1	0	0
339	73	relevance to default reasoning	[relevance to default reasoning]	1.5849625007211563	4	1	0	0
340	73	set of parsing flexibility	[set of parsing flexibilities]	1.5849625007211563	4	1	0	0
341	73	such as English	[such as English]	1.5849625007211563	3	1	0	0
342	73	source interval projection	[source interval projections]	1.5849625007211563	3	1	0	0
343	73	corpus of sentence	[corpus of sentences]	1.5849625007211563	3	1	0	0
344	73	continuous markov model cdhmm	[continuous Markov models CDHMM]	1.5849625007211563	4	1	0	0
345	73	object of the structure	[objects of the structures]	1.5849625007211563	4	1	0	0
346	73	machine translation (mt) evaluation	[machine translation (MT) evaluation]	1.5849625007211563	4	1	0	0
347	73	multiple machine translation system	[multiple machine translation systems]	1.5849625007211563	4	1	0	0
348	73	overview of the research	[overview of the research]	1.5849625007211563	4	1	0	0
349	73	sentence-aligned parallel corpus	[sentence-aligned parallel corpus]	1.5849625007211563	3	1	0	0
350	73	learning algorithm from datum	[learning algorithm from data]	1.5849625007211563	4	1	0	0
351	73	procedural dialog model	[procedural dialog model]	1.5849625007211563	3	1	0	0
352	73	interaction with the user	[interaction with the user]	1.5849625007211563	4	1	0	0
353	73	evaluation of language learner	[evaluation of language learners]	1.5849625007211563	4	1	0	0
354	73	framework of isomorphic Grammar	[framework of Isomorphic Grammars]	1.5849625007211563	4	1	0	0
355	73	principle-and-parameters language framework	[principle-and-parameters language framework]	1.5849625007211563	3	1	0	0
356	73	implicit reference resolution	[implicit reference resolution]	1.5849625007211563	3	1	0	0
357	73	true text through transformation	[true text through transformation]	1.5849625007211563	4	1	0	0
358	73	multi-level answer resolution algorithm	[multi-level answer resolution algorithm]	1.5849625007211563	4	1	0	0
359	73	question answer session	[question answering session]	1.5849625007211563	3	1	0	0
360	73	phone n-gram model	[phone n-gram model]	1.5849625007211563	3	1	0	0
361	73	people' answer to question	[people's answers to questions]	1.5849625007211563	4	2	1	1
362	73	resolve the ambiguity	[resolving the ambiguity]	1.5849625007211563	3	1	0	0
363	73	good-quality mt system	[good-quality MT system]	1.5849625007211563	3	1	0	0
364	73	issue of system response	[issue of system response]	1.5849625007211563	4	1	0	0
365	73	application back end	[application back end]	1.5849625007211563	3	1	0	0
366	73	series of modification	[series of modifications]	1.5849625007211563	3	1	0	0
367	73	role of user modeling	[role of user modeling]	1.5849625007211563	4	1	0	0
368	73	message pass protocol	[message passing protocols]	1.5849625007211563	3	1	0	0
369	73	predict top-level boundary	[predicting top-level boundaries]	1.5849625007211563	3	1	0	0
370	73	logical form language	[logical form language]	1.5849625007211563	3	1	0	0
371	73	natural language generator	[natural language generator]	1.5849625007211563	3	1	0	0
372	73	method for utterance classification	[method for utterance classification]	1.5849625007211563	4	1	0	0
373	73	strength of potential antecedence	[strength of potential antecedence]	1.5849625007211563	4	1	0	0
374	73	bi-directional english-spanish fragment	[bi-directional English-Spanish fragment]	1.5849625007211563	3	1	0	0
375	73	ne email .g he/she/man/woman	[NE e .g he/she/man/woman]	1.5849625007211563	4	1	0	0
376	73	method on a coverage	[method on a coverage]	1.5849625007211563	4	1	0	0
377	73	rule-based translation system	[rule-based translation systems]	1.5849625007211563	3	1	0	0
378	73	easy word classification	[easy word classification]	1.5849625007211563	3	1	0	0
379	73	information about meeting context	[information about meeting context]	1.5849625007211563	4	2	1	1
380	73	unsupervised learning approach	[unsupervised learning approach]	1.5849625007211563	3	2	1	1
381	73	means for lexicon specification	[means for lexicon specification]	1.5849625007211563	4	1	0	0
382	73	typical kind of user	[typical kinds of users]	1.5849625007211563	4	1	0	0
383	73	additional source of constraint	[additional source of constraint]	1.5849625007211563	4	1	0	0
384	73	heuristically-produced complete sentence	[heuristically-produced complete sentences]	1.5849625007211563	3	1	0	0
385	73	context of a dialogue	[context of a dialogue]	1.5849625007211563	4	1	0	0
386	73	polarization of the object	[polarization of the objects]	1.5849625007211563	4	1	0	0
387	73	novel customizable Ie paradigm	[novel customizable IE paradigm]	1.5849625007211563	4	1	0	0
388	73	intelligent interactive system	[intelligent interactive systems]	1.5849625007211563	3	1	0	0
389	73	impediment to progress	[impediment to progress]	1.5849625007211563	3	1	0	0
390	73	right of the island	[right of the islands]	1.5849625007211563	4	1	0	0
391	73	novel training method	[novel training method]	1.5849625007211563	3	1	0	0
392	73	so many possible structure	[so many possible structures]	1.5849625007211563	4	1	0	0
393	73	power of nlp technique	[power of NLP techniques]	1.5849625007211563	4	1	0	0
394	73	most likely answer candidate	[most likely answer candidates]	1.5849625007211563	4	1	0	0
395	73	such as BLEU	[such as BLEU]	1.5849625007211563	3	1	0	0
396	73	other cluster method	[other clustering methods]	1.5849625007211563	3	1	0	0
397	73	behaviour of cue phrase	[behaviour of cue phrases]	1.5849625007211563	4	1	0	0
398	73	translation quality of EBMT	[translation quality of EBMT]	1.5849625007211563	4	1	0	0
399	73	two core module	[two core modules]	1.5849625007211563	3	1	0	0
400	73	variant of tag	[variant of TAGs]	1.5849625007211563	3	1	0	0
401	73	general user modeling system	[general user modeling systems]	1.5849625007211563	4	1	0	0
402	73	improve ocr accuracy	[improving OCR accuracy]	1.5849625007211563	3	1	0	0
403	73	very noisy portion	[very noisy portion]	1.5849625007211563	3	1	0	0
404	73	proprietary stemmer above	[proprietary stemmer above]	1.5849625007211563	3	1	0	0
405	73	hand-crafted template-based approach	[hand-crafted template-based approaches]	1.5849625007211563	3	1	0	0
406	73	few common noun	[few common noun]	1.5849625007211563	3	1	0	0
407	73	even in the presence	[even in the presence]	1.5849625007211563	4	1	0	0
408	73	SMT research community	[SMT research community]	1.5849625007211563	3	1	0	0
409	73	mostly young woman	[mostly young women]	1.5849625007211563	3	1	0	0
410	73	reason for that difference	[reason for that difference]	1.5849625007211563	4	1	0	0
411	73	determine the conversational vehicle	[determining the conversational vehicle]	1.5849625007211563	4	1	0	0
412	73	owe to the problem	[Owing to the problem]	1.5849625007211563	4	1	0	0
413	73	degree of ambiguity	[degree of ambiguity]	1.5849625007211563	3	1	0	0
414	73	method for discourse segmentation	[method for discourse segmentation]	1.5849625007211563	4	1	0	0
415	73	use maximum likelihood method	[using maximum likelihood method]	1.5849625007211563	4	1	0	0
416	73	multiple consecutive word	[multiple consecutive words]	1.5849625007211563	3	1	0	0
417	73	approximation to all	[approximations to all]	1.5849625007211563	3	1	0	0
418	73	twenty switchboard dialogue	[twenty Switchboard dialogues]	1.5849625007211563	3	1	0	0
419	73	block identity email .g	[block identities e .g]	1.5849625007211563	4	1	0	0
420	73	NIST evaluation metric	[NIST evaluation metric]	1.5849625007211563	3	1	0	0
421	73	powerful personal computer	[powerful personal computers]	1.5849625007211563	3	1	0	0
422	73	retrieve example most similar	[retrieving examples most similar]	1.5849625007211563	4	1	0	0
423	73	dictionary of word form	[dictionaries of word forms]	1.5849625007211563	4	1	0	0
424	73	statistical learning Model SLM	[Statistical Learning Model SLM]	1.5849625007211563	4	1	0	0
425	73	modify the transition probability	[modifying the transition probabilities]	1.5849625007211563	4	1	0	0
426	73	large corpus resource	[large corpus resource]	1.5849625007211563	3	1	0	0
427	73	algorithm on a corpus	[algorithm on a corpus]	1.5849625007211563	4	1	0	0
428	73	basic of SMT	[basics of SMT]	1.5849625007211563	3	1	0	0
429	73	oracle use a network	[oracle using a network]	1.5849625007211563	4	1	0	0
430	73	novel classification method	[novel classification method]	1.5849625007211563	3	1	0	0
431	73	document preference sets( dop	[document preference sets( DoPS]	1.5849625007211563	4	1	0	0
432	73	frequency of event	[frequency of events]	1.5849625007211563	3	1	0	0
433	73	leave this counterexample unexplained	[leaving these counterexamples unexplained]	1.5849625007211563	4	1	0	0
434	73	specification of character size	[specification of character size]	1.5849625007211563	4	1	0	0
435	73	breakdown of the performance	[breakdown of the performance]	1.5849625007211563	4	1	0	0
436	73	series of test	[series of tests]	1.5849625007211563	3	1	0	0
437	73	arbitrary set of feature	[arbitrary set of features]	1.5849625007211563	4	1	0	0
438	73	further grammar tailoring	[further grammar tailoring]	1.5849625007211563	3	1	0	0
439	73	aspect of communication	[aspects of communication]	1.5849625007211563	3	2	2	2
440	73	quality of utterance	[quality of utterances]	1.5849625007211563	3	1	0	0
441	73	establish dependency relation	[establishing dependency relations]	1.5849625007211563	3	1	0	0
442	73	use this approach	[using this approach]	1.5849625007211563	3	2	1	1
443	73	attention of the participant	[attention of the participants]	1.5849625007211563	4	1	0	0
444	73	issue with regard	[issues with regard]	1.5849625007211563	3	1	0	0
445	73	understanding of scruffy text	[understanding of scruffy texts]	1.5849625007211563	4	1	0	0
446	73	rule-based machine translation module	[rule-based machine translation modules]	1.5849625007211563	4	1	0	0
447	73	paraphrase in one language	[paraphrases in one language]	1.5849625007211563	4	1	0	0
448	73	retrieval of long phrase	[retrieval of long phrases]	1.5849625007211563	4	1	0	0
449	73	standard 360K floppy	[standard 360K floppy]	1.5849625007211563	3	1	0	0
450	73	dialogue translation system	[dialogue translation system]	1.5849625007211563	3	1	0	0
451	73	method of structure-sharing	[method of structure-sharing]	1.5849625007211563	3	1	0	0
452	73	node-based inference rule	[Node-based inference rules]	1.5849625007211563	3	1	0	0
453	73	definition of the structure	[definitions of the structure]	1.5849625007211563	4	1	0	0
454	73	use of nlp technique	[use of NLP techniques]	1.5849625007211563	4	1	0	0
455	73	number of year	[number of years]	1.5849625007211563	3	1	0	0
456	73	new language pair	[new language pairs]	1.5849625007211563	3	1	0	0
457	73	current decoder implementation	[current decoder implementations]	1.5849625007211563	3	1	0	0
458	73	machine learning experiment center	[machine learning experiments centering]	1.5849625007211563	4	1	0	0
459	73	discourse understanding process	[discourse understanding process]	1.5849625007211563	3	2	1	1
460	73	neutral meaning representation	[neutral meaning representation]	1.5849625007211563	3	1	0	0
461	73	lexical-functional Grammar LFG	[Lexical-Functional Grammars LFG]	1.5849625007211563	3	1	0	0
462	73	set of graph	[sets of graphs]	1.5849625007211563	3	1	0	0
463	73	U .s military	[U .S military]	1.5849625007211563	3	1	0	0
464	73	most difficult SENSEVAL-2 noun	[most difficult SENSEVAL-2 nouns]	1.5849625007211563	4	1	0	0
465	73	topic detection task	[Topic Detection tasks]	1.5849625007211563	3	1	0	0
466	73	middle of the scale	[middle of the scale]	1.5849625007211563	4	1	0	0
467	73	use length-based criterion	[using length-based criterion]	1.5849625007211563	3	1	0	0
468	73	existence of another path	[existence of another path]	1.5849625007211563	4	1	0	0
469	73	leverage human language technology	[leveraging human language technology]	1.5849625007211563	4	1	0	0
470	73	sentence alignment task	[sentence alignment tasks]	1.5849625007211563	3	1	0	0
471	73	structure of the sequence	[structure of the sequence]	1.5849625007211563	4	1	0	0
472	73	maximization of translation accuracy	[maximization of translation accuracy]	1.5849625007211563	4	1	0	0
473	73	set of candidate parse	[set of candidate parses]	1.5849625007211563	4	1	0	0
474	73	focus post-process the output	[focus post-processing the output]	1.5849625007211563	4	1	0	0
475	73	existence of a instance	[existence of an instance]	1.5849625007211563	4	1	0	0
476	73	performance of the stemmer	[performance of the stemmer]	1.5849625007211563	4	1	0	0
477	73	vector space model	[vector space model]	1.5849625007211563	3	1	0	0
478	73	tree adjoin grammar	[tree adjoining grammar]	1.5849625007211563	3	1	0	0
479	73	modeling language acquisition	[modeling language acquisition]	1.5849625007211563	3	1	0	0
480	73	interactive environment TGE	[interactive environment TGE]	1.5849625007211563	3	1	0	0
481	73	order-sensitive string comparison method	[order-sensitive string comparison methods]	1.5849625007211563	4	1	0	0
482	73	use a DCG formalism	[using an DCG formalism]	1.5849625007211563	4	1	0	0
483	73	Prolog a programming language	[Prolog a programming language]	1.5849625007211563	4	1	0	0
484	73	example a disposition	[example a disposition]	1.5849625007211563	3	1	0	0
485	73	design for conversational agent	[design for conversational agents]	1.5849625007211563	4	1	0	0
486	73	strong impact on performance	[strong impact on performance]	1.5849625007211563	4	1	0	0
487	73	2 a real-time response	[2 a real-time response]	1.5849625007211563	4	1	0	0
488	73	example paraphrase ambiguous sentence	[examples paraphrasing ambiguous sentences]	1.5849625007211563	4	1	0	0
489	73	average retrieval time	[average retrieval times]	1.5849625007211563	3	1	0	0
490	73	natural language environment	[natural language environment]	1.5849625007211563	3	1	0	0
491	73	preliminary qualitative user evaluation	[preliminary qualitative user evaluation]	1.5849625007211563	4	1	0	0
492	73	relationship among intention	[relationships among intentions]	1.5849625007211563	3	1	0	0
493	73	such complex linguistic database	[such complex linguistic databases]	1.5849625007211563	4	1	0	0
494	73	many computational system	[many computational systems]	1.5849625007211563	3	1	0	0
495	73	determination of the property	[determination of the properties]	1.5849625007211563	4	1	0	0
496	73	linguistically sophisticated unit	[linguistically sophisticated units]	1.5849625007211563	3	1	0	0
497	73	previous paper [Zernik87]	[previous papers [Zernik87]]	1.5849625007211563	3	1	0	0
498	73	experimental computer program SPRINT	[experimental computer program SPRINT]	1.5849625007211563	4	1	0	0
499	73	pattern of node structure	[pattern of node structures]	1.5849625007211563	4	1	0	0
500	73	problem identify subtopic boundary	[problem identifying subtopic boundaries]	1.5849625007211563	4	1	0	0
501	73	force a particular interpretation	[forcing a particular interpretation]	1.5849625007211563	4	1	0	0
502	73	answer to such question	[answers to such questions]	1.5849625007211563	4	1	0	0
503	73	comparable non-parallel corpus	[comparable non-parallel corpora]	1.5849625007211563	3	1	0	0
504	73	English-Japanese mt system	[English-Japanese MT system]	1.5849625007211563	3	1	0	0
505	73	deterministic finite automaton	[deterministic finite automaton]	1.5849625007211563	3	1	0	0
506	73	simultaneously use less memory	[simultaneously using less memory]	1.5849625007211563	4	1	0	0
507	73	speed of processing	[speed of processing]	1.5849625007211563	3	1	0	0
508	73	logic-based grammar formalism	[logic-based grammar formalism]	1.5849625007211563	3	1	0	0
509	73	decision making process	[decision making process]	1.5849625007211563	3	1	0	0
510	73	sentence condensation system	[sentence condensation systems]	1.5849625007211563	3	1	0	0
511	73	predict subtopic boundary	[predicting subtopic boundaries]	1.5849625007211563	3	2	1	1
512	73	proper treatment of syntax	[proper treatment of syntax]	1.5849625007211563	4	1	0	0
513	73	translation memory system	[translation memory system]	1.5849625007211563	3	1	0	0
514	73	purpose of this research	[purpose of this research]	1.5849625007211563	4	1	0	0
515	73	language processing system	[language processing systems]	1.5849625007211563	3	2	1	1
516	73	37% reduction in error	[37% reduction in error]	1.5849625007211563	4	1	0	0
517	73	natural language generation community	[natural language generation community]	1.5849625007211563	4	1	0	0
518	73	Paramax language understand system	[Paramax language understanding system]	1.5849625007211563	4	1	0	0
519	73	numerous user study	[numerous user studies]	1.5849625007211563	3	1	0	0
520	73	idea of Pereira	[ideas of Pereira]	1.5849625007211563	3	1	0	0
521	73	cooccurrence-based word vector	[cooccurrence-based word vectors]	1.5849625007211563	3	1	0	0
522	73	less precise temporal adverb	[less precise temporal adverbs]	1.5849625007211563	4	1	0	0
523	73	strength of each	[strengths of each]	1.5849625007211563	3	1	0	0
524	73	formulas with disjunctive value	[formulas with disjunctive values]	1.5849625007211563	4	1	0	0
525	73	story link detection	[story link detection]	1.5849625007211563	3	1	0	0
526	73	distribution of nonverbal behavior	[distribution of nonverbal behaviors]	1.5849625007211563	4	1	0	0
527	73	Japanese homophone error	[Japanese homophone errors]	1.5849625007211563	3	2	2	2
528	73	answer resolution algorithm	[answer resolution algorithm]	1.5849625007211563	3	2	2	2
529	73	number of word form	[number of word forms]	1.5849625007211563	4	1	0	0
530	73	probabilistic lr parser	[probabilistic LR parser]	1.5849625007211563	3	1	0	0
531	73	more robust---an scheduling parser	[more robust---an scheduling parser]	1.5849625007211563	4	1	0	0
532	73	additional increase of correlation	[additional increase of correlation]	1.5849625007211563	4	1	0	0
533	73	map estimation approach	[MAP estimation approach]	1.5849625007211563	3	1	0	0
534	73	use of lexical substitution	[use of lexical substitutions]	1.5849625007211563	4	1	0	0
535	73	human annotation performance	[human annotation performance]	1.5849625007211563	3	1	0	0
536	73	grammar model for concurrent	[grammar model for concurrent]	1.5849625007211563	4	1	0	0
537	73	two rule-based sentence planner	[two rule-based sentence planners]	1.5849625007211563	4	1	0	0
538	73	combination with a range	[combination with a range]	1.5849625007211563	4	1	0	0
539	73	recovery technique for parse	[recovery technique for parses]	1.5849625007211563	4	1	0	0
540	73	implicit fuzzy quantifier	[implicit fuzzy quantifiers]	1.5849625007211563	3	1	0	0
541	73	mt evaluation technique	[MT evaluation techniques]	1.5849625007211563	3	1	0	0
542	73	prediction of missing fragment	[predictions of missing fragments]	1.5849625007211563	4	1	0	0
543	73	planning algorithm into Prolog	[planning algorithm into Prolog]	1.5849625007211563	4	1	0	0
544	73	modification of the document	[modification of the document]	1.5849625007211563	4	1	0	0
545	73	NTHU's statistic-based system	[NTHU's statistic-based system]	1.5849625007211563	3	1	0	0
546	73	natural language presupposition	[natural language presuppositions]	1.5849625007211563	3	2	1	1
547	73	dependency network representation	[dependency network representation]	1.5849625007211563	3	1	0	0
548	73	human language learning process	[human language learning process]	1.5849625007211563	4	1	0	0
549	73	style and/or topic	[style and/or topic]	1.5849625007211563	3	1	0	0
550	73	Japanese patent claim sentence	[Japanese patent claim sentences]	1.5849625007211563	4	1	0	0
551	73	evaluation of dialogue system	[evaluation of dialogue systems]	1.5849625007211563	4	1	0	0
552	73	amount of speech	[amount of speech]	1.5849625007211563	3	2	2	2
553	73	conversational context feature	[conversational context features]	1.5849625007211563	3	1	0	0
554	73	negative impact on model	[negative impact on models]	1.5849625007211563	4	1	0	0
555	73	Gaussian mixture observation density	[Gaussian mixture observation densities]	1.5849625007211563	4	1	0	0
556	73	result logical expression	[resulting logical expression]	1.5849625007211563	3	1	0	0
557	73	text from the web	[text from the web]	1.5849625007211563	4	1	0	0
558	73	application of synchronous tag	[application of synchronous TAGs]	1.5849625007211563	4	1	0	0
559	73	many attractive property	[many attractive properties]	1.5849625007211563	3	1	0	0
560	73	arbitrarily long phrase	[arbitrarily long phrases]	1.5849625007211563	3	1	0	0
561	73	much recent work	[much recent work]	1.5849625007211563	3	1	0	0
562	73	existence of a number	[existence of a number]	1.5849625007211563	4	1	0	0
563	73	focus on practical consideration	[focus on practical considerations]	1.5849625007211563	4	1	0	0
564	73	use this visualization method	[Using this visualization method]	1.5849625007211563	4	1	0	0
565	73	monolingual material Roget's Thesaurus	[monolingual material Roget's Thesaurus]	1.5849625007211563	4	1	0	0
566	73	most promising feature	[most promising features]	1.5849625007211563	3	1	0	0
567	73	mobile intelligent agent	[mobile intelligent agent]	1.5849625007211563	3	1	0	0
568	73	law of Hong Kong	[laws of Hong Kong]	1.5849625007211563	4	1	0	0
569	73	automatic scheme collect statistic	[automatic scheme collecting statistics]	1.5849625007211563	4	1	0	0
570	73	use the parse record	[using the parse record]	1.5849625007211563	4	1	0	0
571	73	Japanese revision support system	[Japanese revision support systems]	1.5849625007211563	4	1	0	0
572	73	maximum likelihood criterion	[maximum likelihood criterion]	1.5849625007211563	3	1	0	0
573	73	bidirectional grammar generation system	[bidirectional grammar generation system]	1.5849625007211563	4	1	0	0
574	73	frequent omission of argument	[frequent omissions of arguments]	1.5849625007211563	4	1	0	0
575	73	variety of SMT algorithm	[variety of SMT algorithms]	1.5849625007211563	4	1	0	0
576	73	principle of the system	[principles of the system]	1.5849625007211563	4	1	0	0
577	73	parsing-base ne rule	[parsing-based NE rules]	1.5849625007211563	3	1	0	0
578	73	such complex database	[such complex databases]	1.5849625007211563	3	1	0	0
579	73	legal text hierarchy	[legal text hierarchy]	1.5849625007211563	3	1	0	0
580	73	maximum entropy classifier	[maximum entropy classifier]	1.5849625007211563	3	1	0	0
581	73	76 million word	[76 million word]	1.5849625007211563	3	1	0	0
582	73	syntax-based machine translation	[Syntax-based machine translation]	1.5849625007211563	3	2	2	2
583	73	adjust the parameter	[adjusting the parameters]	1.5849625007211563	3	1	0	0
584	73	reduction of search space	[reduction of search space]	1.5849625007211563	4	1	0	0
585	73	high level of performance	[highest levels of performance]	1.5849625007211563	4	1	0	0
586	73	attend this workshop	[attending this workshop]	1.5849625007211563	3	1	0	0
587	73	stochastic tree-to-tree transducer	[stochastic tree-to-tree transducer]	1.5849625007211563	3	1	0	0
588	73	term of semantic coherence	[terms of semantic coherence]	1.5849625007211563	4	1	0	0
589	73	experiment evaluate the effectiveness	[Experiments evaluating the effectiveness]	1.5849625007211563	4	1	0	0
590	73	language learning model	[language learning model]	1.5849625007211563	3	1	0	0
591	73	research in user modeling	[research in user modeling]	1.5849625007211563	4	1	0	0
592	73	calibrate the rank	[calibrating the ranks]	1.5849625007211563	3	1	0	0
593	73	description of implementation	[description of implementation]	1.5849625007211563	3	1	0	0
594	73	literature on machine translation	[literature on machine translation]	1.5849625007211563	4	1	0	0
595	73	TAP-XL Automated analyst assistant	[TAP-XL Automated Analyst's Assistant]	1.5849625007211563	4	1	0	0
596	73	prior to training	[prior to training]	1.5849625007211563	3	1	0	0
597	73	current typical SMT model	[current typical SMT models]	1.5849625007211563	4	1	0	0
598	73	statistical machine translation (smt)	[statistical machine translation (SMT)]	1.5849625007211563	4	1	0	0
599	73	synchronous dependency insertion grammar	[synchronous dependency insertion grammar]	1.5849625007211563	4	2	1	1
600	73	different dialog schemata	[different dialog schemata]	1.5849625007211563	3	1	0	0
601	73	current statistical mt architecture	[current statistical MT architectures]	1.5849625007211563	4	1	0	0
602	73	speech datum from speaker	[speech data from speakers]	1.5849625007211563	4	1	0	0
603	73	less than 100 word	[less than 100 words]	1.5849625007211563	4	1	0	0
604	73	problem of training datum	[problem of training data]	1.5849625007211563	4	1	0	0
605	73	use sentence co-occurrence	[using sentence co-occurrences]	1.5849625007211563	3	1	0	0
606	73	distribution of demonstrative form	[distribution of demonstrative forms]	1.5849625007211563	4	1	0	0
607	73	apartment a discourse task	[apartment a discourse task]	1.5849625007211563	4	1	0	0
608	73	multimedia articulation of answer	[multimedia articulation of answers]	1.5849625007211563	4	1	0	0
609	73	pronoun with np-	[pronouns with NP-]	1.5849625007211563	3	1	0	0
610	73	use organizational strategy	[using organizational strategies]	1.5849625007211563	3	1	0	0
611	73	conditional loglinear model	[conditional loglinear models]	1.5849625007211563	3	1	0	0
612	73	certain layout structure	[certain layout structures]	1.5849625007211563	3	1	0	0
613	73	standard parser evaluation method	[standard parser evaluation methods]	1.5849625007211563	4	1	0	0
614	73	sense-tagged training datum	[sense-tagged training data]	1.5849625007211563	3	1	0	0
615	73	natural language datum	[natural language data]	1.5849625007211563	3	1	0	0
616	73	log-linear block bigram model	[log-linear block bigram model]	1.5849625007211563	4	1	0	0
617	73	abstract for digital learning	[abstracts for digital learning]	1.5849625007211563	4	1	0	0
618	73	derivation forest output	[derivation forest output]	1.5849625007211563	3	1	0	0
619	73	black-box ocr system	[black-box OCR systems]	1.5849625007211563	3	1	0	0
620	73	Thai key prediction	[Thai key prediction]	1.5849625007211563	3	1	0	0
621	73	status of a request	[status of a request]	1.5849625007211563	4	1	0	0
622	73	potential classify SCF datum	[potential classifying SCF data]	1.5849625007211563	4	1	0	0
623	73	aggregation of utterance	[aggregation of utterances]	1.5849625007211563	3	1	0	0
624	73	speaker group modeling	[speaker group modeling]	1.5849625007211563	3	1	0	0
625	73	term of the ability	[terms of the ability]	1.5849625007211563	4	1	0	0
626	73	chemical biological warfare	[chemical biological warfare]	1.5849625007211563	3	1	0	0
627	73	utilize the numbering system	[utilizing the numbering system]	1.5849625007211563	4	1	0	0
628	73	robust parsing mechanism	[robust parsing mechanism]	1.5849625007211563	3	1	0	0
629	73	vocabulary size by extrapolation	[vocabulary size by extrapolation]	1.5849625007211563	4	1	0	0
630	73	fair copy of text	[fair copy of texts]	1.5849625007211563	4	1	0	0
631	73	use of CCR	[use of CCRs]	1.5849625007211563	3	1	0	0
632	73	new translation model	[new translation model]	1.5849625007211563	3	1	0	0
633	73	basis of a theory	[basis of a theory]	1.5849625007211563	4	1	0	0
634	73	word N-gram model	[word N-gram models]	1.5849625007211563	3	1	0	0
635	73	same two node	[same two nodes]	1.5849625007211563	3	1	0	0
636	73	performance versus sentence	[performance vs sentences]	1.5849625007211563	3	1	0	0
637	73	example speech recognition	[example speech recognition]	1.5849625007211563	3	1	0	0
638	73	quality with paraphrase	[quality with paraphrases]	1.5849625007211563	3	1	0	0
639	73	error correction methodology	[error correction methodology]	1.5849625007211563	3	1	0	0
640	73	implicit intention component	[implicit intention component]	1.5849625007211563	3	1	0	0
641	73	construction of domain-independent lexica	[construction of domain-independent lexica]	1.5849625007211563	4	1	0	0
642	73	allow a speaker leeway	[allowing a speaker leeway]	1.5849625007211563	4	1	0	0
643	73	increasingly common speculative claim	[increasingly common speculative claim]	1.5849625007211563	4	1	0	0
644	73	title-driven name recognition	[title-driven name recognition]	1.5849625007211563	3	1	0	0
645	73	similar phrase-based model	[similar phrase-based models]	1.5849625007211563	3	1	0	0
646	73	DARPA communicator program	[DARPA Communicator program]	1.5849625007211563	3	1	0	0
647	73	scalability against vocabulary size	[scalability against vocabulary size]	1.5849625007211563	4	1	0	0
648	73	insufficient training datum	[insufficient training data]	1.5849625007211563	3	1	0	0
649	73	correspondence between language	[correspondences between languages]	1.5849625007211563	3	1	0	0
650	73	semantic error rate	[semantic error rate]	1.5849625007211563	3	1	0	0
651	73	more remote meteorological event	[more remote meteorological events]	1.5849625007211563	4	1	0	0
652	73	primarily knowledge-based mechanism	[primarily knowledge-based mechanisms]	1.5849625007211563	3	1	0	0
653	73	number of example discourse	[number of example discourses]	1.5849625007211563	4	1	0	0
654	73	most large text-understanding system	[Most large text-understanding systems]	1.5849625007211563	4	1	0	0
655	73	large number of experiment	[large number of experiments]	1.5849625007211563	4	1	0	0
656	73	Wall Street Journal	[Wall Street Journal]	1.5849625007211563	3	2	1	1
657	73	concept of sublanguage	[concept of sublanguage]	1.5849625007211563	3	1	0	0
658	73	first project stage	[first project stage]	1.5849625007211563	3	1	0	0
659	73	Unification categorial Grammar UCG	[Unification Categorial Grammar UCG]	1.5849625007211563	4	1	0	0
660	73	recall of 98%	[recall of 98%]	1.5849625007211563	3	1	0	0
661	73	non-redundant lexicon specification	[non-redundant lexicon specification]	1.5849625007211563	3	1	0	0
662	73	various lexical feature	[various lexical features]	1.5849625007211563	3	1	0	0
663	73	automatic parse-based evaluation	[automatic parse-based evaluation]	1.5849625007211563	3	1	0	0
664	73	method analyze metaphor	[method analyzing metaphors]	1.5849625007211563	3	1	0	0
665	73	Senseval series of workshop	[Senseval series of workshops]	1.5849625007211563	4	1	0	0
666	73	two different language pair	[two different language pairs]	1.5849625007211563	4	1	0	0
667	73	design of this module	[design of this module]	1.5849625007211563	4	1	0	0
668	73	degree of success	[degree of success]	1.5849625007211563	3	1	0	0
669	73	prototype concordancer care	[prototype concordancer CARE]	1.5849625007211563	3	1	0	0
670	73	representation of the equivalence	[representation of the equivalence]	1.5849625007211563	4	1	0	0
671	73	significant improvement in performance	[significant improvements in performance]	1.5849625007211563	4	1	0	0
672	73	use a different lm	[using a different LM]	1.5849625007211563	4	1	0	0
673	73	syntax in machine translation	[syntax in machine translation]	1.5849625007211563	4	1	0	0
674	73	various structure string	[various structures strings]	1.5849625007211563	3	1	0	0
675	73	genuinely useful software infrastructure	[genuinely useful software infrastructure]	1.5849625007211563	4	1	0	0
676	73	very powerful tool	[very powerful tools]	1.5849625007211563	3	1	0	0
677	73	point in a conversation	[point in a conversation]	1.5849625007211563	4	1	0	0
678	73	structure of node	[structure of nodes]	1.5849625007211563	3	1	0	0
679	73	much-studied discourse task	[much-studied discourse task]	1.5849625007211563	3	1	0	0
680	73	assign rank to referent	[assigning ranks to referents]	1.5849625007211563	4	1	0	0
681	73	value of this approach	[value of this approach]	1.5849625007211563	4	1	0	0
682	73	method identify systematic pattern	[method identifying systematic patterns]	1.5849625007211563	4	1	0	0
683	73	sum of each character	[sum of each character]	1.5849625007211563	4	1	0	0
684	73	CSR pilot corpus	[CSR pilot corpus]	1.5849625007211563	3	1	0	0
685	73	system against the datum	[system against the data]	1.5849625007211563	4	1	0	0
686	73	practical translation use	[practical translation use]	1.5849625007211563	3	1	0	0
687	73	mixture trigram model	[mixture trigram models]	1.5849625007211563	3	1	0	0
688	73	referent of the predicate	[referents of the predicate]	1.5849625007211563	4	1	0	0
689	73	other current syntax formalism	[other current syntax formalisms]	1.5849625007211563	4	1	0	0
690	73	evaluation of system	[evaluation of system]	1.5849625007211563	3	1	0	0
691	73	part of speech information	[part of speech information]	1.5849625007211563	4	1	0	0
692	73	optical character recognition	[optical character recognition]	1.5849625007211563	3	2	2	2
693	73	new similarity measure	[new similarity measures]	1.5849625007211563	3	1	0	0
694	73	handle unknown word	[handling unknown words]	1.5849625007211563	3	1	0	0
695	73	two distinct phase	[two distinct phases]	1.5849625007211563	3	1	0	0
696	73	motivation for this effort	[motivation for this effort]	1.5849625007211563	4	1	0	0
697	73	set of problem	[set of problems]	1.5849625007211563	3	1	0	0
698	73	use of this approach	[use of this approach]	1.5849625007211563	4	1	0	0
699	73	use a variant	[using a variant]	1.5849625007211563	3	1	0	0
700	73	web-based computer-assisted writing	[Web-based computer-assisted writing]	1.5849625007211563	3	1	0	0
701	73	area of anaphora resolution	[area of anaphora resolution]	1.5849625007211563	4	1	0	0
702	73	additional transfer mapping	[additional transfer mappings]	1.5849625007211563	3	1	0	0
703	73	lexical information for use	[lexical information for use]	1.5849625007211563	4	1	0	0
704	73	human language learner	[human language learners]	1.5849625007211563	3	1	0	0
705	73	suitable superset of L	[suitable superset of L]	1.5849625007211563	4	1	0	0
706	73	acquire the knowledge	[acquiring the knowledge]	1.5849625007211563	3	1	0	0
707	73	use a 5K vocabulary	[using a 5K vocabulary]	1.5849625007211563	4	1	0	0
708	73	standard chart-based algorithm	[standard chart-based algorithms]	1.5849625007211563	3	1	0	0
709	73	b perspective-taking in reference	[b perspective-taking in reference]	1.5849625007211563	4	1	0	0
710	73	text retrieval application	[text retrieval application]	1.5849625007211563	3	1	0	0
711	73	hand-crafted template-based generation component	[hand-crafted template-based generation component]	1.5849625007211563	4	1	0	0
712	73	set of up extract	[set of up extracts]	1.5849625007211563	4	1	0	0
713	73	context free language	[context free languages]	1.5849625007211563	3	1	0	0
714	73	transcription error inevitable	[transcription errors inevitable]	1.5849625007211563	3	1	0	0
715	73	understand system non-monotonic reasoning	[understanding system non-monotonic reasoning]	1.5849625007211563	4	1	0	0
716	73	connection to Montague semantics	[connection to Montague semantics]	1.5849625007211563	4	1	0	0
717	73	two distinct dataset	[two distinct datasets]	1.5849625007211563	3	1	0	0
718	73	kind of signal	[kinds of signals]	1.5849625007211563	3	1	0	0
719	73	integration of image understanding	[integration of image understanding]	1.5849625007211563	4	1	0	0
720	73	process transform a disposition	[process transforming a disposition]	1.5849625007211563	4	1	0	0
721	73	few mutually equivalent reading	[fewer mutually equivalent readings]	1.5849625007211563	4	1	0	0
722	73	current mt architecture	[current MT architectures]	1.5849625007211563	3	1	0	0
723	73	million of feature	[millions of features]	1.5849625007211563	3	1	0	0
724	73	technical abstracting industry	[technical abstracting industry]	1.5849625007211563	3	1	0	0
725	73	decision on the basis	[decisions on the basis]	1.5849625007211563	4	1	0	0
726	73	loss in translation quality	[loss in translation quality]	1.5849625007211563	4	1	0	0
727	73	source dependency parse	[source dependency parse]	1.5849625007211563	3	1	0	0
728	73	local segment contiguity model	[local segment contiguity models]	1.5849625007211563	4	1	0	0
729	73	model of verbal interaction	[models of verbal interaction]	1.5849625007211563	4	1	0	0
730	73	cover that lexical gap	[covering that lexical gap]	1.5849625007211563	4	1	0	0
731	73	statistical machine translation quality	[statistical machine translation quality]	1.5849625007211563	4	1	0	0
732	73	two parsing algorithm	[two parsing algorithms]	1.5849625007211563	3	1	0	0
733	73	enrichment of human-machine interaction	[enrichment of human-machine interactions]	1.5849625007211563	4	1	0	0
734	73	speech recognition hypothese	[speech recognition hypotheses]	1.5849625007211563	3	2	2	2
735	73	descriptor from individual phrase	[descriptors from individual phrases]	1.5849625007211563	4	1	0	0
736	73	model of the world	[model of the world]	1.5849625007211563	4	1	0	0
737	73	Arabic word segmenter	[Arabic word segmenter]	1.5849625007211563	3	1	0	0
738	73	analysis of corpus	[analysis of corpus]	1.5849625007211563	3	1	0	0
739	73	translation output sufficient	[translation output sufficient]	1.5849625007211563	3	1	0	0
740	73	precision of 70%	[precision of 70%]	1.5849625007211563	3	1	0	0
741	73	commercial application of some	[commercial applications of some]	1.5849625007211563	4	1	0	0
742	73	undisambiguated corpus datum	[undisambiguated corpus data]	1.5849625007211563	3	1	0	0
743	73	sole training resource	[sole training resources]	1.5849625007211563	3	1	0	0
744	73	development of a model	[development of a model]	1.5849625007211563	4	1	0	0
745	73	sentence in a abstract	[sentences in a abstract]	1.5849625007211563	4	1	0	0
746	73	demonstration at ACL	[demonstration at ACL]	1.5849625007211563	3	1	0	0
747	73	new sentence hypothese	[N-Best sentence hypotheses]	1.5849625007211563	3	1	0	0
748	73	significantly good translation quality	[significantly better translation quality]	1.5849625007211563	4	1	0	0
749	73	incremental grammar development	[incremental grammar development]	1.5849625007211563	3	1	0	0
750	73	block pair of phrase	[blocks pairs of phrases]	1.5849625007211563	4	1	0	0
751	73	interesting information piece	[interesting information piece]	1.5849625007211563	3	1	0	0
752	73	counterexample to the theory	[counterexamples to the theory]	1.5849625007211563	4	1	0	0
753	73	employ machine learning technique	[employing machine learning techniques]	1.5849625007211563	4	1	0	0
754	73	problem in machine translation	[problems in machine translation]	1.5849625007211563	4	1	0	0
755	73	goal of this work	[goal of this work]	1.5849625007211563	4	1	0	0
756	73	lexical cohesion-based approach alone	[lexical cohesion-based approach alone]	1.5849625007211563	4	1	0	0
757	73	quality of the paraphrase	[quality of the paraphrases]	1.5849625007211563	4	1	0	0
758	73	level of textual representation	[levels of textual representation]	1.5849625007211563	4	1	0	0
759	73	standard term weighting	[standard term weighting]	1.5849625007211563	3	1	0	0
760	73	target language word segmentation	[target language word segmentation]	1.5849625007211563	4	1	0	0
761	73	traditional information retrieval technique	[Traditional information retrieval techniques]	1.5849625007211563	4	1	0	0
762	73	currently under implementation	[currently under implementation]	1.5849625007211563	3	1	0	0
763	73	factor of 2 .5	[factor of 2 .5]	1.5849625007211563	4	1	0	0
764	73	big performance gain	[bigger performance gains]	1.5849625007211563	3	1	0	0
765	73	marine weather forecast	[marine weather forecasts]	1.5849625007211563	3	1	0	0
766	73	about 110 ,000 word	[about 110 ,000 words]	1.5849625007211563	4	1	0	0
767	73	other linguistic tool	[other linguistic tools]	1.5849625007211563	3	1	0	0
768	73	about 75% accuracy	[about 75% accuracy]	1.5849625007211563	3	1	0	0
769	73	miss out word	[missing out words]	1.5849625007211563	3	1	0	0
770	73	result of the comparison	[result of the comparison]	1.5849625007211563	4	1	0	0
771	73	predict new linguistic concept	[predicting new linguistic concepts]	1.5849625007211563	4	1	0	0
772	73	number of attribute sequence	[number of attribute sequences]	1.5849625007211563	4	1	0	0
773	73	speech recognition word	[speech recognition word]	1.5849625007211563	3	1	0	0
774	73	method interactively visualize	[method interactively visualizing]	1.5849625007211563	3	1	0	0
775	73	use in error correction	[use in error correction]	1.5849625007211563	4	1	0	0
776	73	set of paraphrase	[set of paraphrases]	1.5849625007211563	3	1	0	0
777	73	arbitrary HTML document	[arbitrary HTML documents]	1.5849625007211563	3	1	0	0
778	73	output of mt system	[outputs of MT system]	1.5849625007211563	4	1	0	0
779	73	phrase boundary heuristic	[phrase boundary heuristics]	1.5849625007211563	3	1	0	0
780	73	power of phrasal SMT	[power of phrasal SMT]	1.5849625007211563	4	1	0	0
781	73	lexical cross-relation between English	[lexical cross-relations between English]	1.5849625007211563	4	1	0	0
782	73	portable English understanding system	[portable English understanding system]	1.5849625007211563	4	1	0	0
783	73	monolingual unannotated text	[Monolingual unannotated text]	1.5849625007211563	3	1	0	0
784	73	head-orientation via valency constraint	[head-orientation via valency constraints]	1.5849625007211563	4	1	0	0
785	73	other Slavonic language	[other Slavonic languages]	1.5849625007211563	3	1	0	0
786	73	number of question	[number of questions]	1.5849625007211563	3	1	0	0
787	73	use of expectation	[use of expectations]	1.5849625007211563	3	1	0	0
788	73	information retrieval task	[information retrieval task]	1.5849625007211563	3	1	0	0
789	73	advantage of predicate-argument structure	[advantage of predicate-argument structures]	1.5849625007211563	4	1	0	0
790	73	result span edge	[resulting spanning edges]	1.5849625007211563	3	1	0	0
791	73	similarity-driven transfer System SimTran	[Similarity-driven Transfer System SimTran]	1.5849625007211563	4	1	0	0
792	73	impact on performance	[impact on performance]	1.5849625007211563	3	2	1	1
793	73	language learning experiment	[language learning experiment]	1.5849625007211563	3	1	0	0
794	73	unit of translation	[units of translation]	1.5849625007211563	3	1	0	0
795	73	point of the discourse	[point of the discourse]	1.5849625007211563	4	1	0	0
796	73	speak in fragment	[speaking in fragments]	1.5849625007211563	3	1	0	0
797	73	set of coherence rule	[set of coherence rules]	1.5849625007211563	4	1	0	0
798	73	total processing time	[total processing time]	1.5849625007211563	3	1	0	0
799	73	past 5 year	[past 5 years]	1.5849625007211563	3	1	0	0
800	73	role in syntax proper	[role in syntax proper]	1.5849625007211563	4	1	0	0
801	73	head-oriented notion of valency	[head-oriented notions of valency]	1.5849625007211563	4	1	0	0
802	73	sum of a representation	[sum of a representation]	1.5849625007211563	4	1	0	0
803	73	individual partial theory	[individual partial theories]	1.5849625007211563	3	1	0	0
804	73	three relatively domain-independent capability	[three relatively domain-independent capabilities]	1.5849625007211563	4	1	0	0
805	73	reference word string	[reference word string]	1.5849625007211563	3	1	0	0
806	73	two-step clustering process	[two-step clustering process]	1.5849625007211563	3	1	0	0
807	73	reappraise this insightful counterexample	[reappraising these insightful counterexamples]	1.5849625007211563	4	1	0	0
808	73	language of interest	[language of interest]	1.5849625007211563	3	1	0	0
809	73	spatial description in Japanese	[spatial descriptions in Japanese]	1.5849625007211563	4	1	0	0
810	73	paraphrasing of multiple output	[paraphrasing of multiple output]	1.5849625007211563	4	1	0	0
811	73	out-of-domain bilingual corpus	[out-of-domain bilingual corpus]	1.5849625007211563	3	2	1	1
812	73	task of ambiguity resolution	[task of ambiguity resolution]	1.5849625007211563	4	1	0	0
813	73	entire language system	[entire language system]	1.5849625007211563	3	1	0	0
814	73	version of synchronous grammar	[version of synchronous grammars]	1.5849625007211563	4	1	0	0
815	73	basis of feedback	[basis of feedback]	1.5849625007211563	3	1	0	0
816	73	process describe the layout	[process describing the layout]	1.5849625007211563	4	1	0	0
817	73	good other strategy	[best other strategies]	1.5849625007211563	3	1	0	0
818	73	Educational Research method course	[Educational Research Methods course]	1.5849625007211563	4	1	0	0
819	73	over 90 ,000 entry	[over 90 ,000 entries]	1.5849625007211563	4	1	0	0
820	73	take negligible runtime	[taking negligible runtime]	1.5849625007211563	3	1	0	0
821	73	use of BLEU	[use of BLEU]	1.5849625007211563	3	2	1	1
822	73	characteristic of this method	[characteristics of this method]	1.5849625007211563	4	1	0	0
823	73	use of dependency pointer	[use of dependency pointers]	1.5849625007211563	4	1	0	0
824	73	abstraction of the focus	[abstraction of the focus]	1.5849625007211563	4	1	0	0
825	73	re-utterance of that portion	[re-utterance of that portion]	1.5849625007211563	4	1	0	0
826	73	corpus of ,000 word	[corpus of ,000 words]	1.5849625007211563	4	2	2	2
827	73	contrast to previous work	[contrast to previous work]	1.5849625007211563	4	1	0	0
828	73	limited-domain computer system	[limited-domain computer system]	1.5849625007211563	3	1	0	0
829	73	10 .4m English word	[10 .4M English words]	1.5849625007211563	4	1	0	0
830	73	inheritance in hierarchy	[inheritance in hierarchies]	1.5849625007211563	3	1	0	0
831	73	MIT Lincoln Laboratory	[MIT Lincoln Laboratory]	1.5849625007211563	3	1	0	0
832	73	concurrency of computation	[concurrency of computation]	1.5849625007211563	3	1	0	0
833	73	number of copy	[number of copies]	1.5849625007211563	3	1	0	0
834	73	introduction of substitution cost	[introduction of substitution costs]	1.5849625007211563	4	1	0	0
835	73	probabilistic context-free grammar work	[probabilistic context-free grammars working]	1.5849625007211563	4	1	0	0
836	73	namely parameter smoothing	[namely parameter smoothing]	1.5849625007211563	3	1	0	0
837	73	generate from such redundancy	[generating from such redundancy]	1.5849625007211563	4	1	0	0
838	73	virtue of the fact	[virtue of the fact]	1.5849625007211563	4	1	0	0
839	73	statistical Machine Translation SMT	[Statistical Machine Translation SMT]	1.5849625007211563	4	1	0	0
840	73	three minute per extract	[three minutes per extract]	1.5849625007211563	4	1	0	0
841	73	knowledge about category transition	[knowledge about category transitions]	1.5849625007211563	4	1	0	0
842	73	tree-based ordering model	[tree-based ordering model]	1.5849625007211563	3	1	0	0
843	73	task of equivalence classification	[task of equivalence classification]	1.5849625007211563	4	1	0	0
844	73	man like young woman	[men like young women]	1.5849625007211563	4	2	2	2
845	73	mixture of method	[mixture of methods]	1.5849625007211563	3	1	0	0
846	73	such case-based mt cbmt	[such case-based MT CBMT]	1.5849625007211563	4	1	0	0
847	73	mt system design	[MT system design]	1.5849625007211563	3	1	0	0
848	73	light of several aspect	[light of several aspects]	1.5849625007211563	4	1	0	0
849	73	standard text browser	[standard text browser]	1.5849625007211563	3	1	0	0
850	73	assessment of language pair	[assessment of language pairs]	1.5849625007211563	4	1	0	0
851	73	full linguistic analysis	[full linguistic analysis]	1.5849625007211563	3	1	0	0
852	73	presence of unknown word	[presence of unknown words]	1.5849625007211563	4	1	0	0
853	73	Machine Translation (mt) system	[Machine Translation (MT) systems]	1.5849625007211563	4	1	0	0
854	73	small-sized bilingual corpus	[small-sized bilingual corpus]	1.5849625007211563	3	1	0	0
855	73	three-tiered approach for evaluation	[three-tiered approach for evaluation]	1.5849625007211563	4	1	0	0
856	73	source language sentence	[source language sentences]	1.5849625007211563	3	1	0	0
857	73	coordination use parsing strategy	[coordination using parsing strategies]	1.5849625007211563	4	1	0	0
858	73	unification-based grammar formalism	[unification-based grammar formalism]	1.5849625007211563	3	2	1	1
859	73	highly inflective language	[highly inflective languages]	1.5849625007211563	3	1	0	0
860	73	syntactic ambiguity resolution	[syntactic ambiguity resolution]	1.5849625007211563	3	1	0	0
861	73	inheritance as a means	[inheritance as a means]	1.5849625007211563	4	1	0	0
862	73	Markov model hmm	[Markov models HMM]	1.5849625007211563	3	1	0	0
863	73	IE-enhanced text browser	[IE-enhanced text browsers]	1.5849625007211563	3	1	0	0
864	73	certainty of the conclusion	[certainty of the conclusions]	1.5849625007211563	4	1	0	0
865	73	structure of purpose	[structure of purposes]	1.5849625007211563	3	1	0	0
866	73	new user of tool	[new users of tool]	1.5849625007211563	4	1	0	0
867	73	even large improvement	[even larger improvements]	1.5849625007211563	3	1	0	0
868	73	Air Travel Information Service	[Air Travel Information Service]	1.5849625007211563	4	1	0	0
869	73	second language learning	[second language learning]	1.5849625007211563	3	1	0	0
870	73	knowledge-based acquisition of grammar	[knowledge-based acquisition of grammars]	1.5849625007211563	4	1	0	0
871	73	two evaluation measure	[two evaluation measures]	1.5849625007211563	3	1	0	0
872	73	Japanese kanji-kana character	[Japanese kanji-kana characters]	1.5849625007211563	3	1	0	0
873	73	utterance classification performance	[utterance classification performance]	1.5849625007211563	3	1	0	0
874	73	broad coverage lexicon	[broad coverage lexicon]	1.5849625007211563	3	1	0	0
875	73	individual lexical rule	[individual lexical rules]	1.5849625007211563	3	1	0	0
876	73	more effective CFG filter	[more effective CFG filter]	1.5849625007211563	4	1	0	0
877	73	semantics for feature structure	[semantics for feature structures]	1.5849625007211563	4	1	0	0
878	73	word sense disambigation model	[word sense disambigation models]	1.5849625007211563	4	1	0	0
879	73	adaptive dynamic word formation	[adaptive dynamic word formation]	1.5849625007211563	4	1	0	0
880	73	object-oriented natural language parsing	[object-oriented natural language parsing]	1.5849625007211563	4	1	0	0
881	73	kind of word vector	[kind of word vectors]	1.5849625007211563	4	1	0	0
882	73	different information in memory	[different information in memory]	1.5849625007211563	4	1	0	0
883	73	current SMT model	[current SMT models]	1.5849625007211563	3	2	1	1
884	73	multilingual multimedia datum	[multilingual multimedia data]	1.5849625007211563	3	1	0	0
885	73	machine learning approach	[machine learning approach]	1.5849625007211563	3	1	0	0
886	73	starting point the beginning	[starting point the beginning]	1.5849625007211563	4	1	0	0
887	73	issue of domain dependence	[issue of domain dependence]	1.5849625007211563	4	1	0	0
888	73	precision of 96%	[precision of 96%]	1.5849625007211563	3	1	0	0
889	73	database query paraphrase	[database query paraphrase]	1.5849625007211563	3	1	0	0
890	73	crucial issue recently	[crucial issue recently]	1.5849625007211563	3	1	0	0
891	73	advantage of the sparsity	[advantage of the sparsity]	1.5849625007211563	4	1	0	0
892	73	several level of chunk	[several levels of chunks]	1.5849625007211563	4	1	0	0
893	73	purpose build spelling-checker	[purpose building spelling-checkers]	1.5849625007211563	3	1	0	0
894	73	improve machine translation output	[improving machine translation outputs]	1.5849625007211563	4	1	0	0
895	73	significantly good result	[significantly better results]	1.5849625007211563	3	1	0	0
896	73	sentence-level semantic equivalence classification	[sentence-level semantic equivalence classification]	1.5849625007211563	4	1	0	0
897	73	other approach to wsi	[other approaches to WSI]	1.5849625007211563	4	1	0	0
898	73	translation of each other	[translations of each other]	1.5849625007211563	4	1	0	0
899	73	uncertainty in expert system	[uncertainty in expert systems]	1.5849625007211563	4	1	0	0
900	73	application of this technology	[applications of this technology]	1.5849625007211563	4	1	0	0
901	73	word-dependent substitution cost	[word-dependent substitution costs]	1.5849625007211563	3	1	0	0
902	73	cluster polysemic verb	[clustering polysemic verbs]	1.5849625007211563	3	1	0	0
903	73	work computer program	[working computer program]	1.5849625007211563	3	1	0	0
904	73	information in English	[information in English]	1.5849625007211563	3	1	0	0
905	73	use this idea together	[Using these ideas together]	1.5849625007211563	4	1	0	0
906	73	conceptual Dependency diagram CDD	[Conceptual Dependency Diagram CDD]	1.5849625007211563	4	1	0	0
907	73	performance of a oracle	[performance of an oracle]	1.5849625007211563	4	1	0	0
908	73	various property of discourse	[Various properties of discourse]	1.5849625007211563	4	1	0	0
909	73	knowledge of surface English	[knowledge of surface English]	1.5849625007211563	4	1	0	0
910	73	knowledge-based information retrieval	[knowledge-based information retrieval]	1.5849625007211563	3	1	0	0
911	73	such as word	[such as words]	1.5849625007211563	3	1	0	0
912	73	paper in English	[papers in English]	1.5849625007211563	3	1	0	0
913	73	English-Japanese machine translation	[English-Japanese machine translation]	1.5849625007211563	3	1	0	0
914	73	statistical model of wh-question	[statistical models of WH-questions]	1.5849625007211563	4	1	0	0
915	73	underlying linguistic theory	[underlying linguistic theory]	1.5849625007211563	3	1	0	0
916	73	approach of KPSG	[approach of KPSG]	1.5849625007211563	3	1	0	0
917	73	two stage of 1	[two stages of 1]	1.5849625007211563	4	1	0	0
918	73	natural language text	[natural language texts]	1.5849625007211563	3	2	1	1
919	73	log-linear maximum-entropy model	[log-linear maximum-entropy models]	1.5849625007211563	3	1	0	0
920	73	underlying concurrent computation model	[underlying concurrent computation model]	1.5849625007211563	4	1	0	0
921	73	robust island-based parsing method	[robust island-based parsing method]	1.5849625007211563	4	1	0	0
922	73	poor syntactic construction	[poor syntactic construction]	1.5849625007211563	3	1	0	0
923	73	boundary between root	[boundary between root]	1.5849625007211563	3	1	0	0
924	73	adaptive learning procedure	[adaptive learning procedure]	1.5849625007211563	3	1	0	0
925	73	success of ensemble method	[success of ensemble methods]	1.5849625007211563	4	1	0	0
926	73	quality of multilingual summary	[quality of multilingual summaries]	1.5849625007211563	4	1	0	0
927	73	exist probabilistic parser	[existing probabilistic parser]	1.5849625007211563	3	1	0	0
928	73	language pair like English-Chinese	[language pairs like English-Chinese]	1.5849625007211563	4	1	0	0
929	73	smta double check	[SmTA double check]	1.5849625007211563	3	1	0	0
930	73	path-based inference rule	[Path-based inference rules]	1.5849625007211563	3	2	1	1
931	73	individual coordinate conjunction	[individual coordinate conjunction]	1.5849625007211563	3	1	0	0
932	73	block unigram model	[block unigram model]	1.5849625007211563	3	1	0	0
933	73	full linguistic analysis component	[full linguistic analysis component]	1.5849625007211563	4	1	0	0
934	73	long distance constraint	[long distance constraints]	1.5849625007211563	3	1	0	0
935	73	costly dependency pointer	[costly dependency pointers]	1.5849625007211563	3	1	0	0
936	73	model analyze English sentence	[model analyzing English sentences]	1.5849625007211563	4	1	0	0
937	73	part of the parser	[part of the parser]	1.5849625007211563	4	1	0	0
938	73	current natural language interface	[Current natural language interfaces]	1.5849625007211563	4	1	0	0
939	73	single Intel i860 chip	[single Intel i860 chip]	1.5849625007211563	4	1	0	0
940	73	part of a unit	[part of a unit]	1.5849625007211563	4	1	0	0
941	73	order of magnitude	[orders of magnitude]	1.5849625007211563	3	1	0	0
942	73	correct minor error	[correcting minor errors]	1.5849625007211563	3	1	0	0
943	73	description of feature structure	[description of feature structures]	1.5849625007211563	4	2	1	1
944	73	definite noun phrase reiteration	[definite noun phrase reiteration]	1.5849625007211563	4	1	0	0
945	73	simple mixture language model	[simple mixture language model]	1.5849625007211563	4	1	0	0
946	73	traditional statistical approach	[traditional statistical approaches]	1.5849625007211563	3	1	0	0
947	73	two different algorithm	[two different algorithms]	1.5849625007211563	3	1	0	0
948	73	Hidden Markov Model	[Hidden Markov Model]	1.5849625007211563	3	1	0	0
949	73	one unique author	[one unique author]	1.5849625007211563	3	1	0	0
950	73	basic of monolingual UCG	[basics of monolingual UCG]	1.5849625007211563	4	1	0	0
951	73	disjunctive normal form	[disjunctive normal form]	1.5849625007211563	3	1	0	0
952	73	one of this theory	[one of these theories]	1.5849625007211563	4	1	0	0
953	73	set of inter-related task	[set of inter-related tasks]	1.5849625007211563	4	1	0	0
954	73	structure-sharing of graph	[structure-sharing of graphs]	1.5849625007211563	3	1	0	0
955	73	new CSR corpus	[new CSR corpus]	1.5849625007211563	3	1	0	0
956	73	grammar checker for Basque	[grammar checker for Basque]	1.5849625007211563	4	1	0	0
957	73	Case Marker System	[Case Marker System]	1.5849625007211563	3	1	0	0
958	73	MIT ATIS system	[MIT ATIS system]	1.5849625007211563	3	1	0	0
959	73	experiment in this article	[experiments in this article]	1.5849625007211563	4	1	0	0
960	73	Horn logic program	[Horn logic program]	1.5849625007211563	3	1	0	0
961	73	method for error correction	[method for error correction]	1.5849625007211563	4	1	0	0
962	73	use a neural network	[using a neural network]	1.5849625007211563	4	1	0	0
963	73	2 .284 srh	[2 .284 SRHs]	1.5849625007211563	3	1	0	0
964	73	small Arabic corpus	[small Arabic corpus]	1.5849625007211563	3	1	0	0
965	73	notion of noun group	[notions of noun group]	1.5849625007211563	4	1	0	0
966	73	include Natural language processing	[including Natural Language Processing]	1.5849625007211563	4	1	0	0
967	73	offer advantage like reproducability	[Offering advantages like reproducability]	1.5849625007211563	4	1	0	0
968	73	grammar for this generator	[grammar for this generator]	1.5849625007211563	4	1	0	0
969	73	document understanding technology	[document understanding technology]	1.5849625007211563	3	1	0	0
970	73	performance of the system	[performance of the system]	1.5849625007211563	4	1	0	0
971	73	sequence of utterance	[sequence of utterances]	1.5849625007211563	3	1	0	0
972	73	32 .8% improvement	[32 .8% improvement]	1.5849625007211563	3	1	0	0
973	73	domain of navy message	[domain of Navy messages]	1.5849625007211563	4	1	0	0
974	73	operation of the mechanism	[operation of the mechanisms]	1.5849625007211563	4	1	0	0
975	73	several phrase-based translation model	[several phrase-based translation models]	1.5849625007211563	4	1	0	0
976	73	approximation of HPSG	[approximation of HPSG]	1.5849625007211563	3	1	0	0
977	73	difference between sentence structure	[difference between sentence structure]	1.5849625007211563	4	1	0	0
978	73	formalization of defeasible reasoning	[formalization of defeasible reasoning]	1.5849625007211563	4	1	0	0
979	73	reasonably neat form	[reasonably neat form]	1.5849625007211563	3	1	0	0
980	73	assumption in neat form	[assumption in neat form]	1.5849625007211563	4	1	0	0
981	73	large unsegmented Arabic corpus	[large unsegmented Arabic corpus]	1.5849625007211563	4	1	0	0
982	73	dynamic language model combination	[dynamic language model combination]	1.5849625007211563	4	1	0	0
983	73	carefully control the order	[carefully controlling the order]	1.5849625007211563	4	1	0	0
984	73	empirical conversation analysis	[empirical conversation analysis]	1.5849625007211563	3	1	0	0
985	73	degree of hastiness	[degree of hastiness]	1.5849625007211563	3	1	0	0
986	73	other mt system	[other MT systems]	1.5849625007211563	3	1	0	0
987	73	February '92 benchmark evaluation	[February '92 benchmark evaluation]	1.5849625007211563	4	1	0	0
988	73	use concept-based seed	[using concept-based seeds]	1.5849625007211563	3	1	0	0
989	73	importance of the issue	[importance of the issue]	1.5849625007211563	4	1	0	0
990	73	discourse understand accuracy	[discourse understanding accuracy]	1.5849625007211563	3	1	0	0
991	73	knowledge about task-oriented dialog	[knowledge about task-oriented dialogs]	1.5849625007211563	4	1	0	0
992	73	disjoint English phrase	[disjoint English phrases]	1.5849625007211563	3	1	0	0
993	73	manual evaluation of string	[manual evaluation of strings]	1.5849625007211563	4	1	0	0
994	73	behavior of the oracle	[behavior of the oracle]	1.5849625007211563	4	1	0	0
995	73	datum collection effort	[data collection effort]	1.5849625007211563	3	1	0	0
996	73	very simple sentence-plan-generator SPG	[very simple sentence-plan-generator SPG]	1.5849625007211563	4	1	0	0
997	73	underspecified chart representation	[underspecified chart representations]	1.5849625007211563	3	1	0	0
998	73	broad range of text	[broad range of texts]	1.5849625007211563	4	1	0	0
999	73	special method describe inflection	[special method describing inflection]	1.5849625007211563	4	1	0	0
1000	73	various relation between entity	[various relations between entities]	1.5849625007211563	4	1	0	0
1001	73	lack of automatic method	[lack of automatic methods]	1.5849625007211563	4	1	0	0
1002	73	performance of system	[performance of systems]	1.5849625007211563	3	1	0	0
1003	73	continuous speech recognition	[continuous speech recognition]	1.5849625007211563	3	2	2	2
1004	73	kind of similarity	[kind of similarity]	1.5849625007211563	3	2	2	2
1005	73	future research topic	[future research topics]	1.5849625007211563	3	1	0	0
1006	73	training reference speaker	[training reference speaker]	1.5849625007211563	3	1	0	0
1007	73	effectiveness of this index	[effectiveness of these indices]	1.5849625007211563	4	1	0	0
1008	73	bilingual parallel corpus	[bilingual parallel corpus]	1.5849625007211563	3	2	1	1
1009	73	only bilingual resource	[only bilingual resource]	1.5849625007211563	3	1	0	0
1010	73	word-semantic information email .g	[word-semantic information e .g]	1.5849625007211563	4	1	0	0
1011	73	DARPA resource management corpus	[DARPA Resource Management corpus]	1.5849625007211563	4	1	0	0
1012	73	example of the difference	[examples of the difference]	1.5849625007211563	4	1	0	0
1013	73	use part-of-speech tag sequence	[using part-of-speech tag sequences]	1.5849625007211563	4	1	0	0
1014	73	TREC 2003 qa track	[TREC 2003 QA tracks]	1.5849625007211563	4	1	0	0
1015	73	context-dependent phonetic modelling	[context-dependent phonetic modelling]	1.5849625007211563	3	1	0	0
1016	73	focus of attention	[focus of attention]	1.5849625007211563	3	2	1	1
1017	73	hope generate the parse	[hope generating the parse]	1.5849625007211563	4	1	0	0
1018	73	last two decade	[last two decades]	1.5849625007211563	3	1	0	0
1019	73	method resolve this ambiguity	[method resolving this ambiguity]	1.5849625007211563	4	1	0	0
1020	73	natural language generation	[natural language generation]	1.5849625007211563	3	2	1	1
1021	73	theory of attention	[theory of attention]	1.5849625007211563	3	1	0	0
1022	73	robust parsing capability	[robust parsing capability]	1.5849625007211563	3	1	0	0
1023	73	category Cooccurrence Restriction CCRs	[Category Cooccurrence Restrictions CCRs]	1.5849625007211563	4	1	0	0
1024	73	support to the use	[support to the use]	1.5849625007211563	4	1	0	0
1025	73	disambiguate the noun	[disambiguating the nouns]	1.5849625007211563	3	1	0	0
1026	73	prior to content determination	[prior to content determination]	1.5849625007211563	4	1	0	0
1027	73	completely missing fragment	[completely missing fragments]	1.5849625007211563	3	1	0	0
1028	73	standard evaluation metric	[standard evaluation metric]	1.5849625007211563	3	1	0	0
1029	73	CSR speech Recognition corpus	[CSR Speech Recognition corpus]	1.5849625007211563	4	1	0	0
1030	73	demonstrative expression in English	[demonstrative expressions in English]	1.5849625007211563	4	1	0	0
1031	73	Japanese sentence structure	[Japanese sentence structure]	1.5849625007211563	3	1	0	0
1032	73	distinction among this component	[distinction among these components]	1.5849625007211563	4	1	0	0
1033	73	use of sophisticated representation	[use of sophisticated representations]	1.5849625007211563	4	1	0	0
1034	73	language processing module	[language processing modules]	1.5849625007211563	3	1	0	0
1035	73	recognition with this model	[recognition with this model]	1.5849625007211563	4	1	0	0
1036	73	young man like woman	[young men like women]	1.5849625007211563	4	2	2	2
1037	73	consistency problem for formulas	[consistency problem for formulas]	1.5849625007211563	4	1	0	0
1038	73	relationship between language	[relationships between language]	1.5849625007211563	3	1	0	0
1039	73	instance statistical mt system	[instance statistical MT systems]	1.5849625007211563	4	1	0	0
1040	73	novel approach to q/a	[novel approach to Q/A]	1.5849625007211563	4	1	0	0
1041	73	top human-ranked sentence plan	[top human-ranked sentence plan]	1.5849625007211563	4	1	0	0
1042	73	prior RCL parser	[prior RCL parser]	1.5849625007211563	3	1	0	0
1043	73	context of a task	[context of a task]	1.5849625007211563	4	1	0	0
1044	73	introduce this approach	[introducing this approach]	1.5849625007211563	3	1	0	0
1045	73	useful disambiguation tool	[useful disambiguation tool]	1.5849625007211563	3	1	0	0
1046	73	number of comparative evaluation	[number of comparative evaluations]	1.5849625007211563	4	1	0	0
1047	73	Arabic' rich morphology	[Arabic's rich morphology]	1.5849625007211563	3	1	0	0
1048	73	baseline model’ score	[baseline model’s score]	1.5849625007211563	3	1	0	0
1049	73	unsupervised word alignment component	[unsupervised word alignment component]	1.5849625007211563	4	1	0	0
1050	73	placement of function word	[placement of function words]	1.5849625007211563	4	1	0	0
1051	73	error for the purpose	[error for the purposes]	1.5849625007211563	4	1	0	0
1052	73	natural language parsing nlp	[natural language parsing NLP]	1.5849625007211563	4	1	0	0
1053	73	error in machine translation	[errors in machine translation]	1.5849625007211563	4	1	0	0
1054	73	management of uncertainty	[management of uncertainty]	1.5849625007211563	3	1	0	0
1055	73	18 .3m Chinese character	[18 .3M Chinese characters]	1.5849625007211563	4	1	0	0
1056	73	method detect speech repair	[method detecting speech repairs]	1.5849625007211563	4	1	0	0
1057	73	machine learning technique	[machine learning techniques]	1.5849625007211563	3	2	2	2
1058	73	long possible span	[longest possible span]	1.5849625007211563	3	1	0	0
1059	73	IBM model 1-2	[IBM Models 1-2]	1.5849625007211563	3	1	0	0
1060	73	standard Arabic-English translation task	[standard Arabic-English translation task]	1.5849625007211563	4	1	0	0
1061	73	sentence error rate	[sentence error rates]	1.5849625007211563	3	1	0	0
1062	73	enlarge the separation margin	[enlarging the separation margin]	1.5849625007211563	4	1	0	0
1063	73	different language family	[different language families]	1.5849625007211563	3	1	0	0
1064	73	accord to individual need	[according to individual needs]	1.5849625007211563	4	1	0	0
1065	73	concept of a syllogism	[concept of a syllogism]	1.5849625007211563	4	1	0	0
1066	73	new flexible inference method	[new flexible inference method]	1.5849625007211563	4	1	0	0
1067	73	arc between two node	[arcs between two nodes]	1.5849625007211563	4	1	0	0
1068	73	set of subpredicate	[sets of subpredicates]	1.5849625007211563	3	1	0	0
1069	73	look up phrase translation	[looking up phrase translations]	1.5849625007211563	4	1	0	0
1070	73	target recognition task	[target recognition task]	1.5849625007211563	3	1	0	0
1071	73	new interactive disambiguation scheme	[new interactive disambiguation scheme]	1.5849625007211563	4	1	0	0
1072	73	construction of statistical model	[construction of statistical models]	1.5849625007211563	4	1	0	0
1073	73	accurate word order generation	[accurate word order generation]	1.5849625007211563	4	1	0	0
1074	73	yet more redundancy	[yet more redundancy]	1.5849625007211563	3	1	0	0
1075	73	polynomial time decoding algorithm	[polynomial time decoding algorithm]	1.5849625007211563	4	1	0	0
1076	73	generation of refer expression	[Generation of Referring Expressions]	1.5849625007211563	4	1	0	0
1077	73	generate the correct parse	[generating the correct parse]	1.5849625007211563	4	1	0	0
1078	73	use of demonstrative expression	[use of demonstrative expressions]	1.5849625007211563	4	1	0	0
1079	73	avoid failure of reference	[avoiding failures of reference]	1.5849625007211563	4	1	0	0
1080	73	internal lexical variation	[internal lexical variation]	1.5849625007211563	3	1	0	0
1081	73	submanifold of datum	[submanifold of data]	1.5849625007211563	3	1	0	0
1082	73	various interesting ambiguity	[various interesting ambiguities]	1.5849625007211563	3	1	0	0
1083	73	most importantly Chinese-English text	[Most importantly Chinese-English texts]	1.5849625007211563	4	1	0	0
1084	73	connect referent with predicate	[connecting referents with predicates]	1.5849625007211563	4	1	0	0
1085	73	term of retrieval accuracy	[terms of retrieval accuracy]	1.5849625007211563	4	1	0	0
1086	73	high quality Ie result	[high quality IE results]	1.5849625007211563	4	1	0	0
1087	73	corpus of word	[corpus of words]	1.5849625007211563	3	2	2	2
1088	73	example of parse	[examples of parses]	1.5849625007211563	3	1	0	0
1089	73	input Arabic document	[input Arabic documents]	1.5849625007211563	3	1	0	0
1090	73	dictionary-based word vector	[dictionary-based word vectors]	1.5849625007211563	3	1	0	0
1091	73	selection of font	[selection of fonts]	1.5849625007211563	3	1	0	0
1092	73	to reasoning with disposition	[to reasoning with dispositions]	1.5849625007211563	4	1	0	0
1093	73	accurate predicate-argument structure	[accurate predicate-argument structures]	1.5849625007211563	3	1	0	0
1094	73	potential of text browser	[potential of text browsers]	1.5849625007211563	4	1	0	0
1095	73	current state of research	[current state of research]	1.5849625007211563	4	1	0	0
1096	73	target word selection	[target word selection]	1.5849625007211563	3	1	0	0
1097	73	vagueness in semantic annotation	[vagueness in semantic annotation]	1.5849625007211563	4	1	0	0
1098	73	heuristic Parsing Model HPM	[Heuristic Parsing Model HPM]	1.5849625007211563	4	1	0	0
1099	73	performance for sophisticated representation	[performance for sophisticated representations]	1.5849625007211563	4	1	0	0
1100	73	syllogistic reasoning with disposition	[Syllogistic reasoning with dispositions]	1.5849625007211563	4	1	0	0
1101	73	case of multi-document summarization	[case of multi-document summarization]	1.5849625007211563	4	1	0	0
1102	73	verbal grounding act	[verbal grounding acts]	1.5849625007211563	3	1	0	0
1103	73	sentence utilize example effectively	[sentence utilizing examples effectively]	1.5849625007211563	4	1	0	0
1104	73	explanation for the behaviour	[explanations for the behaviour]	1.5849625007211563	4	1	0	0
1105	73	single understanding result	[single understanding result]	1.5849625007211563	3	1	0	0
1106	73	use a trigram model	[using a trigram model]	1.5849625007211563	4	1	0	0
1107	73	path between the node	[path between the nodes]	1.5849625007211563	4	1	0	0
1108	73	treelet translation pair	[treelet translation pairs]	1.5849625007211563	3	1	0	0
1109	73	movement of word block	[movements of word blocks]	1.5849625007211563	4	1	0	0
1110	73	parsimonious instantiation scheme	[parsimonious instantiation scheme]	1.5849625007211563	3	1	0	0
1111	73	operation of the system	[operation of the system]	1.5849625007211563	4	1	0	0
1112	73	new evaluation measure	[new evaluation measure]	1.5849625007211563	3	1	0	0
1113	73	train two successive learner	[training two successive learners]	1.5849625007211563	4	1	0	0
1114	73	new target speaker	[new target speaker]	1.5849625007211563	3	1	0	0
1115	73	computational complexity of unification	[computational complexity of unification]	1.5849625007211563	4	1	0	0
1116	73	use real dialogue datum	[using real dialogue data]	1.5849625007211563	4	1	0	0
1117	73	reasonable classification accuracy	[reasonable classification accuracy]	1.5849625007211563	3	1	0	0
1118	73	document over a representation	[documents over a representation]	1.5849625007211563	4	1	0	0
1119	73	attentional state dynamic	[attentional state dynamic]	1.5849625007211563	3	1	0	0
1120	73	build a SMT system	[building an SMT system]	1.5849625007211563	4	1	0	0
1121	73	various rhetorical function	[various rhetorical functions]	1.5849625007211563	3	1	0	0
1122	73	resource sensitive logic	[resource sensitive logic]	1.5849625007211563	3	1	0	0
1123	73	improvement of word	[improvement of word]	1.5849625007211563	3	1	0	0
1124	73	October dry-run test set	[October dry-run test set]	1.5849625007211563	4	1	0	0
1125	73	head-to-head test against one	[head-to-head tests against one]	1.5849625007211563	4	1	0	0
1126	73	last few year	[last few years]	1.5849625007211563	3	1	0	0
1127	73	near neighbour method	[nearest neighbour methods]	1.5849625007211563	3	1	0	0
1128	73	solution to this problem	[solution to these problems]	1.5849625007211563	4	1	0	0
1129	73	histogram of keyword	[histogram of keywords]	1.5849625007211563	3	1	0	0
1130	73	missing word ellipsis	[missing words ellipsis]	1.5849625007211563	3	1	0	0
1131	73	description of a approach	[description of an approach]	1.5849625007211563	4	1	0	0
1132	73	performance of the algorithm	[Performance of the algorithm]	1.5849625007211563	4	1	0	0
1133	73	strength of approach	[strength of approach]	1.5849625007211563	3	1	0	0
1134	73	method produce such phrase	[method producing such phrases]	1.5849625007211563	4	1	0	0
1135	73	application of this approach	[application of this approach]	1.5849625007211563	4	1	0	0
1136	73	design of unification algorithm	[design of unification algorithms]	1.5849625007211563	4	1	0	0
1137	73	work translation system	[working translation system]	1.5849625007211563	3	1	0	0
1138	73	machine translation pipeline	[machine translation pipeline]	1.5849625007211563	3	1	0	0
1139	73	complex natural language processing	[complex natural language processing]	1.5849625007211563	4	1	0	0
1140	73	category in local tree	[categories in local trees]	1.5849625007211563	4	1	0	0
1141	73	much recent debate	[much recent debate]	1.5849625007211563	3	1	0	0
1142	73	use translation probability	[using translation probabilities]	1.5849625007211563	3	1	0	0
1143	73	field of speech processing	[field of speech processing]	1.5849625007211563	4	1	0	0
1144	73	use of lexical feature	[use of lexical features]	1.5849625007211563	4	2	2	2
1145	73	block with orientation	[blocks with orientation]	1.5849625007211563	3	1	0	0
1146	73	accurate structural disambiguation	[accurate structural disambiguation]	1.5849625007211563	3	1	0	0
1147	73	DARPA speech recognition technology	[DARPA speech recognition technology]	1.5849625007211563	4	2	1	1
1148	73	explanation of a ambiguity	[explanation of an ambiguity]	1.5849625007211563	4	1	0	0
1149	73	in-domain monolingual corpus	[in-domain monolingual corpus]	1.5849625007211563	3	1	0	0
1150	73	use 109 training speaker	[using 109 training speakers]	1.5849625007211563	4	1	0	0
1151	73	logon mt demonstrator	[LOGON MT demonstrator]	1.5849625007211563	3	1	0	0
1152	73	lack of negative feedback	[lack of negative feedback]	1.5849625007211563	4	1	0	0
1153	73	publication of paper	[publication of papers]	1.5849625007211563	3	1	0	0
1154	73	term of Dynamic grammar	[terms of Dynamic Grammars]	1.5849625007211563	4	1	0	0
1155	73	operate on parse forest	[operating on parse forests]	1.5849625007211563	4	1	0	0
1156	73	validity of lexical entry	[validity of lexical entries]	1.5849625007211563	4	1	0	0
1157	73	scale of exist spelling-checker	[scale of existing spelling-checkers]	1.5849625007211563	4	1	0	0
1158	73	duplicate the experiment	[duplicating the experiment]	1.5849625007211563	3	1	0	0
1159	73	theory of language presupposition	[theory of language presuppositions]	1.5849625007211563	4	1	0	0
1160	73	fundamental problem of SMT	[fundamental problems of SMT]	1.5849625007211563	4	1	0	0
1161	73	novel evaluation scheme	[novel evaluation scheme]	1.5849625007211563	3	1	0	0
1162	73	honorifics respectful word	[honorifics respectful words]	1.5849625007211563	3	1	0	0
1163	73	same source text	[same source text]	1.5849625007211563	3	1	0	0
1164	73	relatively free word order	[relatively free word order]	1.5849625007211563	4	1	0	0
1165	73	however the only area	[However the only area]	1.5849625007211563	4	1	0	0
1166	73	framework describe the processing	[framework describing the processing]	1.5849625007211563	4	1	0	0
1167	73	general preference of approach	[general preference of approach]	1.5849625007211563	4	1	0	0
1168	73	disjunctive feature structure	[disjunctive feature structures]	1.5849625007211563	3	1	0	0
1169	73	use complicated large knowledgebase	[using complicated large knowledgebases]	1.5849625007211563	4	1	0	0
1170	73	clustering of word co-occurrence	[clustering of word co-occurrences]	1.5849625007211563	4	1	0	0
1171	73	word-sense disambiguation algorithm	[word-sense disambiguation algorithm]	1.5849625007211563	3	1	0	0
1172	73	instead of word	[instead of words]	1.5849625007211563	3	1	0	0
1173	73	method acquire sense-tagged datum	[method acquiring sense-tagged data]	1.5849625007211563	4	1	0	0
1174	73	system response to user	[system response to users]	1.5849625007211563	4	1	0	0
1175	73	temporal relation between segment	[temporal relations between segments]	1.5849625007211563	4	1	0	0
1176	73	English-Chinese parallel corpus	[English-Chinese parallel corpora]	1.5849625007211563	3	1	0	0
1177	73	form of n-gram	[form of N-grams]	1.5849625007211563	3	1	0	0
1178	73	unit on protocol analysis	[unit on protocol analysis]	1.5849625007211563	4	1	0	0
1179	73	certain kind of phrase	[certain kinds of phrases]	1.5849625007211563	4	1	0	0
1180	73	use the information bottleneck	[using the Information Bottleneck]	1.5849625007211563	4	1	0	0
1181	73	dependency structure of sentence	[dependency structure of sentences]	1.5849625007211563	4	1	0	0
1182	73	Prolog subset of logic	[Prolog subset of logic]	1.5849625007211563	4	1	0	0
1183	73	method achieve robustness	[methods achieving robustness]	1.5849625007211563	3	1	0	0
1184	73	informally a proposition	[Informally a proposition]	1.5849625007211563	3	1	0	0
1185	73	method for computational analysis	[method for computational analysis]	1.5849625007211563	4	1	0	0
1186	73	and/or answer level	[and/or answer levels]	1.5849625007211563	3	1	0	0
1187	73	CMU's SMT system	[CMU's SMT system]	1.5849625007211563	3	1	0	0
1188	73	list of word string	[list of word strings]	1.5849625007211563	4	1	0	0
1189	73	result of the evaluation	[results of the evaluations]	1.5849625007211563	4	1	0	0
1190	73	predictive performance of model	[predictive performance of models]	1.5849625007211563	4	1	0	0
1191	73	evaluate wsd program	[evaluating WSD programs]	1.5849625007211563	3	1	0	0
1192	73	use domain preference knowledge	[using domain preference knowledge]	1.5849625007211563	4	1	0	0
1193	73	approach to reasoning	[approach to reasoning]	1.5849625007211563	3	1	0	0
1194	73	hand-built symbolic resource	[hand-built symbolic resources]	1.5849625007211563	3	1	0	0
1195	73	segment of the discourse	[segments of the discourse]	1.5849625007211563	4	1	0	0
1196	73	bad coverage than account	[worse coverage than accounts]	1.5849625007211563	4	1	0	0
1197	73	process example in context	[processing examples in context]	1.5849625007211563	4	1	0	0
1198	73	local phrase re-ordering	[local phrase re-ordering]	1.5849625007211563	3	1	0	0
1199	73	precision on both system	[precision on both systems]	1.5849625007211563	4	1	0	0
1200	73	recognize the intention	[recognizing the intentions]	1.5849625007211563	3	1	0	0
1201	73	resolve zero pronoun	[resolving zero pronouns]	1.5849625007211563	3	1	0	0
1202	73	non-native language essay	[non-native language essays]	1.5849625007211563	3	1	0	0
1203	73	natural language parsing	[natural language parsing]	1.5849625007211563	3	2	2	2
1204	73	utility of this constraint	[utility of this constraint]	1.5849625007211563	4	1	0	0
1205	73	phrase with gap	[phrases with gaps]	1.5849625007211563	3	1	0	0
1206	73	shortcoming of different strategy	[shortcomings of different strategies]	1.5849625007211563	4	1	0	0
1207	73	subset of large database	[subsets of larger database]	1.5849625007211563	4	1	0	0
1208	73	language as a pivot	[language as a pivot]	1.5849625007211563	4	1	0	0
1209	73	statement of generalization	[statement of generalizations]	1.5849625007211563	3	1	0	0
1210	73	statistical translation model	[statistical translation model]	1.5849625007211563	3	1	0	0
1211	73	segmental k-means algorithm	[segmental k-means algorithm]	1.5849625007211563	3	1	0	0
1212	73	especially personal name	[especially personal names]	1.5849625007211563	3	1	0	0
1213	73	correct chain of syllable	[correct chain of syllables]	1.5849625007211563	4	1	0	0
1214	73	selection of bag-of-word	[selection of bag-of-words]	1.5849625007211563	3	1	0	0
1215	73	idea behind method	[idea behind method]	1.5849625007211563	3	1	0	0
1216	73	implementation of the model	[implementation of the model]	1.5849625007211563	4	1	0	0
1217	73	computational complexity of some	[computational complexity of some]	1.5849625007211563	4	1	0	0
1218	73	conversation for documentation	[conversation for documentation]	1.5849625007211563	3	1	0	0
1219	73	predicate calculus notation	[predicate calculus notation]	1.5849625007211563	3	1	0	0
1220	73	method of abductive inference	[method of abductive inference]	1.5849625007211563	4	1	0	0
1221	73	almost inevitable inference	[almost inevitable inferences]	1.5849625007211563	3	1	0	0
1222	73	yield different way	[yielding different ways]	1.5849625007211563	3	1	0	0
1223	73	approach to pronoun resolution	[approach to pronoun resolution]	1.5849625007211563	4	1	0	0
1224	73	problem extract specific information	[problem extracting specific information]	1.5849625007211563	4	1	0	0
1225	73	two baseline sentence planner	[two baseline sentence planners]	1.5849625007211563	4	1	0	0
1226	73	balance matching operation	[balance matching operation]	1.5849625007211563	3	1	0	0
1227	73	finite State Model FSM	[Finite State Model FSM]	1.5849625007211563	4	1	0	0
1228	73	trigram language model	[trigram language model]	1.5849625007211563	3	2	1	1
1229	73	valid semantic interpretation	[valid semantic interpretation]	1.5849625007211563	3	1	0	0
1230	73	sophisticated representation of document	[sophisticated representation of documents]	1.5849625007211563	4	2	1	1
1231	73	most difficult problem	[most difficult problems]	1.5849625007211563	3	1	0	0
1232	73	Chinese Natural language processing	[Chinese Natural Language Processing]	1.5849625007211563	4	1	0	0
1233	73	exist spelling-checker for English	[existing spelling-checkers for English]	1.5849625007211563	4	1	0	0
1234	73	two kind of similarity	[two kinds of similarity]	1.5849625007211563	4	1	0	0
1235	73	more sophisticated annotation	[more sophisticated annotation]	1.5849625007211563	3	1	0	0
1236	73	suffix array-based datum structure	[suffix array-based data structure]	1.5849625007211563	4	1	0	0
1237	73	many classical grammatical formalism	[many classical grammatical formalisms]	1.5849625007211563	4	1	0	0
1238	73	left corner parsing algorithm	[left corner parsing algorithm]	1.5849625007211563	4	1	0	0
1239	73	answer to semantic question	[answers to semantic questions]	1.5849625007211563	4	1	0	0
1240	73	framework of international collaboration	[framework of international collaborations]	1.5849625007211563	4	1	0	0
1241	73	result about the effectiveness	[results about the effectiveness]	1.5849625007211563	4	1	0	0
1242	73	temporal anaphora resolution	[temporal anaphora resolution]	1.5849625007211563	3	1	0	0
1243	73	same test set	[same test set]	1.5849625007211563	3	1	0	0
1244	73	verbal means for grounding	[verbal means for grounding]	1.5849625007211563	4	1	0	0
1245	74	order	[order]	1.5	1	16	3	3
1246	74	novel approach	[novel approach]	1.5	2	3	3	2
1247	74	multiple candidate	[multiple candidates]	1.5	2	3	3	2
1248	74	anaphora resolution	[anaphora resolution]	1.5	2	3	3	2
1249	74	phrase translation	[phrase translations]	1.5	2	3	3	2
1250	74	refer expression	[referring expressions]	1.5	2	3	3	2
1251	74	verbal interaction	[verbal interaction]	1.5	2	3	3	2
1252	75	utterance	[utterance]	1.4857142857142858	1	16	8	7
1253	76	knowledge	[knowledge]	1.4846153846153847	1	16	15	13
1254	77	improvement	[improvement]	1.476923076923077	1	16	16	13
1255	78	inference	[inference]	1.4666666666666668	1	16	8	6
1256	79	input	[input]	1.4000000000000001	1	15	8	8
1257	80	example	[example]	1.3833333333333335	1	15	7	6
1258	81	target	[target]	1.3800000000000001	1	15	6	5
1259	82	meaning	[meaning]	1.3	1	14	7	7
1260	82	technique	[technique]	1.3	1	14	8	8
1261	82	parser	[parser]	1.3	1	14	7	7
1262	83	dependency	[dependency]	1.2800000000000002	1	14	6	5
1263	84	output	[output]	1.2666666666666666	1	14	8	6
1264	85	time	[time]	1.2000000000000002	1	13	10	10
1265	85	speaker	[speaker]	1.2000000000000002	1	13	5	5
1266	86	reference	[reference]	1.1800000000000002	1	13	6	5
1267	87	way	[way]	1.1	1	12	11	11
1268	87	decision	[decision]	1.1	1	12	5	5
1269	87	generation	[generation]	1.1	1	12	8	8
1270	88	representation	[representation]	1.0857142857142859	1	12	8	7
1271	88	interpretation	[interpretation]	1.0857142857142859	1	12	8	7
1272	89	context	[context]	1.0666666666666667	1	12	8	6
1273	90	concept	[concept]	1.042857142857143	1	12	11	7
1274	91	tracking algorithm	[tracking algorithm]	1.0	2	1	0	0
1275	91	implementation in a voice system	[implementation in a voice system]	1.0	5	1	0	0
1276	91	minor extension of pre-exist component	[minor extensions of pre-existing components]	1.0	5	1	0	0
1277	91	bag-of-word method	[bag-of-words methods]	1.0	2	1	0	0
1278	91	experimental research	[experimental research]	1.0	2	1	0	0
1279	91	definition of typicality a concept	[definition of typicality a concept]	1.0	5	1	0	0
1280	91	head Grammar	[Head Grammars]	1.0	2	1	0	0
1281	91	m-component mixture of trigram model	[m-component mixture of trigram models]	1.0	5	1	0	0
1282	91	empirical viewpoint	[empirical viewpoint]	1.0	2	1	0	0
1283	91	semantic constraint in broad domain	[semantic constraints in broad domains]	1.0	5	1	0	0
1284	91	suppress the fuzzy quantifier most	[suppressing the fuzzy quantifier most]	1.0	5	1	0	0
1285	91	low precision	[low precision]	1.0	2	1	0	0
1286	91	complete interpretation	[complete interpretations]	1.0	2	1	0	0
1287	91	chart parsing	[Chart parsing]	1.0	2	2	1	1
1288	91	template-based generation	[template-based generation]	1.0	2	1	0	0
1289	91	many speaker prior to training	[many speakers prior to training]	1.0	5	1	0	0
1290	91	log( would	[log( d]	1.0	2	1	0	0
1291	91	roughly corresponding in a chart	[roughly corresponding in a chart]	1.0	5	1	0	0
1292	91	many case	[many cases]	1.0	2	1	0	0
1293	91	appropriate database	[appropriate database]	1.0	2	1	0	0
1294	91	practical approach employ n-gram model	[practical approach employing n-gram models]	1.0	5	1	0	0
1295	91	polarity-tagged corpus	[polarity-tagged corpus]	1.0	2	1	0	0
1296	91	eye gaze	[eye gaze]	1.0	2	1	0	0
1297	91	CFG filter	[CFG filtering]	1.0	2	2	2	2
1298	91	copy of the derivation tree	[copies of the derivation tree]	1.0	5	1	0	0
1299	91	extend the n-b speech/language integration architecture	[extending the n-best speech/language integration architecture]	1.0	6	1	0	0
1300	91	Naive Bayes	[Naive Bayes]	1.0	2	1	0	0
1301	91	early one	[earlier ones]	1.0	2	1	0	0
1302	91	Multimedia answer	[Multimedia answers]	1.0	2	1	0	0
1303	91	mathematical expression	[mathematical expressions]	1.0	2	1	0	0
1304	91	syntactic description	[syntactic descriptions]	1.0	2	1	0	0
1305	91	knowledge about the system user	[knowledge about the system users]	1.0	5	1	0	0
1306	91	additional index such as the time	[additional indices such as the time]	1.0	6	1	0	0
1307	91	experiment result	[Experiment results]	1.0	2	2	1	1
1308	91	spatial concept	[spatial concepts]	1.0	2	1	0	0
1309	91	other model	[other models]	1.0	2	1	0	0
1310	91	access to a pharmaceutical news archive	[access to a pharmaceutical news archive]	1.0	6	1	0	0
1311	91	dependency tree	[dependency trees]	1.0	2	1	0	0
1312	91	direct the process translate a sentence	[directing the process translating a sentence]	1.0	6	1	0	0
1313	91	use the unification-based grammar formalism	[using the unification-based grammar formalism]	1.0	5	1	0	0
1314	91	system for incremental grammar development	[system for incremental grammar development]	1.0	5	1	0	0
1315	91	other document	[other documents]	1.0	2	1	0	0
1316	91	similar criterion	[similar criteria]	1.0	2	1	0	0
1317	91	follow idea	[following ideas]	1.0	2	1	0	0
1318	91	basic goal build that editor	[basic goal building that editor]	1.0	5	1	0	0
1319	91	JAVELIN system	[JAVELIN system]	1.0	2	1	0	0
1320	91	information about the language learning process	[information about the language learning process]	1.0	6	1	0	0
1321	91	HK law	[HK laws]	1.0	2	1	0	0
1322	91	direct application of exist metric	[direct application of existing metrics]	1.0	5	1	0	0
1323	91	theory about this probably-correct rule	[theory about these probably-correct rules]	1.0	5	1	0	0
1324	91	dialogue state	[dialogue state]	1.0	2	1	0	0
1325	91	conventional algorithm for the analysis	[conventional algorithms for the analysis]	1.0	5	1	0	0
1326	91	.1% a 45% reduction in error	[.1% a 45% reduction in error]	1.0	6	1	0	0
1327	91	korean-to-english machine translation system CCLINC	[Korean-to-English machine translation system CCLINC]	1.0	5	1	0	0
1328	91	system' response	[system's response]	1.0	2	1	0	0
1329	91	rank task	[ranking tasks]	1.0	2	1	0	0
1330	91	success of the parsing mechanism	[success of the parsing mechanism]	1.0	5	1	0	0
1331	91	lexical weighting of phrase translation	[lexical weighting of phrase translations]	1.0	5	1	0	0
1332	91	incorporate a approximation of the analysis	[incorporating an approximation of the analysis]	1.0	6	1	0	0
1333	91	language definition	[language definition]	1.0	2	2	2	2
1334	91	evaluate qa system performance on question	[evaluating QA system performance on questions]	1.0	6	1	0	0
1335	91	semantic interpretation	[semantic interpretation]	1.0	2	2	2	2
1336	91	two node	[two nodes]	1.0	2	2	1	1
1337	91	such sentence	[such sentences]	1.0	2	1	0	0
1338	91	formalism of category Cooccurrence Restriction	[formalism of Category Cooccurrence Restrictions]	1.0	5	1	0	0
1339	91	topical report	[topical report]	1.0	2	1	0	0
1340	91	path value	[path values]	1.0	2	1	0	0
1341	91	average the statistic of model	[averaging the statistics of models]	1.0	5	1	0	0
1342	91	right translation of the word	[right translation of the words]	1.0	5	1	0	0
1343	91	increase the parsing accuracy rate	[increasing the parsing accuracy rate]	1.0	5	1	0	0
1344	91	pattern prefix*-stem-suffix*	[pattern prefix*-stem-suffix*]	1.0	2	1	0	0
1345	91	assume great	[assuming greater]	1.0	2	1	0	0
1346	91	use of machine translation system	[use of machine translation systems]	1.0	5	1	0	0
1347	91	reference of the pronoun in sentence	[references of the pronoun in sentences]	1.0	6	1	0	0
1348	91	linguistic concept	[linguistic concepts]	1.0	2	2	1	1
1349	91	new interactive corpus exploration tool	[new interactive corpus exploration tool]	1.0	5	1	0	0
1350	91	demonstrative expression	[demonstrative expressions]	1.0	2	1	0	0
1351	91	isomorphic derivation	[isomorphic derivations]	1.0	2	1	0	0
1352	91	human listener	[human listeners]	1.0	2	1	0	0
1353	91	performance of a search engine	[performance of a search engine]	1.0	5	1	0	0
1354	91	same discourse	[same discourse]	1.0	2	1	0	0
1355	91	fluent speech	[fluent speech]	1.0	2	1	0	0
1356	91	stochastic process	[stochastic processes]	1.0	2	1	0	0
1357	91	error for the language system	[error for the language system]	1.0	5	1	0	0
1358	91	final constituent	[final constituent]	1.0	2	1	0	0
1359	91	conjunction with structure finding method	[conjunction with structure finding methods]	1.0	5	1	0	0
1360	91	complex task	[complex task]	1.0	2	1	0	0
1361	91	HTML document	[HTML documents]	1.0	2	2	2	2
1362	91	dialogue move	[dialogue move]	1.0	2	1	0	0
1363	91	abstract move	[abstract moves]	1.0	2	1	0	0
1364	91	traditional word	[traditional word]	1.0	2	1	0	0
1365	91	set of machine learning experiment	[set of machine learning experiments]	1.0	5	1	0	0
1366	91	parallel datum	[parallel data]	1.0	2	1	0	0
1367	91	bootstrapping procedure	[bootstrapping procedure]	1.0	2	1	0	0
1368	91	test	[test]	1.0	1	11	5	5
1369	91	relative effect of segment order	[relative effects of segment order]	1.0	5	1	0	0
1370	91	certain sublanguage with text structure	[certain sublanguages with text structure]	1.0	5	1	0	0
1371	91	true text	[true text]	1.0	2	1	0	0
1372	91	Syntactic Role	[Syntactic Role]	1.0	2	1	0	0
1373	91	processing account	[processing accounts]	1.0	2	1	0	0
1374	91	non-terminal edge	[non-terminal edges]	1.0	2	1	0	0
1375	91	activity discuss	[activity discussing]	1.0	2	1	0	0
1376	91	multilingual summary	[multilingual summaries]	1.0	2	1	0	0
1377	91	independent event-based extractive mutli-document summarization approach	[independent event-based extractive mutli-document summarization approaches]	1.0	6	1	0	0
1378	91	surface representation of domain entity	[surface representation of domain entities]	1.0	5	1	0	0
1379	91	treatment for Japanese sentence analysis	[treatment for Japanese sentence analyses]	1.0	5	1	0	0
1380	91	over a third of the sentence	[over a third of the sentences]	1.0	6	1	0	0
1381	91	type location	[typing location]	1.0	2	1	0	0
1382	91	effective way improve system performance	[effective way improving system performance]	1.0	5	1	0	0
1383	91	prior work	[prior work]	1.0	2	1	0	0
1384	91	only way	[only way]	1.0	2	1	0	0
1385	91	extract specific information from unrestricted text	[extracting specific information from unrestricted texts]	1.0	6	1	0	0
1386	91	examination of the effect of feature	[Examination of the effect of features]	1.0	6	1	0	0
1387	91	such decoding	[such decoding]	1.0	2	1	0	0
1388	91	multiple token	[multiple tokens]	1.0	2	1	0	0
1389	91	accord to the average precision metric	[according to the average precision metric]	1.0	6	1	0	0
1390	91	principled technique for syntactic analysis	[principled techniques for syntactic analysis]	1.0	5	1	0	0
1391	91	experimental object	[experimental objects]	1.0	2	1	0	0
1392	91	then perform cluster number estimation	[then performing cluster number estimation]	1.0	5	1	0	0
1393	91	member of the semantic set	[member of the semantic set]	1.0	5	1	0	0
1394	91	task-based evaluation use Arabic information retrieval	[Task-based evaluation using Arabic information retrieval]	1.0	6	1	0	0
1395	91	ill-formed input	[ill-formed input]	1.0	2	1	0	0
1396	91	large vocabulary	[large vocabulary]	1.0	2	1	0	0
1397	91	interface browsing	[interface browsing]	1.0	2	1	0	0
1398	91	use the new SI corpus	[using the new SI corpus]	1.0	5	1	0	0
1399	91	cast doubt on this common assumption	[casting doubt on this common assumption]	1.0	6	1	0	0
1400	91	lexical feature	[lexical features]	1.0	2	2	1	1
1401	91	tagging strategy	[tagging strategies]	1.0	2	2	1	1
1402	91	such miscommunication	[such miscommunications]	1.0	2	1	0	0
1403	91	NIST automatic mt evaluation software	[NIST automatic MT evaluation software]	1.0	5	1	0	0
1404	91	free indexation	[free indexation]	1.0	2	2	1	1
1405	91	average precision over unstemmed text	[average precision over unstemmed text]	1.0	5	1	0	0
1406	91	several experiment corresponding to semantic domain	[Several experiments corresponding to semantic domains]	1.0	6	1	0	0
1407	91	7 .5% word error rate	[7 .5% word error rate]	1.0	5	1	0	0
1408	91	new model	[new model]	1.0	2	1	0	0
1409	91	English alphabet	[English alphabets]	1.0	2	1	0	0
1410	91	generation module	[generation modules]	1.0	2	1	0	0
1411	91	goal naturally sound dialog system	[goal naturally sounding dialog systems]	1.0	5	1	0	0
1412	91	use the BU recognition system	[Using the BU recognition system]	1.0	5	1	0	0
1413	91	same concept	[same concepts]	1.0	2	1	0	0
1414	91	large extent	[large extent]	1.0	2	1	0	0
1415	91	important role in human cognition	[important role in human cognition]	1.0	5	1	0	0
1416	91	anaphoric expression	[anaphoric expressions]	1.0	2	1	0	0
1417	91	parallel sentence in comparable non-parallel corpus	[parallel sentences in comparable non-parallel corpora]	1.0	6	1	0	0
1418	91	text-plan input	[text-plan input]	1.0	2	1	0	0
1419	91	unstemmed text	[unstemmed text]	1.0	2	1	0	0
1420	91	unification framework	[unification framework]	1.0	2	1	0	0
1421	91	automatic detection	[automatic detection]	1.0	2	1	0	0
1422	91	automatic translation of natural language	[automatic translation of natural language]	1.0	5	1	0	0
1423	91	recognition task	[recognition task]	1.0	2	1	0	0
1424	91	most benefit from the integration	[most benefit from the integration]	1.0	5	1	0	0
1425	91	provably optimal free indexation algorithm	[provably optimal free indexation algorithm]	1.0	5	1	0	0
1426	91	model parameter with the vocabulary	[model parameters with the vocabulary]	1.0	5	1	0	0
1427	91	transform a disposition into a proposition	[transforming a disposition into a proposition]	1.0	6	1	0	0
1428	91	grammaticality of the system output	[grammaticality of the system output]	1.0	5	1	0	0
1429	91	implement a user modeling component	[implementing a user modeling component]	1.0	5	1	0	0
1430	91	dialogue duration for skilled user	[dialogue duration for skilled users]	1.0	5	1	0	0
1431	91	combine several language model LMs	[combining several language models LMs]	1.0	5	1	0	0
1432	91	independently valuable general-purpose nlp component	[independently valuable general-purpose NLP components]	1.0	5	1	0	0
1433	91	automatic alignment	[automatic alignments]	1.0	2	1	0	0
1434	91	event map	[event map]	1.0	2	1	0	0
1435	91	meaning structure on the sentence	[meaning structure on the sentence]	1.0	5	1	0	0
1436	91	word-based alignment	[word-based alignments]	1.0	2	1	0	0
1437	91	base parser	[base parser]	1.0	2	1	0	0
1438	91	dialogue corpus	[dialogue corpora]	1.0	2	1	0	0
1439	91	linguistic annotation	[linguistic annotation]	1.0	2	1	0	0
1440	91	knowledge level on the target domain	[knowledge level on the target domain]	1.0	6	1	0	0
1441	91	entry from machine readable dictionary	[entries from machine readable dictionaries]	1.0	5	1	0	0
1442	91	model TDMT	[model TDMT]	1.0	2	2	2	2
1443	91	affix removal	[affix removal]	1.0	2	1	0	0
1444	91	model parameter	[model parameters]	1.0	2	2	2	2
1445	91	rule-based system	[rule-based systems]	1.0	2	1	0	0
1446	91	system support of mission success	[system support of mission success]	1.0	5	1	0	0
1447	91	output of black-box ocr system	[output of black-box OCR systems]	1.0	5	1	0	0
1448	91	proprietary stemmer	[proprietary stemmer]	1.0	2	2	2	2
1449	91	scenario template	[scenario templates]	1.0	2	1	0	0
1450	91	two corpus	[two corpora]	1.0	2	1	0	0
1451	91	scruffy text	[scruffy texts]	1.0	2	2	1	1
1452	91	high performance in real application	[high performance in real applications]	1.0	5	1	0	0
1453	91	grammar model	[grammar model]	1.0	2	2	2	2
1454	91	predominant sense	[predominant senses]	1.0	2	1	0	0
1455	91	correct structure	[correct structure]	1.0	2	2	1	1
1456	91	multi-strategy approach	[multi-strategy approach]	1.0	2	1	0	0
1457	91	other thing	[other things]	1.0	2	1	0	0
1458	91	phrase-based model	[phrase-based models]	1.0	2	2	1	1
1459	91	use a fake non-terminal symbol	[using a fake non-terminal symbol]	1.0	5	1	0	0
1460	91	digital learning	[digital learning]	1.0	2	1	0	0
1461	91	test set	[test set]	1.0	2	2	1	1
1462	91	simple unsupervised technique learn morphology	[simple unsupervised technique learning morphology]	1.0	5	1	0	0
1463	91	cooperative Model	[Cooperative Model]	1.0	2	2	1	1
1464	91	predictive performance	[predictive performance]	1.0	2	2	1	1
1465	91	promising approach	[promising approach]	1.0	2	2	1	1
1466	91	human understanding of natural language	[human understanding of natural language]	1.0	5	1	0	0
1467	91	diagnostic tool	[diagnostic tool]	1.0	2	1	0	0
1468	91	young woman	[young women]	1.0	2	2	2	2
1469	91	use alignment technique from machine translation	[Using alignment techniques from machine translation]	1.0	6	1	0	0
1470	91	experimental evaluation	[experimental evaluation]	1.0	2	2	1	1
1471	91	build translation system for language pair	[building translation systems for language pairs]	1.0	6	1	0	0
1472	91	baseline system	[baseline system]	1.0	2	2	1	1
1473	91	denotational semantics	[denotational semantics]	1.0	2	1	0	0
1474	91	fundamental problem	[fundamental problem]	1.0	2	2	1	1
1475	91	commercial application	[commercial applications]	1.0	2	1	0	0
1476	91	parse record of the utterance	[parse record of the utterance]	1.0	5	1	0	0
1477	91	computation involve	[computations involving]	1.0	2	1	0	0
1478	91	mt evaluation	[MT evaluation]	1.0	2	2	1	1
1479	91	special attention	[Special attention]	1.0	2	1	0	0
1480	91	use of the discourse constraint	[use of the discourse constraint]	1.0	5	1	0	0
1481	91	use a underlying word alignment	[using an underlying word alignment]	1.0	5	1	0	0
1482	91	alternative marker	[alternative markers]	1.0	2	1	0	0
1483	91	large-scale acquisition	[large-scale acquisition]	1.0	2	1	0	0
1484	91	sentence ambiguity	[Sentence ambiguities]	1.0	2	1	0	0
1485	91	large volume	[larger volume]	1.0	2	1	0	0
1486	91	miss word	[missing words]	1.0	2	1	0	0
1487	91	other's proposal	[other's proposals]	1.0	2	1	0	0
1488	91	linguistic theory	[linguistic theory]	1.0	2	2	1	1
1489	91	initial rank of this parse	[initial ranking of these parses]	1.0	5	1	0	0
1490	91	document understanding	[document understanding]	1.0	2	2	1	1
1491	91	Arabic-English translation	[Arabic-English translation]	1.0	2	1	0	0
1492	91	relative performance	[relative performances]	1.0	2	1	0	0
1493	91	individual user	[individual users]	1.0	2	1	0	0
1494	91	alignment algorithm	[alignment algorithm]	1.0	2	1	0	0
1495	91	sample output a expert translation	[sample output an expert translation]	1.0	5	1	0	0
1496	91	state-of-the-art word sense disambiguation model	[state-of-the-art word sense disambiguation model]	1.0	5	1	0	0
1497	91	variant of a calculus notation	[variant of a calculus notation]	1.0	5	1	0	0
1498	91	disambiguation WSD	[disambiguation WSD]	1.0	2	1	0	0
1499	91	information extraction	[information extraction]	1.0	2	2	1	1
1500	91	use a out-of-domain bilingual corpus	[using an out-of-domain bilingual corpus]	1.0	5	1	0	0
1501	91	other area of natural language processing	[other areas of natural language processing]	1.0	6	1	0	0
1502	91	form a utterance about a task	[forming an utterance about a task]	1.0	6	1	0	0
1503	91	predictive question	[predictive questioning]	1.0	2	1	0	0
1504	91	baseline model	[baseline model]	1.0	2	1	0	0
1505	91	several experiment	[Several experiments]	1.0	2	2	1	1
1506	91	concept of a char t	[concept of a char t]	1.0	5	1	0	0
1507	91	experiment show the wsd accuracy low	[experiments showing the WSD accuracy lower]	1.0	6	1	0	0
1508	91	other resource	[other resources]	1.0	2	1	0	0
1509	91	WSD model	[WSD models]	1.0	2	2	1	1
1510	91	important bear on commonsense reasoning	[important bearing on commonsense reasoning]	1.0	5	1	0	0
1511	91	different reason	[different reasons]	1.0	2	1	0	0
1512	91	large number	[large number]	1.0	2	2	2	2
1513	91	126 sentence	[126 sentences]	1.0	2	1	0	0
1514	91	few speaker	[few speakers]	1.0	2	1	0	0
1515	91	application of ambiguity packing technique	[application of ambiguity packing techniques]	1.0	5	1	0	0
1516	91	pose special problem for reader	[posing special problems for readers]	1.0	5	1	0	0
1517	91	evidence scalable	[evidence scalable]	1.0	2	1	0	0
1518	91	ungrammatical input	[ungrammatical input]	1.0	2	1	0	0
1519	91	prediction model	[prediction model]	1.0	2	1	0	0
1520	91	document representation	[document representation]	1.0	2	1	0	0
1521	91	document summarization	[document summarization]	1.0	2	1	0	0
1522	91	recognition network	[recognition network]	1.0	2	1	0	0
1523	91	apply general-purpose language understand technology	[applying general-purpose language understanding technology]	1.0	5	1	0	0
1524	91	semantic frame	[semantic frame]	1.0	2	2	1	1
1525	91	referential information	[referential information]	1.0	2	1	0	0
1526	91	find more datum relevant to task	[finding more data relevant to task]	1.0	6	1	0	0
1527	91	path of arc between node	[path of arcs between nodes]	1.0	5	1	0	0
1528	91	approximation of the formal analysis	[approximation of the formal analysis]	1.0	5	1	0	0
1529	91	automatic word sense induction wsi	[automatic word sense induction WSI]	1.0	5	1	0	0
1530	91	inter-related task	[inter-related tasks]	1.0	2	1	0	0
1531	91	multimodal Communication	[Multimodal Communication]	1.0	2	1	0	0
1532	91	original question	[original question]	1.0	2	1	0	0
1533	91	new approach	[new approach]	1.0	2	2	1	1
1534	91	such movement	[such movements]	1.0	2	1	0	0
1535	91	datum collection	[data collection]	1.0	2	2	2	2
1536	91	accuracy rate	[accuracy rate]	1.0	2	2	1	1
1537	91	result algorithm	[resulting algorithm]	1.0	2	1	0	0
1538	91	use class-dependent interpolation of n-gram	[using class-dependent interpolation of N-grams]	1.0	5	1	0	0
1539	91	face-to-face meeting use Bayesian Network classifier	[face-to-face meetings using Bayesian Network classifiers]	1.0	6	1	0	0
1540	91	conversation transcript	[conversation transcripts]	1.0	2	1	0	0
1541	91	speech repair	[speech repairs]	1.0	2	2	2	2
1542	91	non-parallel corpus	[non-parallel corpus]	1.0	2	2	2	2
1543	91	non-constituent coordination	[non-constituent coordination]	1.0	2	1	0	0
1544	91	generation system	[generation system]	1.0	2	2	1	1
1545	91	distinguishing feature	[distinguishing features]	1.0	2	1	0	0
1546	91	present natural language search engine	[present natural language search engines]	1.0	5	1	0	0
1547	91	Human-Machine Communication include language processing	[Human-Machine Communication including Language Processing]	1.0	5	1	0	0
1548	91	phone-string classifier	[phone-string classifier]	1.0	2	1	0	0
1549	91	human text	[human text]	1.0	2	1	0	0
1550	91	raw text	[raw text]	1.0	2	1	0	0
1551	91	fulfillment of logical well-formedness condition	[fulfillment of logical well-formedness conditions]	1.0	5	1	0	0
1552	91	stochastic output	[stochastic output]	1.0	2	1	0	0
1553	91	LRE project	[LRE project]	1.0	2	1	0	0
1554	91	technique superior	[techniques superior]	1.0	2	1	0	0
1555	91	pronoun seed	[pronoun seeds]	1.0	2	1	0	0
1556	91	tool kit	[tool kit]	1.0	2	1	0	0
1557	91	edge adjacent	[edges adjacent]	1.0	2	1	0	0
1558	91	final structure	[final structure]	1.0	2	1	0	0
1559	91	retrieval time	[retrieval time]	1.0	2	2	1	1
1560	91	rough draft	[rough drafts]	1.0	2	1	0	0
1561	91	syntactic analysis	[syntactic analysis]	1.0	2	2	2	2
1562	91	easy exploitation of limited domain semantics	[easy exploitation of limited domain semantics]	1.0	6	1	0	0
1563	91	describe the layout of a apartment	[describing the layout of an apartment]	1.0	6	1	0	0
1564	91	word with multiple meaning ambiguity	[words with multiple meanings ambiguity]	1.0	5	1	0	0
1565	91	vital requirement of language translation	[vital requirements of language translation]	1.0	5	1	0	0
1566	91	Natural language	[Natural Language]	1.0	2	2	1	1
1567	91	focus especially on the change	[focusing especially on the changes]	1.0	5	1	0	0
1568	91	high accuracy	[high accuracy]	1.0	2	1	0	0
1569	91	feature generation	[feature generation]	1.0	2	1	0	0
1570	91	technical abstracting	[technical abstracting]	1.0	2	1	0	0
1571	91	similarity measure	[similarity measure]	1.0	2	2	2	2
1572	91	major new problem include vocabulary	[major new problems including vocabulary]	1.0	5	1	0	0
1573	91	retrieval accuracy	[retrieval accuracy]	1.0	2	2	1	1
1574	91	non-contiguous phrase	[non-contiguous phrases]	1.0	2	1	0	0
1575	91	segment contiguity	[segment contiguity]	1.0	2	2	1	1
1576	91	SYSCONJ system	[SYSCONJ system]	1.0	2	1	0	0
1577	91	actual application	[actual applications]	1.0	2	1	0	0
1578	91	input sentence	[input sentence]	1.0	2	1	0	0
1579	91	logistics system	[logistics system]	1.0	2	1	0	0
1580	91	observation complete	[observation complete]	1.0	2	1	0	0
1581	91	educational tool	[educational tool]	1.0	2	1	0	0
1582	91	multi-document summarization	[multi-document summarization]	1.0	2	1	0	0
1583	91	contain 28 ,449 word token	[containing 28 ,449 word tokens]	1.0	5	1	0	0
1584	91	analyze how	[analyzing how]	1.0	2	1	0	0
1585	91	computational problem	[computational problems]	1.0	2	1	0	0
1586	91	discourse-relevant purpose	[discourse-relevant purposes]	1.0	2	1	0	0
1587	91	new paradigm for human interaction	[new paradigm for human interaction]	1.0	5	1	0	0
1588	91	skilled user	[skilled users]	1.0	2	1	0	0
1589	91	gather information related to a scenario	[gathering information related to a scenario]	1.0	6	1	0	0
1590	91	dependency grammar	[dependency grammars]	1.0	2	1	0	0
1591	91	boosting approach	[boosting approach]	1.0	2	1	0	0
1592	91	combination with conventional SMT model	[combination with conventional SMT models]	1.0	5	1	0	0
1593	91	part-of-speech tagger	[part-of-speech tagger]	1.0	2	2	1	1
1594	91	experiment use the SemCor datum set	[Experiments using the SemCor data sets]	1.0	6	1	0	0
1595	91	regard to the evaluation methodology	[regard to the evaluation methodology]	1.0	5	1	0	0
1596	91	correlation of the new measure	[correlation of the new measure]	1.0	5	1	0	0
1597	91	good way	[best way]	1.0	2	1	0	0
1598	91	natural language in natural setting	[natural language in natural settings]	1.0	5	1	0	0
1599	91	cohesive text	[cohesive text]	1.0	2	1	0	0
1600	91	efficiency—to work	[efficiency—to work]	1.0	2	1	0	0
1601	91	Korean Phonology structure Grammar KPSG	[Korean Phonology Structure Grammar KPSG]	1.0	5	1	0	0
1602	91	valid dictionary on the basis	[valid dictionary on the basis]	1.0	5	1	0	0
1603	91	global behavior of a grammar	[global behavior of a grammar]	1.0	5	1	0	0
1604	91	contrast to the usual approach	[contrast to the usual approach]	1.0	5	1	0	0
1605	91	lexicon with 90 ,000 entry	[lexicon with 90 ,000 entries]	1.0	5	1	0	0
1606	91	learning methodology applicable in general domain	[learning methodology applicable in general domains]	1.0	6	1	0	0
1607	91	automatic generation of hypertext link	[automatic generation of hypertext links]	1.0	5	1	0	0
1608	91	error character	[error characters]	1.0	2	1	0	0
1609	91	two distinct domain tutorial dialogue	[two distinct domains tutorial dialogue]	1.0	5	1	0	0
1610	91	such question	[such questions]	1.0	2	1	0	0
1611	91	ONTOSCORE a system for scoring set	[ONTOSCORE a system for scoring sets]	1.0	6	1	0	0
1612	91	KANA-KANJI conversion	[KANA-KANJI conversion]	1.0	2	1	0	0
1613	91	similar model	[similar models]	1.0	2	1	0	0
1614	91	event element	[event elements]	1.0	2	1	0	0
1615	91	Minimalist grammar	[Minimalist grammars]	1.0	2	1	0	0
1616	91	several aspect	[several aspects]	1.0	2	1	0	0
1617	91	language parsing	[language parsing]	1.0	2	2	2	2
1618	91	predicate-argument structure	[predicate-argument structures]	1.0	2	2	2	2
1619	91	ATIS benchmark	[ATIS benchmark]	1.0	2	1	0	0
1620	91	research context	[research context]	1.0	2	1	0	0
1621	91	bigram feature	[bigram features]	1.0	2	1	0	0
1622	91	right context	[right context]	1.0	2	1	0	0
1623	91	SL expression	[SL expressions]	1.0	2	1	0	0
1624	91	logical model	[logical model]	1.0	2	2	1	1
1625	91	Natural language processing nlp system	[Natural Language Processing NLP systems]	1.0	5	1	0	0
1626	91	key feature of the system	[key features of the system]	1.0	5	1	0	0
1627	91	unigram model for machine translation	[unigram model for machine translation]	1.0	5	1	0	0
1628	91	many system	[many systems]	1.0	2	2	1	1
1629	91	automatically identify predicate argument structure	[automatically identifying predicate argument structures]	1.0	5	1	0	0
1630	91	manual word	[manual word]	1.0	2	1	0	0
1631	91	rule-based model	[rule-based model]	1.0	2	1	0	0
1632	91	standard use	[standard use]	1.0	2	1	0	0
1633	91	source text	[source text]	1.0	2	2	1	1
1634	91	research project	[research projects]	1.0	2	1	0	0
1635	91	wsd accuracy of SMT model	[WSD accuracy of SMT models]	1.0	5	1	0	0
1636	91	other text	[other texts]	1.0	2	1	0	0
1637	91	CCR formalism	[CCR formalism]	1.0	2	1	0	0
1638	91	new theory of discourse structure	[new theory of discourse structure]	1.0	5	1	0	0
1639	91	small corpus of the language	[small corpus of the language]	1.0	5	1	0	0
1640	91	categorial grammar	[categorial grammar]	1.0	2	1	0	0
1641	91	use a logical model in place	[using a logical model in place]	1.0	6	1	0	0
1642	91	nlp technique for document classification	[NLP techniques for document classification]	1.0	5	1	0	0
1643	91	2-character Chinese name without title	[2-character Chinese names without title]	1.0	5	1	0	0
1644	91	NIST score	[NIST score]	1.0	2	1	0	0
1645	91	own knowledge	[own knowledge]	1.0	2	1	0	0
1646	91	conventional word-trigram recognition require manual transcription	[conventional word-trigram recognition requiring manual transcription]	1.0	6	1	0	0
1647	91	other approach	[other approaches]	1.0	2	2	1	1
1648	91	baseline method	[baseline methods]	1.0	2	1	0	0
1649	91	improve the performance of the word-sense	[improving the performance of the word-sense]	1.0	6	1	0	0
1650	91	only resource	[only resource]	1.0	2	2	2	2
1651	91	future need	[future needs]	1.0	2	1	0	0
1652	91	phrase translation from word-based alignment	[phrase translations from word-based alignments]	1.0	5	1	0	0
1653	91	dialogue pattern	[dialogue patterns]	1.0	2	1	0	0
1654	91	same sense	[same sense]	1.0	2	1	0	0
1655	91	phonological knowledge	[phonological knowledge]	1.0	2	1	0	0
1656	91	second model	[second model]	1.0	2	1	0	0
1657	91	introduction to statistical machine translation	[introduction to statistical machine translation]	1.0	5	1	0	0
1658	91	word fragment	[word fragments]	1.0	2	1	0	0
1659	91	manual method	[manual methods]	1.0	2	1	0	0
1660	91	hard decision	[hard decisions]	1.0	2	1	0	0
1661	91	adequate tool	[adequate tool]	1.0	2	1	0	0
1662	91	useful tool	[useful tool]	1.0	2	1	0	0
1663	91	include a frequent word on aps	[including a frequent word on APs]	1.0	6	1	0	0
1664	91	speaker leeway form a utterance	[speaker leeway forming an utterance]	1.0	5	1	0	0
1665	91	Montague semantics	[Montague semantics]	1.0	2	2	2	2
1666	91	variation on the evaluation metric	[variation on the evaluation metric]	1.0	5	1	0	0
1667	91	same problem	[same problem]	1.0	2	1	0	0
1668	91	overlap speech	[overlapping speech]	1.0	2	1	0	0
1669	91	inherent ambiguity	[inherent ambiguity]	1.0	2	1	0	0
1670	91	combination of the training speaker	[combination of the training speakers]	1.0	5	1	0	0
1671	91	accuracy difference between the approach	[accuracy difference between the approaches]	1.0	5	1	0	0
1672	91	good prediction	[better predictions]	1.0	2	1	0	0
1673	91	three tier	[three tiers]	1.0	2	1	0	0
1674	91	behavioral pattern	[behavioral patterns]	1.0	2	1	0	0
1675	91	use a argumentation system by Konolige	[using an argumentation system by Konolige]	1.0	6	1	0	0
1676	91	identify hub in a automaton	[identifying hubs in an automaton]	1.0	5	1	0	0
1677	91	list of output sentence plan	[list of output sentence plans]	1.0	5	1	0	0
1678	91	contextual information	[contextual information]	1.0	2	1	0	0
1679	91	explicit use of precede tag context	[explicit use of preceding tag contexts]	1.0	6	1	0	0
1680	91	unification-based grammar	[unification-based grammar]	1.0	2	1	0	0
1681	91	semantic-head-driven generation	[semantic-head-driven generation]	1.0	2	1	0	0
1682	91	tag LMs with confidence measure	[tagging LMs with confidence measures]	1.0	5	1	0	0
1683	91	equivalence between the standard use	[equivalence between the standard use]	1.0	5	1	0	0
1684	91	stop list	[stop lists]	1.0	2	1	0	0
1685	91	build polarity-tagged corpus from HTML document	[building polarity-tagged corpus from HTML documents]	1.0	6	1	0	0
1686	91	utterance feature	[utterance features]	1.0	2	1	0	0
1687	91	text structure	[text structure]	1.0	2	1	0	0
1688	91	graph unification	[Graph unification]	1.0	2	1	0	0
1689	91	text corpus	[text corpus]	1.0	2	1	0	0
1690	91	extraposition grammar	[extraposition grammars]	1.0	2	1	0	0
1691	91	/Soames 1979/	[/Soames 1979/]	1.0	2	1	0	0
1692	91	several approach	[several approaches]	1.0	2	1	0	0
1693	91	robust analysis	[robust analysis]	1.0	2	1	0	0
1694	91	likely repair	[likely repair]	1.0	2	1	0	0
1695	91	/Mercer 1987/	[/Mercer 1987/]	1.0	2	1	0	0
1696	91	Begin/After tagging	[Begin/After tagging]	1.0	2	1	0	0
1697	91	transfer component for parse reduction	[transfer component for parse reduction]	1.0	5	1	0	0
1698	91	semantic set	[semantic set]	1.0	2	1	0	0
1699	91	language L	[language L]	1.0	2	1	0	0
1700	91	translate in another natural language	[translates in another natural language]	1.0	5	1	0	0
1701	91	commonsense reasoning	[commonsense reasoning]	1.0	2	1	0	0
1702	91	alternative index	[alternative index]	1.0	2	1	0	0
1703	91	statistical machine translation tool kit	[statistical machine translation tool kit]	1.0	5	1	0	0
1704	91	speech-search algorithm	[speech-search algorithm]	1.0	2	1	0	0
1705	91	restricted-domain parsing	[restricted-domain parsing]	1.0	2	1	0	0
1706	91	effect of the one sense	[effect of the one sense]	1.0	5	1	0	0
1707	91	language understand system with agent	[language understanding system with agents]	1.0	5	1	0	0
1708	91	direct imitation of human performance	[direct imitation of human performance]	1.0	5	1	0	0
1709	91	Bayesian learning	[Bayesian learning]	1.0	2	1	0	0
1710	91	understanding process of the description	[understanding process of the descriptions]	1.0	5	1	0	0
1711	91	ambiguity handling	[ambiguity handling]	1.0	2	1	0	0
1712	91	Transition Network	[Transition Network]	1.0	2	1	0	0
1713	91	retrieval accuracy superior to any	[retrieval accuracy superior to any]	1.0	5	1	0	0
1714	91	semantic question	[semantic questions]	1.0	2	1	0	0
1715	91	main dictionary	[main dictionary]	1.0	2	1	0	0
1716	91	dependency relation	[dependency relations]	1.0	2	2	1	1
1717	91	three different language system domain	[three different language system domains]	1.0	5	1	0	0
1718	91	independent approach	[independent approach]	1.0	2	1	0	0
1719	91	translation speed	[translation speed]	1.0	2	1	0	0
1720	91	unsupervised word	[unsupervised word]	1.0	2	1	0	0
1721	91	only system	[only system]	1.0	2	1	0	0
1722	91	5 speed-up	[5 speed-up]	1.0	2	1	0	0
1723	91	great effectiveness of this approach	[great effectiveness of this approach]	1.0	5	1	0	0
1724	91	paraphrase set	[paraphrase sets]	1.0	2	2	1	1
1725	91	performance as a similarity measure	[performance as a similarity measure]	1.0	5	1	0	0
1726	91	simple interpolation method like log-linear interpolation	[simple interpolation methods like log-linear interpolation]	1.0	6	1	0	0
1727	91	/Gazdar 1979/	[/Gazdar 1979/]	1.0	2	1	0	0
1728	91	intelligent agent	[intelligent agent]	1.0	2	2	2	2
1729	91	computational system	[computational system]	1.0	2	2	2	2
1730	91	target text	[target texts]	1.0	2	1	0	0
1731	91	full analysis	[full analysis]	1.0	2	2	1	1
1732	91	Thai prediction	[Thai prediction]	1.0	2	1	0	0
1733	91	practical system	[practical systems]	1.0	2	1	0	0
1734	91	natural manner	[natural manner]	1.0	2	1	0	0
1735	91	line length	[line length]	1.0	2	1	0	0
1736	91	answer capability on free text	[answering capability on free text]	1.0	5	1	0	0
1737	91	need develop polynomial time approximation	[need developing polynomial time approximations]	1.0	5	1	0	0
1738	91	place comma	[placing commas]	1.0	2	2	2	2
1739	91	dictionary-based word	[dictionary-based word]	1.0	2	1	0	0
1740	91	ideal aim for complete interpretation	[ideal aiming for complete interpretations]	1.0	5	1	0	0
1741	91	Arabic stemmer	[Arabic stemmer]	1.0	2	2	2	2
1742	91	evaluation method	[evaluation method]	1.0	2	2	1	1
1743	91	take advantage of the symmetric pattern	[taking advantage of the symmetric patterns]	1.0	6	1	0	0
1744	91	Multi-site ATIS data COllection work group	[Multi-site ATIS Data COllection Working group]	1.0	6	1	0	0
1745	91	efficient decoder	[efficient decoder]	1.0	2	1	0	0
1746	91	syntax-based constraint for word alignment	[syntax-based constraint for word alignment]	1.0	5	1	0	0
1747	91	deterministic automaton	[deterministic automaton]	1.0	2	1	0	0
1748	91	videodisc image	[videodisc images]	1.0	2	1	0	0
1749	91	correct parse	[correct parse]	1.0	2	2	2	2
1750	91	corresponding parser	[corresponding parser]	1.0	2	1	0	0
1751	91	human-ranked sentence	[human-ranked sentence]	1.0	2	1	0	0
1752	91	serious attention	[serious attention]	1.0	2	1	0	0
1753	91	73 .2%	[73 .2%]	1.0	2	1	0	0
1754	91	human cognition	[human cognition]	1.0	2	1	0	0
1755	91	utility cluster	[utility clustering]	1.0	2	1	0	0
1756	91	well-written discourse	[well-written discourse]	1.0	2	1	0	0
1757	91	conversational agent	[conversational agents]	1.0	2	1	0	0
1758	91	current system	[current systems]	1.0	2	1	0	0
1759	91	learning algorithm	[learning algorithm]	1.0	2	2	2	2
1760	91	addressee of a dialogue act	[addressee of a dialogue act]	1.0	5	1	0	0
1761	91	translation candidate	[translation candidates]	1.0	2	1	0	0
1762	91	lack of manually sense-tagged datum	[lack of manually sense-tagged data]	1.0	5	1	0	0
1763	91	restriction on the semantic category	[restrictions on the semantic categories]	1.0	5	1	0	0
1764	91	simple notation	[simple notations]	1.0	2	1	0	0
1765	91	language model of abstract move	[language model of abstract moves]	1.0	5	1	0	0
1766	91	result from experiment with substitution cost	[Results from experiments with substitution costs]	1.0	6	1	0	0
1767	91	such synthesis	[Such synthesis]	1.0	2	1	0	0
1768	91	hot spot	[hot spots]	1.0	2	1	0	0
1769	91	more traditional natural language interface	[more traditional natural language interfaces]	1.0	5	1	0	0
1770	91	syntactic category on the terminal	[syntactic categories on the terminal]	1.0	5	1	0	0
1771	91	computational lexicon	[computational lexicon]	1.0	2	2	1	1
1772	91	language generation	[language generation]	1.0	2	2	2	2
1773	91	conversational feature	[conversational features]	1.0	2	1	0	0
1774	91	generative model	[generative model]	1.0	2	2	1	1
1775	91	place comma with a precision	[placing commas with a precision]	1.0	5	1	0	0
1776	91	automatic method	[automatic methods]	1.0	2	1	0	0
1777	91	paraphrase set from seed sentence	[paraphrase sets from seed sentences]	1.0	5	1	0	0
1778	91	unified approach	[unified approach]	1.0	2	1	0	0
1779	91	possible indexing	[possible indexings]	1.0	2	1	0	0
1780	91	post-process the output of ocr system	[post-processing the output of OCR systems]	1.0	6	1	0	0
1781	91	Markov probability of a correct chain	[Markov probability of a correct chain]	1.0	6	1	0	0
1782	91	addition to a unsupervised component	[addition to an unsupervised component]	1.0	5	1	0	0
1783	91	new problem	[new problem]	1.0	2	2	1	1
1784	91	efficient Prolog cf query optimisation	[efficient Prolog cf query optimisation]	1.0	5	1	0	0
1785	91	information source	[information sources]	1.0	2	1	0	0
1786	91	statistical approach	[Statistical approach]	1.0	2	2	1	1
1787	91	hand-crafted rule	[hand-crafted rules]	1.0	2	1	0	0
1788	91	adjoin word	[adjoining words]	1.0	2	1	0	0
1789	91	real task	[real tasks]	1.0	2	1	0	0
1790	91	result Arabic word segmentation system	[resulting Arabic word segmentation system]	1.0	5	1	0	0
1791	91	nature of the missing fragment	[nature of the missing fragments]	1.0	5	1	0	0
1792	91	finite-state model	[finite-state models]	1.0	2	1	0	0
1793	91	manual determination	[manual determination]	1.0	2	1	0	0
1794	91	social rank of the referent	[social ranks of the referents]	1.0	5	1	0	0
1795	91	erroneous chain	[erroneous chains]	1.0	2	1	0	0
1796	91	knowledge for a user model	[knowledge for a user model]	1.0	5	1	0	0
1797	91	technique for automatically training module	[Techniques for automatically training modules]	1.0	5	1	0	0
1798	91	world knowledge of the situation	[world knowledge of the situation]	1.0	5	1	0	0
1799	91	usually the beginning of the sentence	[usually the beginning of the sentence]	1.0	6	1	0	0
1800	91	new input	[new inputs]	1.0	2	1	0	0
1801	91	potential antecedence of each element	[potential antecedence of each element]	1.0	5	1	0	0
1802	91	use the NIST mt evaluation software	[using the NIST MT evaluation software]	1.0	6	1	0	0
1803	91	conceptual problem	[conceptual problems]	1.0	2	1	0	0
1804	91	combination with decision tree building	[combination with decision tree building]	1.0	5	1	0	0
1805	91	conversational cue such as cue phrase	[conversational cues such as cue phrases]	1.0	6	1	0	0
1806	91	/Soames 1982/	[/Soames 1982/]	1.0	2	1	0	0
1807	91	linguistic knowledge	[linguistic knowledge]	1.0	2	1	0	0
1808	91	word matching	[word matchings]	1.0	2	1	0	0
1809	91	35 .0%	[35 .0%]	1.0	2	1	0	0
1810	91	human judge	[human judges]	1.0	2	1	0	0
1811	91	method detect	[method detecting]	1.0	2	1	0	0
1812	91	include a variety of meta-inference type	[including a variety of meta-inference types]	1.0	6	1	0	0
1813	91	equivalent word	[equivalent words]	1.0	2	1	0	0
1814	91	linguistic generality	[linguistic generality]	1.0	2	1	0	0
1815	91	eigenvector of a adjacency graph's laplacian	[eigenvectors of an adjacency graph's Laplacian]	1.0	6	1	0	0
1816	91	principle advantage of this approach	[principle advantage of this approach]	1.0	5	1	0	0
1817	91	,000 word	[,000 words]	1.0	2	3	2	1
1818	91	particular representational format for meaning	[particular representational format for meaning]	1.0	5	1	0	0
1819	91	annotate scheme	[annotating scheme]	1.0	2	1	0	0
1820	91	preliminary analysis	[preliminary analysis]	1.0	2	1	0	0
1821	91	logical definition of Minimalist grammar	[logical definition of Minimalist grammars]	1.0	5	1	0	0
1822	91	English sentence	[English sentence]	1.0	2	2	1	1
1823	91	two way	[two ways]	1.0	2	1	0	0
1824	91	research use machine learning technique	[research using machine learning techniques]	1.0	5	1	0	0
1825	91	candidate antecedent	[candidate antecedents]	1.0	2	1	0	0
1826	91	logical definition	[logical definition]	1.0	2	2	1	1
1827	91	lateral direction	[lateral directions]	1.0	2	1	0	0
1828	91	syntactic structure	[syntactic structure]	1.0	2	3	2	1
1829	91	correlation between automatic evaluation measure	[correlation between automatic evaluation measures]	1.0	5	1	0	0
1830	91	Paul a computer text generation system	[Paul a computer text generation system]	1.0	6	1	0	0
1831	91	linguistic object	[linguistic objects]	1.0	2	1	0	0
1832	91	event network	[event networks]	1.0	2	1	0	0
1833	91	fully-connected first-order statistical class grammar	[fully-connected first-order statistical class grammar]	1.0	5	1	0	0
1834	91	superordinate substitution	[superordinate substitution]	1.0	2	1	0	0
1835	91	language learning	[language learning]	1.0	2	2	2	2
1836	91	NE performance for some NE type	[NE performance for some NE types]	1.0	6	1	0	0
1837	91	several kind of linguistic knowledge	[several kinds of linguistic knowledge]	1.0	5	1	0	0
1838	91	statistical system	[statistical systems]	1.0	2	1	0	0
1839	91	linguistic pattern	[linguistic pattern]	1.0	2	1	0	0
1840	91	bi-directional fragment	[bi-directional fragment]	1.0	2	1	0	0
1841	91	describe inflection for the purpose	[describing inflection for the purpose]	1.0	5	1	0	0
1842	91	special method	[special method]	1.0	2	2	1	1
1843	91	approach induce a grammar from corpus	[approach inducing a grammar from corpora]	1.0	6	1	0	0
1844	91	state-of-the-art approach	[state-of-the-art approaches]	1.0	2	1	0	0
1845	91	surprising finding	[surprising finding]	1.0	2	1	0	0
1846	91	Canadian Hansards	[Canadian Hansards]	1.0	2	1	0	0
1847	91	evaluate word sense disambiguation WSD model	[evaluating word sense disambiguation WSD models]	1.0	6	1	0	0
1848	91	coherent (given	[coherent (given]	1.0	2	1	0	0
1849	91	development of a message-passing infrastructure	[development of a message-passing infrastructure]	1.0	5	1	0	0
1850	91	conceptual Dependency	[Conceptual Dependency]	1.0	2	1	0	0
1851	91	recent advance in phrasal translation	[recent advances in phrasal translation]	1.0	5	1	0	0
1852	91	subject of much recent debate	[subject of much recent debate]	1.0	5	1	0	0
1853	91	avoidance of copying of unmodified subgraph	[avoidance of copying of unmodified subgraphs]	1.0	6	1	0	0
1854	91	include vocabulary	[including vocabulary]	1.0	2	1	0	0
1855	91	probably-correct rule	[probably-correct rules]	1.0	2	1	0	0
1856	91	perplexing fact	[perplexing fact]	1.0	2	1	0	0
1857	91	center on the construction of model	[centering on the construction of models]	1.0	6	1	0	0
1858	91	full meaning	[full meaning]	1.0	2	1	0	0
1859	91	preliminary experiment	[preliminary experiments]	1.0	2	1	0	0
1860	91	fuzzy syllogism	[fuzzy syllogism]	1.0	2	1	0	0
1861	91	similar indifference	[similar indifference]	1.0	2	1	0	0
1862	91	information redundancy in multilingual input	[information redundancy in multilingual input]	1.0	5	1	0	0
1863	91	system combination for unsupervised wsd	[system combination for unsupervised WSD]	1.0	5	1	0	0
1864	91	information nugget	[information nugget]	1.0	2	1	0	0
1865	91	expression include a frequent word	[expressions including a frequent word]	1.0	5	1	0	0
1866	91	resource management	[Resource Management]	1.0	2	2	2	2
1867	91	syntax-based decoder	[syntax-based decoder]	1.0	2	1	0	0
1868	91	important information	[important information]	1.0	2	1	0	0
1869	91	annotation experiment	[annotation experiment]	1.0	2	1	0	0
1870	91	most important task build the editor	[most important task building the editor]	1.0	6	1	0	0
1871	91	correct this error in Japanese bunsetsu	[correcting these errors in Japanese bunsetsu]	1.0	6	1	0	0
1872	91	semantic verb	[semantic verb]	1.0	2	1	0	0
1873	91	standard grammar	[standard grammar]	1.0	2	1	0	0
1874	91	parse	[parse]	1.0	1	11	5	5
1875	91	training corpus	[training corpus]	1.0	2	2	1	1
1876	91	understand system	[understanding system]	1.0	2	2	2	2
1877	91	example-retrieval er	[example-retrieval ER]	1.0	2	1	0	0
1878	91	primarily pass some computational issue	[primarily passing some computational issues]	1.0	5	1	0	0
1879	91	cohesion constraint	[cohesion constraint]	1.0	2	1	0	0
1880	91	traditional approach	[traditional approach]	1.0	2	2	2	2
1881	91	multiple output	[multiple output]	1.0	2	1	0	0
1882	91	asynchronous message passing between actor	[asynchronous message passing between actors]	1.0	5	1	0	0
1883	91	combine knowledge about dialog schemata	[combining knowledge about dialog schemata]	1.0	5	1	0	0
1884	91	system RAREAS	[system RAREAS]	1.0	2	1	0	0
1885	91	strongly biase	[strongly biasing]	1.0	2	1	0	0
1886	91	three separate component the structure	[three separate components the structure]	1.0	5	1	0	0
1887	91	separation margin between the candidate	[separation margin between the candidate]	1.0	5	1	0	0
1888	91	proper substitution for this element	[proper substitutions for these elements]	1.0	5	1	0	0
1889	91	good confidence	[best confidence]	1.0	2	1	0	0
1890	91	thus the number of edge	[thus the number of edges]	1.0	5	1	0	0
1891	91	hand-crafted system	[hand-crafted system]	1.0	2	1	0	0
1892	91	superior performance	[superior performance]	1.0	2	1	0	0
1893	91	synchronous dependency	[synchronous dependency]	1.0	2	1	0	0
1894	91	structure contain set of feature	[structures containing sets of features]	1.0	5	1	0	0
1895	91	robust parsing	[Robust parsing]	1.0	2	2	1	1
1896	91	cluster algorithm	[clustering algorithm]	1.0	2	1	0	0
1897	91	use of multiple parsing strategy	[use of multiple parsing strategies]	1.0	5	1	0	0
1898	91	corrective training	[corrective training]	1.0	2	1	0	0
1899	91	word-based model	[word-based models]	1.0	2	1	0	0
1900	91	maintain the quasi-destructive scheme' ability	[maintaining the quasi-destructive scheme's ability]	1.0	5	1	0	0
1901	91	original document	[original document]	1.0	2	1	0	0
1902	91	anaphora reference	[anaphora references]	1.0	2	1	0	0
1903	91	semantic annotation	[semantic annotation]	1.0	2	1	0	0
1904	91	experiment use machine translation output	[experiment using machine translation output]	1.0	5	1	0	0
1905	91	term of event type network	[terms of event type networks]	1.0	5	1	0	0
1906	91	identify systematic pattern in translation datum	[identifying systematic patterns in translation data]	1.0	6	1	0	0
1907	91	number of performance enhancing technique	[number of performance enhancing techniques]	1.0	5	1	0	0
1908	91	computational phonological system speech recognition	[computational phonological system speech recognition]	1.0	5	1	0	0
1909	91	information related to a particular scenario	[information related to a particular scenario]	1.0	6	1	0	0
1910	91	advantage of detailed linguistic information	[advantage of detailed linguistic information]	1.0	5	1	0	0
1911	91	conversion rule	[conversion rules]	1.0	2	1	0	0
1912	91	hard decision use the reference	[hard decisions using the reference]	1.0	5	1	0	0
1913	91	experimental evaluation of summarization quality	[experimental evaluation of summarization quality]	1.0	5	1	0	0
1914	91	kanji-kana character	[kanji-kana characters]	1.0	2	2	1	1
1915	91	consist of 126 ,610 sentence	[consisting of 126 ,610 sentences]	1.0	5	1	0	0
1916	91	application at the character level	[application at the character level]	1.0	5	1	0	0
1917	91	conversational speech	[conversational speech]	1.0	2	1	0	0
1918	91	Ie paradigm	[IE paradigm]	1.0	2	2	2	2
1919	91	two strategy for language understanding	[two strategies for language understanding]	1.0	5	1	0	0
1920	91	isomorphic Grammar approach to mt	[Isomorphic Grammars approach to MT]	1.0	5	1	0	0
1921	91	non Verbal	[Non Verbal]	1.0	2	1	0	0
1922	91	induce a grammar from parallel corpus	[inducing a grammar from parallel corpora]	1.0	6	1	0	0
1923	91	query optimisation in a relational database	[query optimisation in a relational database]	1.0	6	1	0	0
1924	91	unsupervised learning algorithm for identification	[unsupervised learning algorithm for identification]	1.0	5	1	0	0
1925	91	speculative claim	[speculative claim]	1.0	2	2	1	1
1926	91	Dynamic Hierarchical Phrasal Lexicon DHPL	[Dynamic Hierarchical Phrasal Lexicon DHPL]	1.0	5	1	0	0
1927	91	English-Chinese bitext	[English-Chinese bitexts]	1.0	2	2	1	1
1928	91	retrieval performance	[retrieval performance]	1.0	2	1	0	0
1929	91	good insight into the potential	[good insight into the potential]	1.0	5	1	0	0
1930	91	cooperative response to each user	[cooperative responses to each user]	1.0	5	1	0	0
1931	91	computational analysis of move structure	[computational analysis of move structures]	1.0	5	1	0	0
1932	91	word string with the performance	[word string with the performance]	1.0	5	1	0	0
1933	91	valuable task	[valuable task]	1.0	2	1	0	0
1934	91	one sentence	[one sentences]	1.0	2	1	0	0
1935	91	parser incorporate the control structure	[parser incorporating the control structure]	1.0	5	1	0	0
1936	91	linguistic analysis	[linguistic analysis]	1.0	2	2	1	1
1937	91	new stem	[new stems]	1.0	2	1	0	0
1938	91	syntactic information	[syntactic information]	1.0	2	1	0	0
1939	91	algorithmic addition	[algorithmic additions]	1.0	2	1	0	0
1940	91	combine the result from answer agent	[combining the results from answering agents]	1.0	6	1	0	0
1941	91	automatically evaluate the summarization quality	[automatically evaluating the summarization quality]	1.0	5	1	0	0
1942	91	chart-based phrase	[chart-based phrase]	1.0	2	1	0	0
1943	91	probabilistic model	[probabilistic model]	1.0	2	2	1	1
1944	91	transfer phase in machine translation	[transfer phase in machine translation]	1.0	5	1	0	0
1945	91	ambiguous sentence	[ambiguous sentences]	1.0	2	1	0	0
1946	91	new strategy	[new strategy]	1.0	2	1	0	0
1947	91	apply statistical model to datum	[applying statistical models to data]	1.0	5	1	0	0
1948	91	execution at the appropriate database	[execution at the appropriate database]	1.0	5	1	0	0
1949	91	speech understanding	[speech understanding]	1.0	2	2	1	1
1950	91	rank rule	[ranking rules]	1.0	2	1	0	0
1951	91	source language	[source language]	1.0	2	1	0	0
1952	91	large Chinese non-parallel newspaper corpus	[large Chinese non-parallel newspaper corpora]	1.0	5	1	0	0
1953	91	effectiveness of answer resolution algorithm	[effectiveness of answer resolution algorithm]	1.0	5	1	0	0
1954	91	dialog schemata	[dialog schemata]	1.0	2	2	2	2
1955	91	further work	[further work]	1.0	2	1	0	0
1956	91	training phase	[training phase]	1.0	2	1	0	0
1957	91	SUN Sparcstation	[SUN Sparcstations]	1.0	2	1	0	0
1958	91	text processing	[text processing]	1.0	2	1	0	0
1959	91	different amount	[different amounts]	1.0	2	1	0	0
1960	91	well exist	[best existing]	1.0	2	1	0	0
1961	91	successive learner	[successive learners]	1.0	2	2	1	1
1962	91	quadratic time	[quadratic time]	1.0	2	1	0	0
1963	91	specialized editor	[specialized editor]	1.0	2	1	0	0
1964	91	significant improvement in alignment quality	[significant improvement in alignment quality]	1.0	5	1	0	0
1965	91	sentence structure	[sentence structure]	1.0	2	2	2	2
1966	91	practical consideration	[practical considerations]	1.0	2	1	0	0
1967	91	semantically classify undisambiguated SCF datum	[semantically classifying undisambiguated SCF data]	1.0	5	1	0	0
1968	91	49 .3%	[49 .3%]	1.0	2	1	0	0
1969	91	processing description	[processing description]	1.0	2	1	0	0
1970	91	use parsing strategy in a fashion	[using parsing strategies in a fashion]	1.0	6	1	0	0
1971	91	general mechanism	[general mechanism]	1.0	2	1	0	0
1972	91	world knowledge	[world knowledge]	1.0	2	2	1	1
1973	91	CCLINC common Coalition Language System	[CCLINC Common Coalition Language System]	1.0	5	1	0	0
1974	91	novel method build polarity-tagged corpus	[novel method building polarity-tagged corpus]	1.0	5	1	0	0
1975	91	decoding algorithm	[decoding algorithm]	1.0	2	1	0	0
1976	91	boost approach	[boosting approach]	1.0	2	2	2	2
1977	91	dramatic improvement	[dramatic improvements]	1.0	2	1	0	0
1978	91	use a binary relational calculus notation	[using a binary relational calculus notation]	1.0	6	1	0	0
1979	91	korean-to-english machine	[Korean-to-English machine]	1.0	2	1	0	0
1980	91	much more commonly available resource	[much more commonly available resource]	1.0	5	1	0	0
1981	91	unsupervised algorithm	[unsupervised algorithm]	1.0	2	2	1	1
1982	91	place of a denotational semantics	[place of a denotational semantics]	1.0	5	1	0	0
1983	91	text collection	[text collection]	1.0	2	2	2	2
1984	91	feature space	[feature space]	1.0	2	1	0	0
1985	91	basic methodology	[Basic methodology]	1.0	2	1	0	0
1986	91	segment order-sensitive method in term	[segment order-sensitive methods in terms]	1.0	5	1	0	0
1987	91	2 srh	[2 SRHs]	1.0	2	1	0	0
1988	91	common ground in human-computer interaction	[common ground in human-computer interaction]	1.0	5	1	0	0
1989	91	word-aligned corpus	[word-aligned corpora]	1.0	2	1	0	0
1990	91	disambiguation the interpretation of ellipse	[disambiguation the interpretation of ellipses]	1.0	5	1	0	0
1991	91	hand-produced set	[hand-produced sets]	1.0	2	1	0	0
1992	91	set of individually high-precision rule	[sets of individually high-precision rules]	1.0	5	1	0	0
1993	91	two new word-sense disambiguation system	[two new word-sense disambiguation systems]	1.0	5	1	0	0
1994	91	FERRET a interactive question-answering q/a system	[FERRET an interactive question-answering Q/A system]	1.0	6	1	0	0
1995	91	user satisfaction	[user satisfaction]	1.0	2	1	0	0
1996	91	speech Recognition	[Speech Recognition]	1.0	2	2	2	2
1997	91	accurate translation	[accurate translation]	1.0	2	1	0	0
1998	91	kind of Semantic role system	[kind of Semantic Role System]	1.0	5	1	0	0
1999	91	ambiguity packing	[ambiguity packing]	1.0	2	2	1	1
2000	91	English-Chinese bitext of the law	[English-Chinese bitexts of the laws]	1.0	5	1	0	0
2001	91	high quality	[High quality]	1.0	2	2	1	1
2002	91	pronominal anaphora	[pronominal anaphora]	1.0	2	1	0	0
2003	91	core technology	[core technology]	1.0	2	1	0	0
2004	91	adaptive nature	[adaptive nature]	1.0	2	1	0	0
2005	91	natural grammar	[natural grammar]	1.0	2	1	0	0
2006	91	chart-like parsing	[Chart-like parsing]	1.0	2	1	0	0
2007	91	state of focus of attention	[state of focus of attention]	1.0	5	1	0	0
2008	91	ocr system	[OCR system]	1.0	2	2	1	1
2009	91	decision tree	[decision tree]	1.0	2	1	0	0
2010	91	logical formalism	[logical formalism]	1.0	2	1	0	0
2011	91	context-free grammar	[context-free grammars]	1.0	2	2	1	1
2012	91	reformulation of the erroneous sentence	[reformulation of the erroneous sentence]	1.0	5	1	0	0
2013	91	require significantly less development time	[requiring significantly less development time]	1.0	5	1	0	0
2014	91	VME bus	[VME bus]	1.0	2	1	0	0
2015	91	synthesis system	[synthesis system]	1.0	2	1	0	0
2016	91	development of machine translation system	[development of machine translation systems]	1.0	5	1	0	0
2017	91	sentence analysis	[sentence analysis]	1.0	2	2	1	1
2018	91	much effort	[Much effort]	1.0	2	1	0	0
2019	91	calculate eigenvector	[calculating eigenvectors]	1.0	2	1	0	0
2020	91	good system	[best system]	1.0	2	1	0	0
2021	91	Computer program	[Computer programs]	1.0	2	1	0	0
2022	91	employ a traditional generative phonological approach	[employing a traditional generative phonological approach]	1.0	6	1	0	0
2023	91	available knowledge	[available knowledge]	1.0	2	1	0	0
2024	91	hard problem	[hard problems]	1.0	2	1	0	0
2025	91	state-of-the-art performance	[state-of-the-art performance]	1.0	2	1	0	0
2026	91	finite State	[Finite State]	1.0	2	1	0	0
2027	91	comparative evaluation	[comparative evaluations]	1.0	2	1	0	0
2028	91	unidentified portion	[unidentified portion]	1.0	2	1	0	0
2029	91	training material	[training material]	1.0	2	1	0	0
2030	91	use this tree-based model in combination	[using these tree-based models in combination]	1.0	6	1	0	0
2031	91	sense coverage	[sense coverage]	1.0	2	1	0	0
2032	91	decision list	[decision list]	1.0	2	1	0	0
2033	91	usually in a rightward manner	[usually in a rightward manner]	1.0	5	1	0	0
2034	91	last decade	[last decade]	1.0	2	2	1	1
2035	91	HBG model	[HBG model]	1.0	2	1	0	0
2036	91	most bird	[most birds]	1.0	2	1	0	0
2037	91	gather statistic on the output	[gathering statistics on the output]	1.0	5	1	0	0
2038	91	official ranking	[official rankings]	1.0	2	1	0	0
2039	91	demonstrative form	[demonstrative forms]	1.0	2	1	0	0
2040	91	unsupervised training	[unsupervised training]	1.0	2	1	0	0
2041	91	respect to the original question	[respect to the original question]	1.0	5	1	0	0
2042	91	character recognition	[character recognition]	1.0	2	2	2	2
2043	91	high cost	[high cost]	1.0	2	2	1	1
2044	91	automate certain aspect of language learning	[automating certain aspects of language learning]	1.0	6	1	0	0
2045	91	on-line community	[on-line communities]	1.0	2	1	0	0
2046	91	input text	[input text]	1.0	2	1	0	0
2047	91	automatic word	[automatic word]	1.0	2	1	0	0
2048	91	non-literal aspect	[non-literal aspects]	1.0	2	2	2	2
2049	91	term of the inference type	[terms of the inference types]	1.0	5	1	0	0
2050	91	find the language for a need	[Finding the language for a need]	1.0	6	1	0	0
2051	91	optimum configuration	[optimum configuration]	1.0	2	1	0	0
2052	91	event term	[event terms]	1.0	2	1	0	0
2053	91	local tree	[local trees]	1.0	2	1	0	0
2054	91	preference knowledge	[preference knowledge]	1.0	2	2	1	1
2055	91	conjunction with a lr parser	[conjunction with a LR parser]	1.0	5	1	0	0
2056	91	LHIP a system for grammar development	[LHIP a system for grammar development]	1.0	6	1	0	0
2057	91	improvement on several machine translation test	[improvement on several machine translation tests]	1.0	6	1	0	0
2058	91	new mechanism	[new mechanism]	1.0	2	1	0	0
2059	91	model use associative processor aps	[model using associative processors APs]	1.0	5	1	0	0
2060	91	zero occurrence of a morpheme	[zero occurrences of a morpheme]	1.0	5	1	0	0
2061	91	addressee identification in four-participant meeting	[addressee identification in four-participants meetings]	1.0	5	1	0	0
2062	91	clear evidence from the input	[clear evidence from the input]	1.0	5	1	0	0
2063	91	trigram model	[trigram model]	1.0	2	2	2	2
2064	91	assignment of descriptor from phrase	[assignment of descriptors from phrases]	1.0	5	1	0	0
2065	91	essential underpin	[essential underpinning]	1.0	2	1	0	0
2066	91	different strategy	[different strategies]	1.0	2	2	1	1
2067	91	deictic reference about the discourse	[Deictic reference about the discourse]	1.0	5	1	0	0
2068	91	requirement for a software infrastructure	[requirements for a software infrastructure]	1.0	5	1	0	0
2069	91	paraphrase extraction	[paraphrase extraction]	1.0	2	1	0	0
2070	91	essential role	[essential roles]	1.0	2	1	0	0
2071	91	semantic grammar	[semantic grammar]	1.0	2	1	0	0
2072	91	interpret metaphor	[Interpreting metaphors]	1.0	2	1	0	0
2073	91	lexical information	[lexical information]	1.0	2	2	2	2
2074	91	robust recognition of extra-grammatical input	[robust recognition of extra-grammatical input]	1.0	5	1	0	0
2075	91	novel likewise automatic evaluation method	[novel likewise automatic evaluation method]	1.0	5	1	0	0
2076	91	full meaning of the sentence	[full meaning of the sentence]	1.0	5	1	0	0
2077	91	implicit quantifier	[implicit quantifiers]	1.0	2	1	0	0
2078	91	syntactic ambiguity	[syntactic ambiguities]	1.0	2	1	0	0
2079	91	STTK a machine translation tool kit	[STTK a machine translation tool kit]	1.0	6	1	0	0
2080	91	standard atn	[standard ATN]	1.0	2	1	0	0
2081	91	problem of the automatic detection	[problem of the automatic detection]	1.0	5	1	0	0
2082	91	build a language model of move	[building a language model of moves]	1.0	6	1	0	0
2083	91	test corpus	[test corpus]	1.0	2	1	0	0
2084	91	target sentence	[target sentence]	1.0	2	1	0	0
2085	91	new type of transfer system	[new type of transfer system]	1.0	5	1	0	0
2086	91	restrictive statement	[restrictive statements]	1.0	2	1	0	0
2087	91	annealing approach	[annealing approach]	1.0	2	1	0	0
2088	91	key prediction	[key prediction]	1.0	2	2	1	1
2089	91	include part of speech tagging	[including part of speech tagging]	1.0	5	1	0	0
2090	91	error correction of ill-formed input	[error correction of ill-formed input]	1.0	5	1	0	0
2091	91	international collaboration	[international collaborations]	1.0	2	1	0	0
2092	91	sentence scoping	[sentence scoping]	1.0	2	1	0	0
2093	91	operational semantics	[operational semantics]	1.0	2	2	1	1
2094	91	computational analysis	[computational analysis]	1.0	2	1	0	0
2095	91	two approach	[two approaches]	1.0	2	1	0	0
2096	91	logical form	[logical form]	1.0	2	1	0	0
2097	91	use by language processing system	[use by Language Processing systems]	1.0	5	1	0	0
2098	91	multilingual datum	[multilingual data]	1.0	2	1	0	0
2099	91	result of a practical evaluation	[results of a practical evaluation]	1.0	5	1	0	0
2100	91	use LCS-Marine	[Using LCS-Marine]	1.0	2	1	0	0
2101	91	naturally sound dialog system within reach	[naturally sounding dialog systems within reach]	1.0	6	1	0	0
2102	91	current capability in dialogue system	[current capabilities in dialogue systems]	1.0	5	1	0	0
2103	91	entity-oriented parser	[entity-oriented parser]	1.0	2	1	0	0
2104	91	additional feature	[additional features]	1.0	2	2	1	1
2105	91	abductive inference	[abductive inference]	1.0	2	1	0	0
2106	91	88 .2%	[88 .2%]	1.0	2	1	0	0
2107	91	fragmentary recognition	[fragmentary recognition]	1.0	2	1	0	0
2108	91	prior in conditional loglinear model	[priors in conditional loglinear models]	1.0	5	1	0	0
2109	91	meaning of a proposition p	[meaning of a proposition p]	1.0	5	1	0	0
2110	91	7% improvement in recognition accuracy	[7% improvement in recognition accuracy]	1.0	5	1	0	0
2111	91	apply standard mt evaluation method (BLEU	[applying standard MT evaluation methods (BLEU]	1.0	6	1	0	0
2112	91	new method	[new method]	1.0	2	2	1	1
2113	91	discourse understanding	[discourse understanding]	1.0	2	2	2	2
2114	91	new technique	[new technique]	1.0	2	1	0	0
2115	91	training algorithm	[training algorithm]	1.0	2	1	0	0
2116	91	computation model	[computation model]	1.0	2	2	1	1
2117	91	pragmatics processing	[pragmatics processing]	1.0	2	1	0	0
2118	91	efficient algorithm	[efficient algorithm]	1.0	2	2	1	1
2119	91	light of various rhetorical function	[light of various rhetorical functions]	1.0	5	1	0	0
2120	91	problem automatically predict segment boundary	[problem automatically predicting segment boundaries]	1.0	5	1	0	0
2121	91	reestimation formulas	[reestimation formulas]	1.0	2	1	0	0
2122	91	40 utterance from the target speaker	[40 utterances from the target speaker]	1.0	6	1	0	0
2123	91	practical way deal with complexity	[practical ways dealing with complexity]	1.0	5	1	0	0
2124	91	shallow linguistic feature of question	[shallow linguistic features of questions]	1.0	5	1	0	0
2125	91	real world	[real world]	1.0	2	1	0	0
2126	91	result tagger	[resulting tagger]	1.0	2	1	0	0
2127	91	parse a set of phrase	[parsing a set of phrases]	1.0	5	1	0	0
2128	91	cooccurrence-based word	[cooccurrence-based word]	1.0	2	1	0	0
2129	91	calculus notation	[calculus notation]	1.0	2	2	2	2
2130	91	global scene	[global scene]	1.0	2	1	0	0
2131	91	also graph indispensable to paper	[also graphs indispensable to papers]	1.0	5	1	0	0
2132	91	limitation in comparison with wsd model	[limitations in comparison with WSD models]	1.0	6	1	0	0
2133	91	comma checker	[comma checker]	1.0	2	1	0	0
2134	91	parsing datum	[parsing data]	1.0	2	1	0	0
2135	91	information request	[information request]	1.0	2	1	0	0
2136	91	pattern in machine translation output	[patterns in machine translation output]	1.0	5	1	0	0
2137	91	first experiment in a series	[first experiment in a series]	1.0	5	1	0	0
2138	91	second requirement	[second requirement]	1.0	2	1	0	0
2139	91	practical technique	[practical techniques]	1.0	2	1	0	0
2140	91	evaluate a representative chinese-to-english SMT model	[evaluating a representative Chinese-to-English SMT model]	1.0	6	1	0	0
2141	91	nonverbal behavior	[nonverbal behaviors]	1.0	2	1	0	0
2142	91	semantic coherence	[semantic coherence]	1.0	2	1	0	0
2143	91	good guidance for novice user	[good guidance for novice users]	1.0	5	1	0	0
2144	91	cluster subcategorization frame SCF distribution	[clustering subcategorization frame SCF distributions]	1.0	5	1	0	0
2145	91	block reordering	[block reordering]	1.0	2	1	0	0
2146	91	good condition	[best condition]	1.0	2	1	0	0
2147	91	efficacy apply automated evaluation technique	[efficacy applying automated evaluation techniques]	1.0	5	1	0	0
2148	91	build a non-English Arabic stemmer	[building a non-English Arabic stemmer]	1.0	5	1	0	0
2149	91	defeat rule	[defeat rules]	1.0	2	1	0	0
2150	91	formal analysis	[formal analysis]	1.0	2	2	1	1
2151	91	tactical personnel	[tactical personnel]	1.0	2	1	0	0
2152	91	Byron's system	[Byron's system]	1.0	2	1	0	0
2153	91	accurate result	[accurate results]	1.0	2	1	0	0
2154	91	first utterance	[first utterance]	1.0	2	1	0	0
2155	91	several dimension	[several dimensions]	1.0	2	1	0	0
2156	91	parallel text	[parallel text]	1.0	2	1	0	0
2157	91	head nod	[head nods]	1.0	2	1	0	0
2158	91	similar approach	[similar approach]	1.0	2	1	0	0
2159	91	compete member	[competing members]	1.0	2	1	0	0
2160	91	overall pattern	[overall pattern]	1.0	2	1	0	0
2161	91	global meaning of a sentence	[global meaning of a sentence]	1.0	5	1	0	0
2162	91	text-to-speech form	[text-to-speech form]	1.0	2	1	0	0
2163	91	redundant copying	[redundant copying]	1.0	2	1	0	0
2164	91	human user	[human users]	1.0	2	1	0	0
2165	91	straight c	[straight C]	1.0	2	1	0	0
2166	91	decision how	[decision how]	1.0	2	1	0	0
2167	91	compound noun	[compound noun]	1.0	2	3	2	1
2168	91	manual transcription	[manual transcription]	1.0	2	2	2	2
2169	91	Shieber [11]	[Shieber [11]]	1.0	2	1	0	0
2170	91	commercial system output unsegmented text	[commercial systems outputting unsegmented texts]	1.0	5	1	0	0
2171	91	approximation error	[approximation error]	1.0	2	1	0	0
2172	91	accuracy rate of syntactic disambiguation	[accuracy rate of syntactic disambiguation]	1.0	5	1	0	0
2173	91	Word identification	[Word Identification]	1.0	2	1	0	0
2174	91	underlying theory	[underlying theory]	1.0	2	1	0	0
2175	91	error analysis	[Error analysis]	1.0	2	1	0	0
2176	91	one thing	[one thing]	1.0	2	1	0	0
2177	91	Arabic corpus	[Arabic corpus]	1.0	2	2	2	2
2178	91	basic idea of this method	[basic idea of this method]	1.0	5	1	0	0
2179	91	specific information	[specific information]	1.0	2	1	0	0
2180	91	category transition for speech repair	[category transitions for speech repairs]	1.0	5	1	0	0
2181	91	FlexP a bottom-up pattern-matching parser	[FlexP a bottom-up pattern-matching parser]	1.0	5	1	0	0
2182	91	basic mapping	[basic mapping]	1.0	2	1	0	0
2183	91	Kyoto city bus information system	[Kyoto city bus information system]	1.0	5	1	0	0
2184	91	even in a precise way	[even in a precise way]	1.0	5	1	0	0
2185	91	Listen-Communicate-Show LCS	[Listen-Communicate-Show LCS]	1.0	2	1	0	0
2186	91	human interaction	[human interaction]	1.0	2	1	0	0
2187	91	logical well-formedness condition on tree	[logical well-formedness conditions on trees]	1.0	5	1	0	0
2188	91	automatically evaluate answer to definition question	[automatically evaluating answers to definition questions]	1.0	6	1	0	0
2189	91	part of industry watch function	[part of industry watch function]	1.0	5	1	0	0
2190	91	new view for extensional reference	[new view for extensional reference]	1.0	5	1	0	0
2191	91	represent determiner	[representing determiners]	1.0	2	1	0	0
2192	91	fundamental concern	[fundamental concern]	1.0	2	1	0	0
2193	91	lexical information from the parse tree	[lexical information from the parse tree]	1.0	6	1	0	0
2194	91	initial rank	[initial ranking]	1.0	2	2	1	1
2195	91	complete lexical distribution of grammatical knowledge	[Complete lexical distribution of grammatical knowledge]	1.0	6	1	0	0
2196	91	robust mixed-initiative speech dialogue interaction	[robust mixed-initiative speech dialogue interactions]	1.0	5	1	0	0
2197	91	kernel function	[kernel function]	1.0	2	2	1	1
2198	91	new algorithm	[new algorithm]	1.0	2	2	2	2
2199	91	need for answer to question	[need for answers to questions]	1.0	5	1	0	0
2200	91	conventional treatment	[conventional treatments]	1.0	2	1	0	0
2201	91	work in a 'synchronous' way	[working in a 'synchronous' way]	1.0	5	1	0	0
2202	91	Chinese word	[Chinese word]	1.0	2	1	0	0
2203	91	human annotator	[human annotators]	1.0	2	1	0	0
2204	91	whole sentence	[whole sentence]	1.0	2	1	0	0
2205	91	discourse segment	[discourse segments]	1.0	2	1	0	0
2206	91	identical information	[identical information]	1.0	2	1	0	0
2207	91	cue phrase	[cue phrases]	1.0	2	2	2	2
2208	91	disambiguation algorithm	[disambiguation algorithms]	1.0	2	2	1	1
2209	91	increase worst-case parsing time complexity	[increasing worst-case parsing time complexity]	1.0	5	1	0	0
2210	91	discourse processing	[discourse processing]	1.0	2	2	1	1
2211	91	disambiguation process	[disambiguation process]	1.0	2	1	0	0
2212	91	English stemmer	[English stemmer]	1.0	2	1	0	0
2213	91	naturally aggregate	[naturally aggregate]	1.0	2	1	0	0
2214	91	phrase-based statistical machine translation method	[phrase-based statistical machine translation method]	1.0	5	1	0	0
2215	91	entity-oriented approach to restricted-domain parsing	[entity-oriented approach to restricted-domain parsing]	1.0	5	1	0	0
2216	91	four application	[four applications]	1.0	2	1	0	0
2217	91	expert translation	[expert translation]	1.0	2	2	2	2
2218	91	link detection	[Link detection]	1.0	2	1	0	0
2219	91	conventional method	[conventional methods]	1.0	2	1	0	0
2220	91	manual transcription of training datum	[manual transcription of training data]	1.0	5	1	0	0
2221	91	multiple candidate for the understanding result	[multiple candidates for the understanding result]	1.0	6	1	0	0
2222	91	presuppositional nature	[presuppositional nature]	1.0	2	1	0	0
2223	91	knowledge concern translation equivalence of expression	[knowledge concerning translation equivalence of expressions]	1.0	6	1	0	0
2224	91	similar result	[Similar results]	1.0	2	1	0	0
2225	91	around 97% exact match accuracy	[around 97% exact match accuracy]	1.0	5	1	0	0
2226	91	direction-givegivegivegivegivegivegivion- task	[direction-giving task]	1.0	2	1	0	0
2227	91	communicator participant	[Communicator participants]	1.0	2	1	0	0
2228	91	diverse problem	[diverse problems]	1.0	2	1	0	0
2229	91	little robustness	[little robustness]	1.0	2	1	0	0
2230	91	instead of the traditional practice	[instead of the traditional practice]	1.0	5	1	0	0
2231	91	large class	[large class]	1.0	2	1	0	0
2232	91	Dynamic grammar	[Dynamic Grammars]	1.0	2	1	0	0
2233	91	big corpus	[bigger corpus]	1.0	2	1	0	0
2234	91	computer program	[computer program]	1.0	2	2	2	2
2235	91	separate module	[separate modules]	1.0	2	1	0	0
2236	91	unsupervised component	[unsupervised component]	1.0	2	1	0	0
2237	91	little speech	[little speech]	1.0	2	1	0	0
2238	91	similar fashion	[similar fashion]	1.0	2	1	0	0
2239	91	utility apply mt evaluation method	[utility applying MT evaluation methods]	1.0	5	1	0	0
2240	91	newspaper story	[newspaper stories]	1.0	2	1	0	0
2241	91	lexical-cohesion feature	[lexical-cohesion features]	1.0	2	1	0	0
2242	91	syntax-based statistical machine translation system	[syntax-based statistical machine translation system]	1.0	5	1	0	0
2243	91	CFG filter technique for LTAG	[CFG filtering techniques for LTAG]	1.0	5	1	0	0
2244	91	output quality	[output quality]	1.0	2	1	0	0
2245	91	end-to-end process	[end-to-end process]	1.0	2	1	0	0
2246	91	resource-frugal approach	[resource-frugal approach]	1.0	2	1	0	0
2247	91	everyday rejoinder	[everyday rejoinders]	1.0	2	1	0	0
2248	91	discourse structure	[discourse structure]	1.0	2	2	1	1
2249	91	important role convey the meaning	[important role conveying the meaning]	1.0	5	1	0	0
2250	91	method attain a design overhead	[method attaining a design overheads]	1.0	5	1	0	0
2251	91	program RINA	[program RINA]	1.0	2	1	0	0
2252	91	telephone dialogue	[telephone dialogue]	1.0	2	1	0	0
2253	91	scruffy text in the domain	[scruffy texts in the domain]	1.0	5	1	0	0
2254	91	exploit a large non-parallel corpus	[exploiting a large non-parallel corpus]	1.0	5	1	0	0
2255	91	monolingual corpus	[monolingual corpus]	1.0	2	2	2	2
2256	91	PageRank algorithm	[PageRank algorithm]	1.0	2	1	0	0
2257	91	possibility use the language model	[possibility using the language model]	1.0	5	1	0	0
2258	91	referent anaphora	[referents anaphora]	1.0	2	1	0	0
2259	91	potential antecedence	[potential antecedence]	1.0	2	1	0	0
2260	91	syntactic expectation	[syntactic expectations]	1.0	2	1	0	0
2261	91	concrete example	[concrete examples]	1.0	2	1	0	0
2262	91	two experiment	[two experiments]	1.0	2	1	0	0
2263	91	track the discourse through the operation	[tracking the discourse through the operation]	1.0	6	1	0	0
2264	91	boosting method	[boosting method]	1.0	2	1	0	0
2265	91	robust parser	[robust parser]	1.0	2	1	0	0
2266	91	include the influence of training factor	[including the influence of training factors]	1.0	6	1	0	0
2267	91	semantic category	[semantic categories]	1.0	2	1	0	0
2268	91	state-of-the-art question answering qa system	[State-of-the-art Question Answering QA systems]	1.0	5	1	0	0
2269	91	natural language text email .g	[natural language texts e .g]	1.0	5	1	0	0
2270	91	maximum entropy word alignment algorithm	[maximum entropy word alignment algorithm]	1.0	5	1	0	0
2271	91	February 1992	[February 1992]	1.0	2	1	0	0
2272	91	automatically train spot on the basis	[automatically training SPoT on the basis]	1.0	6	1	0	0
2273	91	13% decrease in f-measure error	[13% decrease in F-measure error]	1.0	5	1	0	0
2274	91	previous study	[previous studies]	1.0	2	1	0	0
2275	91	to reasoning	[to reasoning]	1.0	2	1	0	0
2276	91	SI result	[SI result]	1.0	2	1	0	0
2277	91	conversational context	[conversational context]	1.0	2	1	0	0
2278	91	speaker adaptation	[speaker adaptation]	1.0	2	2	1	1
2279	91	problem of vagueness in annotation	[problems of vagueness in annotation]	1.0	5	1	0	0
2280	91	two style	[Two styles]	1.0	2	2	1	1
2281	91	erroneous sentence	[erroneous sentence]	1.0	2	1	0	0
2282	91	document processing	[document processing]	1.0	2	1	0	0
2283	91	instantiation of the one sense	[instantiation of the one sense]	1.0	5	1	0	0
2284	91	automatic evaluation	[automatic evaluation]	1.0	2	2	1	1
2285	91	correctness of answer to question	[correctness of answers to questions]	1.0	5	1	0	0
2286	91	Prolog form	[Prolog form]	1.0	2	1	0	0
2287	91	Arabic word	[Arabic word]	1.0	2	2	2	2
2288	91	statistical class	[statistical class]	1.0	2	1	0	0
2289	91	IBM model	[IBM Models]	1.0	2	2	1	1
2290	91	trainable component	[trainable components]	1.0	2	1	0	0
2291	91	97 .24%	[97 .24%]	1.0	2	1	0	0
2292	91	understand system with intelligent mobile agent	[understanding system with intelligent mobile agents]	1.0	6	1	0	0
2293	91	English question	[English questions]	1.0	2	1	0	0
2294	91	adopt a simple mt-based paraphrasing technique	[adopting a simple MT-based paraphrasing technique]	1.0	6	1	0	0
2295	91	use a big a corpus	[using a bigger a corpus]	1.0	5	1	0	0
2296	91	boolean condition on the cooccurrence	[Boolean conditions on the cooccurrence]	1.0	5	1	0	0
2297	91	experiment result on ACE corpus	[Experiment results on ACE corpora]	1.0	5	1	0	0
2298	91	source of training datum suitable	[Sources of training data suitable]	1.0	5	1	0	0
2299	91	heuristic rule	[heuristic rules]	1.0	2	1	0	0
2300	91	browse the repository of datum object	[browsing the repository of data objects]	1.0	6	1	0	0
2301	91	real-world environment	[real-world environments]	1.0	2	1	0	0
2302	91	error in the machine translation	[errors in the machine translations]	1.0	5	1	0	0
2303	91	maximum-entropy model	[maximum-entropy model]	1.0	2	2	2	2
2304	91	2 inductive decision tree learning	[2 inductive decision tree learning]	1.0	5	1	0	0
2305	91	kind of similarity between word	[kind of similarity between words]	1.0	5	1	0	0
2306	91	parsing-as-deduction in a resource logic	[parsing-as-deduction in a resource logic]	1.0	5	1	0	0
2307	91	Japanese-English translation	[Japanese-English translation]	1.0	2	1	0	0
2308	91	extensional reference	[extensional reference]	1.0	2	1	0	0
2309	91	semantically coherent speech recognition hypothese	[semantically coherent speech recognition hypotheses]	1.0	5	1	0	0
2310	91	paraphrase probability	[paraphrase probability]	1.0	2	1	0	0
2311	91	trigram language	[trigram language]	1.0	2	2	2	2
2312	91	minimal DFA	[minimal DFA]	1.0	2	1	0	0
2313	91	graphical model	[graphical model]	1.0	2	1	0	0
2314	91	fuzzy quantifier	[fuzzy quantifier]	1.0	2	2	2	2
2315	91	dominance graph	[dominance graphs]	1.0	2	1	0	0
2316	91	language modeling of conversational speech	[language modeling of conversational speech]	1.0	5	1	0	0
2317	91	grammatical formalism	[grammatical formalisms]	1.0	2	2	2	2
2318	91	current sentence	[current sentence]	1.0	2	1	0	0
2319	91	12,000 utterance	[12,000 utterances]	1.0	2	1	0	0
2320	91	relationship among the target variable	[relationships among the target variables]	1.0	5	1	0	0
2321	91	relevant approach	[relevant approach]	1.0	2	1	0	0
2322	91	search for answer in multiple corpus	[searching for answers in multiple corpora]	1.0	6	1	0	0
2323	91	language presupposition	[language presuppositions]	1.0	2	2	2	2
2324	91	first learner	[first learner]	1.0	2	1	0	0
2325	91	edit operation	[edit operation]	1.0	2	1	0	0
2326	91	stem model	[stemming model]	1.0	2	1	0	0
2327	91	logical formulas	[logical formulas]	1.0	2	1	0	0
2328	91	less extent	[lesser extent]	1.0	2	1	0	0
2329	91	certainty condition	[certainty conditions]	1.0	2	1	0	0
2330	91	theoretical work	[theoretical work]	1.0	2	1	0	0
2331	91	new measure	[new measure]	1.0	2	1	0	0
2332	91	exhaustive comparison	[exhaustive comparison]	1.0	2	1	0	0
2333	91	great benefit to language pair	[great benefit to language pairs]	1.0	5	1	0	0
2334	91	theoretical literature	[theoretical literature]	1.0	2	1	0	0
2335	91	English coordinate structure analysis model	[English coordinate structure analysis model]	1.0	5	1	0	0
2336	91	short time	[short time]	1.0	2	1	0	0
2337	91	recent work	[recent work]	1.0	2	2	2	2
2338	91	useful feature for paraphrase classification	[useful features for paraphrase classification]	1.0	5	1	0	0
2339	91	vagueness of the spatial concept	[vagueness of the spatial concepts]	1.0	5	1	0	0
2340	91	polynomial time solution for any	[polynomial time solution for any]	1.0	5	1	0	0
2341	91	many speaker	[many speakers]	1.0	2	2	2	2
2342	91	interactively visualize	[interactively visualizing]	1.0	2	1	0	0
2343	91	early copying	[early copying]	1.0	2	1	0	0
2344	91	case-based reasoning in machine translation	[case-based reasoning in machine translation]	1.0	5	1	0	0
2345	91	sensitive logic	[sensitive logic]	1.0	2	1	0	0
2346	91	blackboard-like architecture	[blackboard-like architecture]	1.0	2	1	0	0
2347	91	neat text	[neat texts]	1.0	2	1	0	0
2348	91	paraphrase classification	[paraphrase classification]	1.0	2	2	2	2
2349	91	thereby reduce the amount of ambiguity	[thereby reducing the amount of ambiguity]	1.0	6	1	0	0
2350	91	word-level alignment	[word-level alignment]	1.0	2	1	0	0
2351	91	spectral approach	[spectral approach]	1.0	2	1	0	0
2352	91	out-of-domain corpus	[out-of-domain corpus]	1.0	2	2	1	1
2353	91	parsing technique	[parsing technique]	1.0	2	1	0	0
2354	91	f-measure error	[F-measure error]	1.0	2	1	0	0
2355	91	/soames 1982/	[/Soames 1982/]	1.0	2	1	0	0
2356	91	training example	[training examples]	1.0	2	1	0	0
2357	91	overview of the CSR corpus	[overview of the CSR corpus]	1.0	5	1	0	0
2358	91	analysis step	[analysis step]	1.0	2	1	0	0
2359	91	czech-english statistical machine translation system	[Czech-English statistical machine translation system]	1.0	5	1	0	0
2360	91	broad domain	[broad domains]	1.0	2	1	0	0
2361	91	image understanding with text understanding	[image understanding with text understanding]	1.0	5	1	0	0
2362	91	advantage of the three strategy	[advantages of the three strategies]	1.0	5	1	0	0
2363	91	textual representation	[textual representation]	1.0	2	1	0	0
2364	91	translingual reach into other language	[translingual reach into other languages]	1.0	5	1	0	0
2365	91	verb group	[verb group]	1.0	2	1	0	0
2366	91	fake symbol	[fake symbol]	1.0	2	1	0	0
2367	91	unification algorithm avoidance of copying	[unification algorithms avoidance of copying]	1.0	5	1	0	0
2368	91	several voting-	[several voting-]	1.0	2	1	0	0
2369	91	translation into a equivalent RCG	[translation into an equivalent RCG]	1.0	5	1	0	0
2370	91	storage media	[storage media]	1.0	2	1	0	0
2371	91	binary feature	[binary features]	1.0	2	1	0	0
2372	91	large-scale grammar	[large-scale grammars]	1.0	2	1	0	0
2373	91	use m-th order Markov chain model	[using m-th order Markov chain model]	1.0	6	1	0	0
2374	91	combination with a clustering process	[combination with a clustering process]	1.0	5	1	0	0
2375	91	substitution cost	[substitution costs]	1.0	2	2	2	2
2376	91	same belief	[same beliefs]	1.0	2	1	0	0
2377	91	monitoring of lack of feedback	[monitoring of lack of feedback]	1.0	5	1	0	0
2378	91	Schutze idea of evaluation algorithm	[Schutze's idea of evaluation algorithms]	1.0	5	1	0	0
2379	91	single system	[single system]	1.0	2	1	0	0
2380	91	sentence-level correlation	[sentence-level correlation]	1.0	2	1	0	0
2381	91	unsegmented corpus	[unsegmented corpus]	1.0	2	2	1	1
2382	91	add linguistic annotation to corpus	[adding linguistic annotation to corpora]	1.0	5	1	0	0
2383	91	grolier' Encyclopedia	[Grolier's Encyclopedia]	1.0	2	1	0	0
2384	91	other language	[other languages]	1.0	2	2	2	2
2385	91	part-of-speech-based criterion	[part-of-speech-based criterion]	1.0	2	1	0	0
2386	91	target document	[target document]	1.0	2	1	0	0
2387	91	domain-independent lexica	[domain-independent lexica]	1.0	2	1	0	0
2388	91	benchmark system	[benchmark system]	1.0	2	1	0	0
2389	91	image understanding	[image understanding]	1.0	2	2	1	1
2390	91	error reduction of 4 .4%	[error reduction of 4 .4%]	1.0	5	1	0	0
2391	91	valuable lexicon	[valuable lexicons]	1.0	2	1	0	0
2392	91	human interaction with datum source	[human interaction with data sources]	1.0	5	1	0	0
2393	91	use command	[using commands]	1.0	2	1	0	0
2394	91	other tool	[other tools]	1.0	2	1	0	0
2395	91	top-down derivation	[top-down derivation]	1.0	2	1	0	0
2396	91	evaluation technique	[evaluation techniques]	1.0	2	2	1	1
2397	91	challenge integrate automatic q/a application	[challenges integrating automatic Q/A applications]	1.0	5	1	0	0
2398	91	describe a case study avoid failure	[describing a case study avoiding failures]	1.0	6	1	0	0
2399	91	reference set	[reference sets]	1.0	2	2	1	1
2400	91	conditional model	[conditional models]	1.0	2	1	0	0
2401	91	sentence planning	[Sentence planning]	1.0	2	1	0	0
2402	91	language understanding work at Paramax	[Language understanding work at Paramax]	1.0	5	1	0	0
2403	91	more complex mixture of technique	[more complex mixtures of techniques]	1.0	5	1	0	0
2404	91	multiple translation	[multiple translations]	1.0	2	1	0	0
2405	91	use of various lexical feature	[use of various lexical features]	1.0	5	1	0	0
2406	91	translation relation	[translation relation]	1.0	2	2	1	1
2407	91	translation tool	[translation tool]	1.0	2	1	0	0
2408	91	negative feedback	[negative feedback]	1.0	2	1	0	0
2409	91	tree-to-tree translation of dependency structure	[tree-to-tree translation of dependency structures]	1.0	5	1	0	0
2410	91	particular domain	[particular domain]	1.0	2	1	0	0
2411	91	dialogue acquisition	[dialogue acquisition]	1.0	2	1	0	0
2412	91	investigate the combinatorics of free indexation	[investigating the combinatorics of free indexation]	1.0	6	1	0	0
2413	91	final language with overt case marker	[final language with overt case markers]	1.0	6	1	0	0
2414	91	several respects	[several respects]	1.0	2	1	0	0
2415	91	elastic constraint	[elastic constraints]	1.0	2	1	0	0
2416	91	several extension of this basic idea	[Several extensions of this basic idea]	1.0	6	1	0	0
2417	91	missing period	[missing periods]	1.0	2	1	0	0
2418	91	polysemous word	[polysemous word]	1.0	2	2	1	1
2419	91	role of purpose in discourse	[role of purpose in discourse]	1.0	5	1	0	0
2420	91	specific domain of HK law	[specific domain of HK laws]	1.0	5	1	0	0
2421	91	semantic network	[semantic network]	1.0	2	2	1	1
2422	91	practical efficiency of RCL parser	[practical efficiency of RCL parsers]	1.0	5	1	0	0
2423	91	correct parse of a sentence	[correct parse of a sentence]	1.0	5	1	0	0
2424	91	classification accuracy	[classification accuracy]	1.0	2	2	2	2
2425	91	indexing accord to simple character bigram	[indexing according to simple character bigrams]	1.0	6	1	0	0
2426	91	many language	[many languages]	1.0	2	1	0	0
2427	91	pilot system	[pilot system]	1.0	2	1	0	0
2428	91	b perspective-taking	[b perspective-taking]	1.0	2	1	0	0
2429	91	useful insight	[useful insights]	1.0	2	1	0	0
2430	91	new edge	[new edge]	1.0	2	1	0	0
2431	91	parse the Wall Street Journal treebank	[parsing the Wall Street Journal treebank]	1.0	6	1	0	0
2432	91	integrate language understanding with speech recognition	[integrating language understanding with speech recognition]	1.0	6	1	0	0
2433	91	strongly biase parsing toward meaning	[strongly biasing parsing toward meanings]	1.0	5	1	0	0
2434	91	study on communication between APs	[study on communications between APs]	1.0	5	1	0	0
2435	91	string with the good performance (typically	[string with the best performance (typically]	1.0	6	1	0	0
2436	91	pick the good hypothesis corresponding	[picking the best hypothesis corresponding]	1.0	5	1	0	0
2437	91	isomorphic Grammar for Machine Translation	[Isomorphic Grammars for Machine Translation]	1.0	5	1	0	0
2438	91	means of a beam-search decoder	[means of a beam-search decoder]	1.0	5	1	0	0
2439	91	use a set of word alignment	[using a set of word alignments]	1.0	6	1	0	0
2440	91	French sentence	[French sentence]	1.0	2	1	0	0
2441	91	prototype system	[prototype system]	1.0	2	1	0	0
2442	91	young man	[young men]	1.0	2	2	2	2
2443	91	bilingual material the Canadian Hansards	[bilingual material the Canadian Hansards]	1.0	5	1	0	0
2444	91	affix list	[affix lists]	1.0	2	1	0	0
2445	91	convey the meaning of a utterance	[conveying the meaning of an utterance]	1.0	6	1	0	0
2446	91	exploratory corpus analysis accessible to researcher	[exploratory corpus analysis accessible to researchers]	1.0	6	1	0	0
2447	91	little inflection	[little inflection]	1.0	2	1	0	0
2448	91	intensional concept	[intensional concepts]	1.0	2	1	0	0
2449	91	event	[event]	1.0	1	11	4	4
2450	91	English grammar	[English grammar]	1.0	2	1	0	0
2451	91	6 .5%	[6 .5%]	1.0	2	1	0	0
2452	91	reference model	[reference model]	1.0	2	1	0	0
2453	91	more robust---an agenda-based scheduling parser	[more robust---an agenda-based scheduling parser]	1.0	5	1	0	0
2454	91	show the wsd accuracy significantly low	[showing the WSD accuracy significantly lower]	1.0	6	1	0	0
2455	91	computer system	[computer system]	1.0	2	2	1	1
2456	91	little inflection such as English	[little inflection such as English]	1.0	5	1	0	0
2457	91	model TDMT Transfer-Driven Machine Translation	[model TDMT Transfer-Driven Machine Translation]	1.0	5	1	0	0
2458	91	speech information of the word	[speech information of the words]	1.0	5	1	0	0
2459	91	such mistake	[Such mistakes]	1.0	2	1	0	0
2460	91	experience plan COMPLEX a computational lexicon	[experience planning COMPLEX a computational lexicon]	1.0	6	1	0	0
2461	91	such system	[such systems]	1.0	2	1	0	0
2462	91	identify a predicate with honorifics	[identifying a predicate with honorifics]	1.0	5	1	0	0
2463	91	output summary	[output summary]	1.0	2	1	0	0
2464	91	learning approach	[learning approach]	1.0	2	2	2	2
2465	91	content-independent formalism	[content-independent formalism]	1.0	2	1	0	0
2466	91	lr parsing	[LR parsing]	1.0	2	1	0	0
2467	91	pre-tagged corpus	[pre-tagged corpus]	1.0	2	1	0	0
2468	91	two time	[two times]	1.0	2	1	0	0
2469	91	cyclic structure	[cyclic structures]	1.0	2	1	0	0
2470	91	expression of natural language to semantics	[expressions of natural languages to semantics]	1.0	6	1	0	0
2471	91	pronoun resolution	[pronoun resolution]	1.0	2	2	2	2
2472	91	ensemble method in machine learning	[ensemble methods in machine learning]	1.0	5	1	0	0
2473	91	test-score semantics	[test-score semantics]	1.0	2	1	0	0
2474	91	high model	[higher models]	1.0	2	1	0	0
2475	91	typical usage	[typical usage]	1.0	2	1	0	0
2476	91	developer of machine translation system	[developers of machine translation systems]	1.0	5	1	0	0
2477	91	Gaussian mixture	[Gaussian mixture]	1.0	2	1	0	0
2478	91	same time	[same time]	1.0	2	1	0	0
2479	91	WSD accuracy of SMT model	[WSD accuracy of SMT models]	1.0	5	1	0	0
2480	91	little difficulty	[little difficulty]	1.0	2	1	0	0
2481	91	objective machine translation evaluation measure	[objective machine translation evaluation measures]	1.0	5	1	0	0
2482	91	linguistic domain	[linguistic domain]	1.0	2	1	0	0
2483	91	increase the dialogue duration for user	[increasing the dialogue duration for users]	1.0	6	1	0	0
2484	91	good performance	[good performance]	1.0	2	2	1	1
2485	91	Semantic information	[Semantic information]	1.0	2	1	0	0
2486	91	synchronous grammar	[synchronous grammars]	1.0	2	1	0	0
2487	91	source code	[source code]	1.0	2	1	0	0
2488	91	implication for current discourse processing algorithm	[implications for current discourse processing algorithms]	1.0	6	1	0	0
2489	91	three type of the error	[three types of the errors]	1.0	5	1	0	0
2490	91	lexical hierarchy	[lexical hierarchy]	1.0	2	2	1	1
2491	91	obviate the need for answer	[obviating the need for answers]	1.0	5	1	0	0
2492	91	yield a treatment of Montague semantics	[yielding a treatment of Montague semantics]	1.0	6	1	0	0
2493	91	BLEU at the character level	[BLEU at the character level]	1.0	5	1	0	0
2494	91	rank method	[ranking methods]	1.0	2	1	0	0
2495	91	BLEU score	[BLEU score]	1.0	2	2	1	1
2496	91	algorithmic efficiency	[algorithmic efficiency]	1.0	2	1	0	0
2497	91	build spelling-checker for such language	[building spelling-checkers for such languages]	1.0	5	1	0	0
2498	91	speed of the result program	[speed of the resulting program]	1.0	5	1	0	0
2499	91	RCL parser	[RCL parser]	1.0	2	2	2	2
2500	91	large database	[large database]	1.0	2	3	2	1
2501	91	program APT	[program APT]	1.0	2	1	0	0
2502	91	cooperative response	[cooperative responses]	1.0	2	2	2	2
2503	91	unigram count	[unigram counts]	1.0	2	1	0	0
2504	91	lexical paraphrase	[lexical paraphrases]	1.0	2	1	0	0
2505	91	language acquisition	[language acquisition]	1.0	2	1	0	0
2506	91	significant improvement over the WI system	[significant improvements over the WI systems]	1.0	6	1	0	0
2507	91	non-overlapping interval	[non-overlapping intervals]	1.0	2	1	0	0
2508	91	syntactic paraphrase	[syntactic paraphrases]	1.0	2	1	0	0
2509	91	extra-grammatical input	[extra-grammatical input]	1.0	2	1	0	0
2510	91	parser evaluation	[parser evaluation]	1.0	2	1	0	0
2511	91	elementary structure	[elementary structures]	1.0	2	1	0	0
2512	91	phrase length	[phrase length]	1.0	2	1	0	0
2513	91	temporary belief about the world	[temporary belief about the world]	1.0	5	1	0	0
2514	91	saturation of the final structure	[saturation of the final structure]	1.0	5	1	0	0
2515	91	apparent tendency	[apparent tendency]	1.0	2	1	0	0
2516	91	Stabler' formalization of Chomsky minimalist program	[Stabler's formalization of Chomsky's minimalist program]	1.0	6	1	0	0
2517	91	1 a set of feature	[1 an set of features]	1.0	5	1	0	0
2518	91	multi-site evaluation	[multi-site evaluation]	1.0	2	1	0	0
2519	91	similar performance	[similar performance]	1.0	2	1	0	0
2520	91	asr output	[ASR output]	1.0	2	2	2	2
2521	91	computational model of the process	[computational model of the process]	1.0	5	1	0	0
2522	91	content understanding of the document	[content understanding of the document]	1.0	5	1	0	0
2523	91	set of manual word alignment	[set of manual word alignments]	1.0	5	1	0	0
2524	91	training method	[training method]	1.0	2	2	1	1
2525	91	new methodology automatically train spot	[new methodology automatically training SPoT]	1.0	5	1	0	0
2526	91	metaphor interpretation	[metaphor interpretation]	1.0	2	1	0	0
2527	91	stereotypic dialogue	[stereotypic dialogue]	1.0	2	1	0	0
2528	91	October '91	[October '91]	1.0	2	2	1	1
2529	91	information need	[information need]	1.0	2	1	0	0
2530	91	directly on word sense disambiguation performance	[directly on word sense disambiguation performance]	1.0	6	1	0	0
2531	91	fundamental problem in user modeling	[fundamental problem in user modeling]	1.0	5	1	0	0
2532	91	novel way	[novel way]	1.0	2	1	0	0
2533	91	Japanese bunsetsu	[Japanese bunsetsu]	1.0	2	2	1	1
2534	91	outstanding problem with the approach	[outstanding problems with the approach]	1.0	5	1	0	0
2535	91	10k sentence	[10K sentences]	1.0	2	1	0	0
2536	91	DARPA speech recognition technology development	[DARPA speech recognition technology development]	1.0	5	1	0	0
2537	91	attentional focus	[attentional focus]	1.0	2	1	0	0
2538	91	evaluation metric	[evaluation metric]	1.0	2	2	2	2
2539	91	important role in the determination	[important role in the determination]	1.0	5	1	0	0
2540	91	multiple way	[multiple ways]	1.0	2	1	0	0
2541	91	practical evaluation	[practical evaluation]	1.0	2	1	0	0
2542	91	research program	[research program]	1.0	2	1	0	0
2543	91	obvious implementation	[obvious implementation]	1.0	2	1	0	0
2544	91	cull information from a large inflow	[culling information from a large inflow]	1.0	6	1	0	0
2545	91	top-ranked plan	[top-ranked plan]	1.0	2	1	0	0
2546	91	EBMT system	[EBMT system]	1.0	2	1	0	0
2547	91	overall summarization quality of the system	[Overall summarization quality of the system]	1.0	6	1	0	0
2548	91	speaker's utterance	[speaker's utterance]	1.0	2	1	0	0
2549	91	SUMMIT recognizer	[SUMMIT recognizer]	1.0	2	1	0	0
2550	91	young man like young woman	[young men like young women]	1.0	5	2	1	1
2551	91	correct candidate for the error	[correct candidates for the errors]	1.0	5	1	0	0
2552	91	dialogue strategy	[Dialogue strategies]	1.0	2	1	0	0
2553	91	linguistic representation	[linguistic representation]	1.0	2	2	1	1
2554	91	search space	[search space]	1.0	2	2	2	2
2555	91	parsing complexity	[parsing complexity]	1.0	2	1	0	0
2556	91	people' answer to everyday question	[people's answers to everyday questions]	1.0	5	1	0	0
2557	91	system froff	[system FROFF]	1.0	2	1	0	0
2558	91	multi engine machine translation system	[multi engine machine translation system]	1.0	5	1	0	0
2559	91	everyday question	[everyday questions]	1.0	2	1	0	0
2560	91	log-linear block	[log-linear block]	1.0	2	1	0	0
2561	91	target variable	[target variables]	1.0	2	2	1	1
2562	91	feature selection method within model	[feature selection methods within models]	1.0	5	1	0	0
2563	91	lexical generalization	[lexical generalizations]	1.0	2	1	0	0
2564	91	systematic pattern	[systematic patterns]	1.0	2	1	0	0
2565	91	linguistic segment	[linguistic segments]	1.0	2	1	0	0
2566	91	rule-reduction algorithm apply mutual information	[rule-reduction algorithm applying mutual information]	1.0	5	1	0	0
2567	91	theoretical account	[theoretical accounts]	1.0	2	1	0	0
2568	91	Korean newspaper article on missile	[Korean newspaper articles on missiles]	1.0	5	1	0	0
2569	91	DoPS system	[DoPS system]	1.0	2	1	0	0
2570	91	idus intelligent document Understanding System	[IDUS Intelligent Document Understanding System]	1.0	5	1	0	0
2571	91	segment of actual tape-recorded description	[segments of actual tape-recorded descriptions]	1.0	5	1	0	0
2572	91	sentence-plan-ranker SPR	[sentence-plan-ranker SPR]	1.0	2	1	0	0
2573	91	component performance	[component performance]	1.0	2	1	0	0
2574	91	use standard wsd evaluation methodology	[using standard WSD evaluation methodology]	1.0	5	1	0	0
2575	91	routine basis	[routine basis]	1.0	2	1	0	0
2576	91	many theory of anaphora resolution	[many theories of anaphora resolution]	1.0	5	1	0	0
2577	91	classification accuracy of the method	[classification accuracy of the method]	1.0	5	1	0	0
2578	91	still satisfy human communication need	[still satisfying human communication needs]	1.0	5	1	0	0
2579	91	annotate honorifics	[Annotating honorifics]	1.0	2	1	0	0
2580	91	formalisme usage	[formalism's usage]	1.0	2	1	0	0
2581	91	cluster number estimation on the eigenvector	[cluster number estimation on the eigenvectors]	1.0	6	1	0	0
2582	91	basic idea	[basic idea]	1.0	2	2	2	2
2583	91	subparagraph level	[subparagraph level]	1.0	2	1	0	0
2584	91	identify unknown word in Chinese newspaper	[identifying unknown words in Chinese newspapers]	1.0	6	1	0	0
2585	91	usual approach	[usual approach]	1.0	2	1	0	0
2586	91	language identification	[language identification]	1.0	2	2	2	2
2587	91	provide a interpretation for value	[providing an interpretation for values]	1.0	5	1	0	0
2588	91	text browser	[text browser]	1.0	2	2	2	2
2589	91	cooccurrence statistic	[cooccurrence statistics]	1.0	2	1	0	0
2590	91	elaborate document	[elaborate documents]	1.0	2	1	0	0
2591	91	reproducability of a gold standard	[reproducability of a gold standard]	1.0	5	1	0	0
2592	91	problem enumerate all possible indexing	[problem enumerating all possible indexings]	1.0	5	1	0	0
2593	91	declarative way	[declarative way]	1.0	2	1	0	0
2594	91	unsupervised wsd	[unsupervised WSD]	1.0	2	1	0	0
2595	91	use a phrase in another language	[using a phrase in another language]	1.0	6	1	0	0
2596	91	constraint-based parser/generator	[constraint-based parser/generator]	1.0	2	1	0	0
2597	91	monolingual UCG	[monolingual UCG]	1.0	2	1	0	0
2598	91	system' output	[system's output]	1.0	2	1	0	0
2599	91	Japanese text	[Japanese texts]	1.0	2	1	0	0
2600	91	syntax proper	[syntax proper]	1.0	2	1	0	0
2601	91	link decision	[link decisions]	1.0	2	1	0	0
2602	91	own width	[own width]	1.0	2	1	0	0
2603	91	original model	[original model]	1.0	2	1	0	0
2604	91	previous research	[Previous research]	1.0	2	1	0	0
2605	91	segmentation accuracy	[segmentation accuracy]	1.0	2	1	0	0
2606	91	workability of the general mechanism	[workability of the general mechanism]	1.0	5	1	0	0
2607	91	people' answer	[people's answers]	1.0	2	2	2	2
2608	91	individual phrase	[individual phrases]	1.0	2	1	0	0
2609	91	dependency structure	[dependency structure]	1.0	2	3	2	1
2610	91	inheritance mechanism	[inheritance mechanisms]	1.0	2	1	0	0
2611	92	constraint	[constraint]	0.9857142857142858	1	11	8	7
2612	93	sense	[sense]	0.9833333333333334	1	11	7	6
2613	94	segment	[segment]	0.9666666666666667	1	11	4	3
2614	94	tree	[tree]	0.9666666666666667	1	11	4	3
2615	95	role	[role]	0.9571428571428572	1	11	10	7
2616	96	number	[number]	0.9500000000000001	1	12	10	4
2617	97	alignment	[alignment]	0.9	1	10	6	6
2618	97	dialog	[dialog]	0.9	1	10	4	4
2619	97	source	[source]	0.9	1	10	3	3
2620	97	character	[character]	0.9	1	10	3	3
2621	97	include	[including]	0.9	1	10	4	4
2622	98	expression	[expression]	0.8800000000000001	1	10	6	5
2623	99	combination	[combination]	0.8500000000000001	1	10	9	6
2624	100	advantage	[advantage]	0.8400000000000001	1	10	8	5
2625	100	candidate	[candidate]	0.8400000000000001	1	10	8	5
2626	101	semantics	[semantics]	0.8	1	9	6	6
2627	101	rank	[ranking]	0.8	1	9	8	8
2628	101	purpose	[purpose]	0.8	1	9	4	4
2629	101	recognition	[recognition]	0.8	1	9	4	4
2630	101	tool	[tool]	0.8	1	9	7	7
2631	101	accuracy	[accuracy]	0.8	1	9	5	5
2632	101	pattern	[pattern]	0.8	1	9	7	7
2633	102	formalism	[formalism]	0.7833333333333333	1	9	7	6
2634	102	database	[database]	0.7833333333333333	1	9	7	6
2635	103	basis	[basis]	0.75	1	9	6	4
2636	103	kind	[kind]	0.75	1	9	9	6
2637	104	block	[block]	0.7000000000000001	1	8	4	4
2638	104	equivalence	[equivalence]	0.7000000000000001	1	8	4	4
2639	104	form	[form]	0.7000000000000001	1	8	5	5
2640	104	wsd	[WSD]	0.7000000000000001	1	8	5	5
2641	104	rule	[rules]	0.7000000000000001	1	8	6	6
2642	104	interaction	[interaction]	0.7000000000000001	1	8	6	6
2643	104	nlp	[NLP]	0.7000000000000001	1	8	2	2
2644	104	edge	[edge]	0.7000000000000001	1	8	7	7
2645	104	resource	[resource]	0.7000000000000001	1	8	6	6
2646	104	process	[process]	0.7000000000000001	1	8	5	5
2647	104	idea	[idea]	0.7000000000000001	1	8	6	6
2648	104	baseline	[baseline]	0.7000000000000001	1	8	2	2
2649	105	strategy	[strategy]	0.6857142857142858	1	8	8	7
2650	106	understanding	[understanding]	0.675	1	8	5	4
2651	106	communication	[communication]	0.675	1	8	5	4
2652	107	similarity	[similarity]	0.6666666666666667	1	8	8	6
2653	107	description	[description]	0.6666666666666667	1	8	8	6
2654	107	tag	[tag]	0.6666666666666667	1	8	4	3
2655	108	state	[state]	0.6400000000000001	1	8	8	5
2656	109	quality	[quality]	0.625	1	8	7	4
2657	110	term	[term]	0.6000000000000001	1	8	6	3
2658	110	graph	[Graph]	0.6000000000000001	1	7	3	3
2659	110	computation	[computation]	0.6000000000000001	1	7	3	3
2660	110	need	[need]	0.6000000000000001	1	7	5	5
2661	110	addition	[addition]	0.6000000000000001	1	7	2	2
2662	110	email	[e]	0.6000000000000001	1	7	2	2
2663	110	implementation	[implementation]	0.6000000000000001	1	7	3	3
2664	110	framework	[framework]	0.6000000000000001	1	7	5	5
2665	110	NIST	[NIST]	0.6000000000000001	1	6	0	0
2666	110	mixture	[mixture]	0.6000000000000001	1	7	5	5
2667	110	homophone	[homophone]	0.6000000000000001	1	7	2	2
2668	110	computer	[computer]	0.6000000000000001	1	7	3	3
2669	110	part	[part]	0.6000000000000001	1	8	8	4
2670	110	verb	[verb]	0.6000000000000001	1	7	4	4
2671	110	development	[development]	0.6000000000000001	1	8	4	2
2672	110	noun	[noun]	0.6000000000000001	1	7	6	6
2673	110	English	[English]	0.6000000000000001	1	7	1	1
2674	110	processing	[processing]	0.6000000000000001	1	7	1	1
2675	110	goal	[goal]	0.6000000000000001	1	7	4	4
2676	111	reasoning	[reasoning]	0.5833333333333334	1	7	7	6
2677	112	focus	[focus]	0.58	1	7	6	5
2678	112	aspect	[aspects]	0.58	1	7	6	5
2679	113	complexity	[complexity]	0.5750000000000001	1	7	5	4
2680	114	proposition	[proposition]	0.5666666666666668	1	7	4	3
2681	114	lexicon	[lexicon]	0.5666666666666668	1	7	4	3
2682	115	CSR	[CSR]	0.5	1	6	1	1
2683	115	annotation	[annotation]	0.5	1	6	2	2
2684	115	precision	[precision]	0.5	1	6	5	5
2685	115	world	[world]	0.5	1	6	1	1
2686	115	cluster	[cluster]	0.5	1	6	3	3
2687	115	BLEU	[BLEU]	0.5	1	5	0	0
2688	115	pronoun	[pronoun]	0.5	1	6	4	4
2689	115	dictionary	[dictionary]	0.5	1	6	4	4
2690	115	search	[search]	0.5	1	6	2	2
2691	115	issue	[issue]	0.5	1	6	5	5
2692	115	exist	[existing]	0.5	1	6	3	3
2693	115	component	[component]	0.5	1	6	5	5
2694	115	design	[design]	0.5	1	6	4	4
2695	116	effect	[effect]	0.48	1	6	6	5
2696	116	amount	[amount]	0.48	1	6	6	5
2697	117	type	[type]	0.47500000000000003	1	6	5	4
2698	117	substitution	[substitution]	0.47500000000000003	1	6	5	4
2699	117	definition	[definition]	0.47500000000000003	1	6	5	4
2700	118	prediction	[prediction]	0.46666666666666673	1	6	4	3
2701	119	disposition	[disposition]	0.45	1	6	3	2
2702	120	relationship	[relationship]	0.43333333333333335	1	6	5	3
2703	120	fragment	[fragment]	0.43333333333333335	1	6	5	3
2704	121	p	[P]	0.4	1	4	0	0
2705	121	Machine	[Machine]	0.4	1	5	1	1
2706	121	evidence	[evidence]	0.4	1	5	5	5
2707	121	study	[study]	0.4	1	5	3	3
2708	121	anaphora	[anaphora]	0.4	1	5	4	4
2709	121	attention	[attention]	0.4	1	5	2	2
2710	121	reduction	[reduction]	0.4	1	5	5	5
2711	121	ability	[ability]	0.4	1	4	0	0
2712	121	author	[author]	0.4	1	5	1	1
2713	121	follow	[following]	0.4	1	4	0	0
2714	121	ATIS	[ATIS]	0.4	1	5	3	3
2715	121	chart	[chart]	0.4	1	5	1	1
2716	121	listener	[listener]	0.4	1	5	1	1
2717	121	abstract	[abstract]	0.4	1	5	3	3
2718	121	article	[article]	0.4	1	4	0	0
2719	121	property	[properties]	0.4	1	5	3	3
2720	121	construction	[construction]	0.4	1	5	5	5
2721	121	unification	[unification]	0.4	1	5	1	1
2722	121	summarization	[summarization]	0.4	1	5	4	4
2723	121	syntax	[syntax]	0.4	1	5	1	1
2724	121	comparison	[comparison]	0.4	1	5	4	4
2725	121	classification	[classification]	0.4	1	5	2	2
2726	121	relation	[relation]	0.4	1	5	3	3
2727	121	logic	[logic]	0.4	1	5	1	1
2728	121	mechanism	[mechanism]	0.4	1	5	3	3
2729	121	metaphor	[metaphor]	0.4	1	5	2	2
2730	121	class	[class]	0.4	1	5	5	5
2731	121	node	[node]	0.4	1	5	1	1
2732	121	explanation	[explanation]	0.4	1	5	4	4
2733	121	disambiguation	[disambiguation]	0.4	1	5	2	2
2734	121	stemmer	[stemmer]	0.4	1	5	3	3
2735	121	program	[program]	0.4	1	5	3	3
2736	121	derivation	[derivation]	0.4	1	5	2	2
2737	121	case	[case]	0.4	1	5	4	4
2738	121	transfer	[transfer]	0.4	1	5	2	2
2739	121	trigram	[trigram]	0.4	1	5	2	2
2740	121	oracle	[oracle]	0.4	1	4	0	0
2741	121	category	[Category]	0.4	1	5	3	3
2742	122	area	[area]	0.375	1	5	5	4
2743	123	factor	[factor]	0.35000000000000003	1	5	3	2
2744	123	segmentation	[segmentation]	0.35000000000000003	1	5	3	2
2745	124	newspaper	[newspaper]	0.30000000000000004	1	4	2	2
2746	124	copying	[copying]	0.30000000000000004	1	4	2	2
2747	124	effectiveness	[effectiveness]	0.30000000000000004	1	4	2	2
2748	124	SPR	[SPR]	0.30000000000000004	1	3	0	0
2749	124	HDAG	[HDAG]	0.30000000000000004	1	3	0	0
2750	124	cue	[cue]	0.30000000000000004	1	4	2	2
2751	124	memory	[memory]	0.30000000000000004	1	4	2	2
2752	124	CCR	[CCR]	0.30000000000000004	1	3	0	0
2753	124	impact	[impact]	0.30000000000000004	1	4	4	4
2754	124	treatment	[treatment]	0.30000000000000004	1	4	4	4
2755	124	approximation	[approximation]	0.30000000000000004	1	4	4	4
2756	124	benefit	[benefit]	0.30000000000000004	1	4	3	3
2757	124	February	[February]	0.30000000000000004	1	4	1	1
2758	124	decoding	[decoding]	0.30000000000000004	1	4	3	3
2759	124	Ie	[IE]	0.30000000000000004	1	4	2	2
2760	124	Markov	[Markov]	0.30000000000000004	1	4	2	2
2761	124	NE	[NE]	0.30000000000000004	1	4	4	4
2762	124	SI	[SI]	0.30000000000000004	1	4	2	2
2763	124	portion	[portion]	0.30000000000000004	1	4	3	3
2764	124	web	[Web]	0.30000000000000004	1	4	1	1
2765	124	topic	[topic]	0.30000000000000004	1	3	0	0
2766	124	literature	[literature]	0.30000000000000004	1	4	2	2
2767	124	entry	[entries]	0.30000000000000004	1	4	3	3
2768	124	Prolog	[Prolog]	0.30000000000000004	1	4	2	2
2769	124	WSD	[WSD]	0.30000000000000004	1	4	2	2
2770	124	entity	[entity]	0.30000000000000004	1	4	2	2
2771	124	difference	[difference]	0.30000000000000004	1	3	0	0
2772	124	measure	[measure]	0.30000000000000004	1	4	2	2
2773	124	point	[point]	0.30000000000000004	1	4	2	2
2774	124	hub	[hub]	0.30000000000000004	1	4	1	1
2775	124	solution	[solution]	0.30000000000000004	1	4	2	2
2776	124	intention	[intention]	0.30000000000000004	1	4	1	1
2777	124	HBG	[HBG]	0.30000000000000004	1	3	0	0
2778	124	referent	[referents]	0.30000000000000004	1	4	3	3
2779	124	variation	[variation]	0.30000000000000004	1	4	4	4
2780	124	hypothesis	[hypothesis]	0.30000000000000004	1	4	1	1
2781	124	Chat-80	[Chat-80]	0.30000000000000004	1	3	0	0
2782	124	control	[control]	0.30000000000000004	1	3	0	0
2783	124	feedback	[feedback]	0.30000000000000004	1	3	0	0
2784	124	place	[place]	0.30000000000000004	1	4	2	2
2785	124	left	[left]	0.30000000000000004	1	3	0	0
2786	124	object	[objects]	0.30000000000000004	1	4	2	2
2787	124	conjunction	[conjunction]	0.30000000000000004	1	4	2	2
2788	124	counterexample	[counterexamples]	0.30000000000000004	1	4	3	3
2789	124	means	[means]	0.30000000000000004	1	4	4	4
2790	124	phrasal	[phrasal]	0.30000000000000004	1	3	0	0
2791	124	LFG	[LFG]	0.30000000000000004	1	3	0	0
2792	124	MIT	[MIT]	0.30000000000000004	1	3	0	0
2793	124	string	[string]	0.30000000000000004	1	4	2	2
2794	124	actor	[actor]	0.30000000000000004	1	3	0	0
2795	124	ocr	[OCR]	0.30000000000000004	1	4	2	2
2796	124	index	[index]	0.30000000000000004	1	4	3	3
2797	124	claim	[claim]	0.30000000000000004	1	4	2	2
2798	124	device	[device]	0.30000000000000004	1	3	0	0
2799	124	activity	[activity]	0.30000000000000004	1	4	1	1
2800	124	extract	[extract]	0.30000000000000004	1	4	2	2
2801	124	correlation	[correlation]	0.30000000000000004	1	4	3	3
2802	124	conversation	[conversation]	0.30000000000000004	1	4	2	2
2803	124	interface	[interface]	0.30000000000000004	1	4	1	1
2804	124	HPSG	[HPSG]	0.30000000000000004	1	3	0	0
2805	124	meeting	[meeting]	0.30000000000000004	1	4	1	1
2806	124	request	[request]	0.30000000000000004	1	3	0	0
2807	124	account	[account]	0.30000000000000004	1	4	2	2
2808	124	capability	[capability]	0.30000000000000004	1	4	2	2
2809	124	classifier	[classifiers]	0.30000000000000004	1	4	1	1
2810	124	instance	[instance]	0.30000000000000004	1	3	0	0
2811	124	space	[space]	0.30000000000000004	1	4	2	2
2812	124	prototype	[prototype]	0.30000000000000004	1	3	0	0
2813	124	tagger	[tagger]	0.30000000000000004	1	4	3	3
2814	124	view	[view]	0.30000000000000004	1	4	1	1
2815	124	statistic	[statistics]	0.30000000000000004	1	4	2	2
2816	124	name	[name]	0.30000000000000004	1	4	3	3
2817	124	scheme	[scheme]	0.30000000000000004	1	4	2	2
2818	124	honorifics	[honorifics]	0.30000000000000004	1	4	3	3
2819	124	n-gram	[n-gram]	0.30000000000000004	1	4	2	2
2820	124	coordinate	[coordinate]	0.30000000000000004	1	3	0	0
2821	124	argument	[argument]	0.30000000000000004	1	4	1	1
2822	125	value	[value]	0.2666666666666667	1	4	4	3
2823	125	one	[one]	0.2666666666666667	1	4	4	3
2824	125	list	[list]	0.2666666666666667	1	4	4	3
2825	126	content	[content]	0.25	1	4	3	2
2826	126	variety	[variety]	0.25	1	4	3	2
2827	126	distribution	[distribution]	0.25	1	4	3	2
2828	126	sample	[sample]	0.25	1	4	3	2
2829	126	range	[range]	0.25	1	4	3	2
2830	127	L	[L]	0.2	1	2	0	0
2831	127	spelling-checker	[spelling-checkers]	0.2	1	3	1	1
2832	127	redundancy	[redundancy]	0.2	1	3	2	2
2833	127	concern	[concern]	0.2	1	3	2	2
2834	127	element	[element]	0.2	1	3	1	1
2835	127	SUN	[SUN]	0.2	1	2	0	0
2836	127	paraphrasing	[paraphrasing]	0.2	1	3	2	2
2837	127	participant	[Participants]	0.2	1	2	0	0
2838	127	TAG	[TAG]	0.2	1	2	0	0
2839	127	SENSEVAL-2	[SENSEVAL-2]	0.2	1	2	0	0
2840	127	island	[islands]	0.2	1	2	0	0
2841	127	format	[format]	0.2	1	3	1	1
2842	127	CFG	[CFG]	0.2	1	2	0	0
2843	127	acquisition	[acquisition]	0.2	1	3	3	3
2844	127	expectation	[expectations]	0.2	1	3	2	2
2845	127	attempt	[attempt]	0.2	1	2	0	0
2846	127	subject	[subject]	0.2	1	3	1	1
2847	127	.g	[.g]	0.2	1	3	1	1
2848	127	clustering	[clustering]	0.2	1	3	1	1
2849	127	predicate	[predicate]	0.2	1	3	2	2
2850	127	subpredicate	[subpredicates]	0.2	1	2	0	0
2851	127	progress	[progress]	0.2	1	3	1	1
2852	127	agent	[agent]	0.2	1	3	2	2
2853	127	modification	[modification]	0.2	1	3	1	1
2854	127	formulas	[formulas]	0.2	1	3	2	2
2855	127	parameter	[parameter]	0.2	1	3	2	2
2856	127	year	[years]	0.2	1	3	2	2
2857	127	planning	[planning]	0.2	1	2	0	0
2858	127	MADCOW	[MADCOW]	0.2	1	2	0	0
2859	127	USR	[USR]	0.2	1	2	0	0
2860	127	ensemble	[ensemble]	0.2	1	2	0	0
2861	127	requirement	[requirement]	0.2	1	3	3	3
2862	127	dominance	[dominance]	0.2	1	2	0	0
2863	127	protocol	[protocol]	0.2	1	3	2	2
2864	127	DARPA	[DARPA]	0.2	1	4	2	1
2865	127	genre	[genre]	0.2	1	2	0	0
2866	127	Grammar	[Grammars]	0.2	1	3	2	2
2867	127	er	[ER]	0.2	1	3	1	1
2868	127	transcription	[transcription]	0.2	1	3	2	2
2869	127	wsi	[WSI]	0.2	1	3	2	2
2870	127	robustness	[robustness]	0.2	1	3	1	1
2871	127	lr	[LR]	0.2	1	2	0	0
2872	127	qa	[QA]	0.2	1	3	3	3
2873	127	Pereira	[Pereira]	0.2	1	2	0	0
2874	127	LTAG	[LTAG]	0.2	1	2	0	0
2875	127	assessor	[assessors]	0.2	1	2	0	0
2876	127	Senseval-3	[Senseval-3]	0.2	1	2	0	0
2877	127	editor	[editor]	0.2	1	3	2	2
2878	127	chain	[chain]	0.2	1	3	3	3
2879	127	potential	[potential]	0.2	1	2	0	0
2880	127	dimension	[dimensions]	0.2	1	3	2	2
2881	127	hmm	[HMM]	0.2	1	2	0	0
2882	127	preference	[preference]	0.2	1	3	3	3
2883	127	P-CFG	[P-CFG]	0.2	1	2	0	0
2884	127	inheritance	[inheritance]	0.2	1	3	2	2
2885	127	disjunction	[disjunctions]	0.2	1	2	0	0
2886	127	compound	[compound]	0.2	1	4	2	1
2887	127	fact	[fact]	0.2	1	3	1	1
2888	127	finding	[finding]	0.2	1	3	1	1
2889	127	detect	[detecting]	0.2	1	2	0	0
2890	127	criterion	[criterion]	0.2	1	3	3	3
2891	127	vocabulary	[vocabulary]	0.2	1	3	2	2
2892	127	interpolation	[interpolation]	0.2	1	3	3	3
2893	127	entailment	[entailment]	0.2	1	2	0	0
2894	127	unit	[unit]	0.2	1	3	2	2
2895	127	identification	[identification]	0.2	1	3	3	3
2896	127	IBM	[IBM]	0.2	1	3	1	1
2897	127	boosting	[boosting]	0.2	1	2	0	0
2898	127	style	[style]	0.2	1	3	2	2
2899	127	interruption	[interruptions]	0.2	1	2	0	0
2900	127	expert	[expert]	0.2	1	3	1	1
2901	127	level	[levels]	0.2	1	3	3	3
2902	127	query	[query]	0.2	1	3	2	2
2903	127	coverage	[coverage]	0.2	1	3	1	1
2904	127	technology	[technology]	0.2	1	3	3	3
2905	127	collocation	[collocation]	0.2	1	2	0	0
2906	127	corresponding	[corresponding]	0.2	1	3	3	3
2907	127	tutorial	[tutorial]	0.2	1	2	0	0
2908	127	function	[function]	0.2	1	3	1	1
2909	127	clause	[clauses]	0.2	1	2	0	0
2910	127	principle	[principle]	0.2	1	3	1	1
2911	127	bootstrapping	[bootstrapping]	0.2	1	2	0	0
2912	127	experience	[experience]	0.2	1	3	2	2
2913	127	RAREAS	[RAREAS]	0.2	1	2	0	0
2914	127	path	[path]	0.2	1	3	2	2
2915	127	bird	[birds]	0.2	1	3	1	1
2916	127	human	[humans]	0.2	1	2	0	0
2917	127	field	[field]	0.2	1	3	1	1
2918	127	boost	[boosting]	0.2	1	2	0	0
2919	127	flexibility	[flexibility]	0.2	1	2	0	0
2920	127	utility	[utility]	0.2	1	3	3	3
2921	127	behavior	[behavior]	0.2	1	3	3	3
2922	127	mistake	[mistakes]	0.2	1	3	1	1
2923	127	morpheme	[morpheme]	0.2	1	3	1	1
2924	127	attribute	[attribute]	0.2	1	3	1	1
2925	127	InfoMagnet	[InfoMagnets]	0.2	1	2	0	0
2926	127	determiner	[determiners]	0.2	1	3	1	1
2927	127	board	[board]	0.2	1	2	0	0
2928	127	TREC	[TREC]	0.2	1	2	0	0
2929	127	hierarchy	[hierarchy]	0.2	1	4	2	1
2930	127	image	[image]	0.2	1	3	1	1
2931	127	frame	[frame]	0.2	1	3	2	2
2932	127	calculus	[calculus]	0.2	1	3	2	2
2933	127	q/a	[Q/A]	0.2	1	3	1	1
2934	127	tagging	[tagging]	0.2	1	3	1	1
2935	127	core	[core]	0.2	1	2	0	0
2936	127	link	[link]	0.2	1	3	2	2
2937	127	cost	[cost]	0.2	1	3	2	2
2938	127	challenge	[challenge]	0.2	1	3	3	3
2939	127	kernel	[kernel]	0.2	1	2	0	0
2940	127	miscommunication	[miscommunication]	0.2	1	3	1	1
2941	127	demonstration	[demonstration]	0.2	1	3	2	2
2942	127	.5%	[.5%]	0.2	1	3	1	1
2943	127	workshop	[workshop]	0.2	1	3	1	1
2944	127	subtopic	[subtopic]	0.2	1	3	2	2
2945	127	transition	[transition]	0.2	1	3	2	2
2946	127	JAVELIN	[JAVELIN]	0.2	1	2	0	0
2947	127	Japanese	[Japanese]	0.2	1	2	0	0
2948	127	bigram	[bigram]	0.2	1	3	2	2
2949	127	summary	[summary]	0.2	1	3	1	1
2950	127	explicitation	[explicitation]	0.2	1	2	0	0
2951	127	comma	[comma]	0.2	1	3	1	1
2952	127	decoder	[decoder]	0.2	1	3	2	2
2953	127	network	[network]	0.2	1	4	2	1
2954	127	Arabic	[Arabic]	0.2	1	2	0	0
2955	127	unigram	[unigram]	0.2	1	3	1	1
2956	127	arc	[arc]	0.2	1	3	2	2
2957	127	art	[art]	0.2	1	2	0	0
2958	127	asr	[ASR]	0.2	1	2	0	0
2959	127	instantiation	[instantiation]	0.2	1	3	2	2
2960	127	editing	[editing]	0.2	1	2	0	0
2961	127	POURPRE	[POURPRE]	0.2	1	2	0	0
2962	127	move	[move]	0.2	1	3	2	2
2963	127	RCL	[RCL]	0.2	1	2	0	0
2964	127	surface	[surface]	0.2	1	3	1	1
2965	127	entropy	[entropy]	0.2	1	2	0	0
2966	127	suffix	[suffix]	0.2	1	2	0	0
2967	127	nature	[nature]	0.2	1	3	3	3
2968	127	generalization	[generalization]	0.2	1	3	2	2
2969	127	efficiency	[efficiency]	0.2	1	3	2	2
2970	127	response	[response]	0.2	1	3	3	3
2971	128	lack	[lack]	0.15000000000000002	1	3	3	2
2972	128	learner	[learner]	0.15000000000000002	1	3	3	2
2973	129	tdmt	[TDMT]	0.1	1	1	0	0
2974	129	scratch	[scratch]	0.1	1	1	0	0
2975	129	gold	[gold]	0.1	1	1	0	0
2976	129	U	[U]	0.1	1	1	0	0
2977	129	neighbor	[neighbors]	0.1	1	1	0	0
2978	129	-speak	[-speaking]	0.1	1	1	0	0
2979	129	STTK	[STTK]	0.1	1	2	1	1
2980	129	polysemy	[polysemy]	0.1	1	1	0	0
2981	129	root	[root]	0.1	1	1	0	0
2982	129	BIA	[BIA]	0.1	1	1	0	0
2983	129	C++	[C++]	0.1	1	1	0	0
2984	129	Byron's	[Byron's]	0.1	1	1	0	0
2985	129	morphology	[morphology]	0.1	1	1	0	0
2986	129	hypertext	[hypertext]	0.1	1	1	0	0
2987	129	regard	[regard]	0.1	1	1	0	0
2988	129	SPG	[SPG]	0.1	1	1	0	0
2989	129	DHPL	[DHPL]	0.1	1	1	0	0
2990	129	x-window	[X-windows]	0.1	1	1	0	0
2991	129	Swedes	[Swedes]	0.1	1	1	0	0
2992	129	creation	[creation]	0.1	1	1	0	0
2993	129	thing	[thing]	0.1	1	2	2	2
2994	129	house	[house]	0.1	1	1	0	0
2995	129	consistency	[consistency]	0.1	1	1	0	0
2996	129	restoration	[restoration]	0.1	1	1	0	0
2997	129	TACITUS	[TACITUS]	0.1	1	1	0	0
2998	129	dag	[dags]	0.1	1	1	0	0
2999	129	SUN4	[SUN4]	0.1	1	1	0	0
3000	129	logistics	[logistics]	0.1	1	1	0	0
3001	129	introduction	[introduction]	0.1	1	2	2	2
3002	129	build	[building]	0.1	1	2	1	1
3003	129	methodology	[methodology]	0.1	1	2	2	2
3004	129	retrieval	[retrieval]	0.1	1	2	1	1
3005	129	LCS-Marine	[LCS-Marine]	0.1	1	2	1	1
3006	129	TGE	[TGE]	0.1	1	1	0	0
3007	129	average	[average]	0.1	1	2	1	1
3008	129	error-correction	[error-correction]	0.1	1	2	1	1
3009	129	right	[right]	0.1	1	2	1	1
3010	129	stage	[stage]	0.1	1	2	1	1
3011	129	balance	[balance]	0.1	1	1	0	0
3012	129	status	[status]	0.1	1	2	1	1
3013	129	usr	[USRs]	0.1	1	1	0	0
3014	129	Wall	[Wall]	0.1	1	2	2	2
3015	129	Information	[Information]	0.1	1	2	2	2
3016	129	inform	[informing]	0.1	1	1	0	0
3017	129	industry	[industry]	0.1	1	1	0	0
3018	129	np-complete	[NP-complete]	0.1	1	1	0	0
3019	129	interest	[interest]	0.1	1	1	0	0
3020	129	layout	[layout]	0.1	1	2	1	1
3021	129	base	[base]	0.1	1	2	1	1
3022	129	stem	[stemming]	0.1	1	2	1	1
3023	129	non-match	[non-matches]	0.1	1	1	0	0
3024	129	extension	[extensions]	0.1	1	2	2	2
3025	129	Korean	[Korean]	0.1	1	1	0	0
3026	129	opinion	[opinion]	0.1	1	1	0	0
3027	129	UCG	[UCG]	0.1	1	2	1	1
3028	129	eca	[ECA]	0.1	1	1	0	0
3029	129	revision	[revision]	0.1	1	1	0	0
3030	129	out-degree	[out-degree]	0.1	1	1	0	0
3031	129	usefulness	[usefulness]	0.1	1	1	0	0
3032	129	SYSCONJ	[SYSCONJ]	0.1	1	1	0	0
3033	129	density	[density]	0.1	1	1	0	0
3034	129	lexicologist	[lexicologists]	0.1	1	1	0	0
3035	129	Treebank	[Treebank]	0.1	1	2	1	1
3036	129	constituent	[constituent]	0.1	1	2	1	1
3037	129	engine	[engine]	0.1	1	1	0	0
3038	129	kanji-kana	[kanji-kana]	0.1	1	2	1	1
3039	129	environment	[environment]	0.1	1	2	2	2
3040	129	HK	[HK]	0.1	1	1	0	0
3041	129	phrasing	[phrasing]	0.1	1	1	0	0
3042	129	LM	[LM]	0.1	1	1	0	0
3043	129	perception	[perceptions]	0.1	1	1	0	0
3044	129	Human-Machine	[Human-Machine]	0.1	1	1	0	0
3045	129	defeasibility	[defeasibility]	0.1	1	1	0	0
3046	129	PC	[PC]	0.1	1	1	0	0
3047	129	hdag	[HDAGs]	0.1	1	1	0	0
3048	129	pivot	[pivot]	0.1	1	1	0	0
3049	129	[Zernik87]	[[Zernik87]]	0.1	1	1	0	0
3050	129	default	[default]	0.1	1	1	0	0
3051	129	metric	[metric]	0.1	1	2	1	1
3052	129	TL	[TL]	0.1	1	1	0	0
3053	129	Spaniards	[Spaniards]	0.1	1	1	0	0
3054	129	conclusion	[conclusions]	0.1	1	1	0	0
3055	129	and/or	[and/or]	0.1	1	2	1	1
3056	129	newswire	[newswire]	0.1	1	1	0	0
3057	129	parsing-base	[parsing-based]	0.1	1	1	0	0
3058	129	Research	[Research]	0.1	1	1	0	0
3059	129	porting	[porting]	0.1	1	1	0	0
3060	129	section	[section]	0.1	1	1	0	0
3061	129	presenter	[presenter]	0.1	1	1	0	0
3062	129	system'	[system's]	0.1	1	2	1	1
3063	129	Roget's	[Roget's]	0.1	1	1	0	0
3064	129	generator	[generator]	0.1	1	1	0	0
3065	129	four-participant	[four-participants]	0.1	1	1	0	0
3066	129	extent	[extent]	0.1	1	2	2	2
3067	129	assumption	[assumption]	0.1	1	2	2	2
3068	129	al	[al]	0.1	1	1	0	0
3069	129	scoring	[scoring]	0.1	1	2	1	1
3070	129	cf	[cf]	0.1	1	1	0	0
3071	129	human-machine	[human-machine]	0.1	1	1	0	0
3072	129	VME	[VME]	0.1	1	1	0	0
3073	129	antecedence	[antecedence]	0.1	1	2	1	1
3074	129	contrary	[contrary]	0.1	1	1	0	0
3075	129	specification	[specification]	0.1	1	2	2	2
3076	129	ir	[IR]	0.1	1	1	0	0
3077	129	belief	[belief]	0.1	1	2	2	2
3078	129	expansion	[expansion]	0.1	1	1	0	0
3079	129	gaze	[gaze]	0.1	1	2	1	1
3080	129	ne	[NE]	0.1	1	2	1	1
3081	129	np	[NP]	0.1	1	1	0	0
3082	129	storage	[storage]	0.1	1	1	0	0
3083	129	head	[head]	0.1	1	2	2	2
3084	129	misunderstanding	[misunderstandings]	0.1	1	2	1	1
3085	129	determination	[determination]	0.1	1	2	1	1
3086	129	seed	[seed]	0.1	1	2	2	2
3087	129	KPSG	[KPSG]	0.1	1	2	1	1
3088	129	grammaticality	[grammaticality]	0.1	1	2	1	1
3089	129	TAP-XL	[TAP-XL]	0.1	1	1	0	0
3090	129	tv	[TV]	0.1	1	1	0	0
3091	129	failure	[failure]	0.1	1	2	1	1
3092	129	gap	[gap]	0.1	1	2	2	2
3093	129	examination	[Examination]	0.1	1	2	2	2
3094	129	insight	[insight]	0.1	1	2	2	2
3095	129	f-measure	[F-measure]	0.1	1	2	1	1
3096	129	WER	[WER]	0.1	1	1	0	0
3097	129	power	[power]	0.1	1	3	2	1
3098	129	pronominalization	[pronominalization]	0.1	1	1	0	0
3099	129	transformation	[transformation]	0.1	1	1	0	0
3100	129	depth	[depth]	0.1	1	1	0	0
3101	129	Model	[Model]	0.1	1	2	2	2
3102	129	addressee	[addressee]	0.1	1	2	1	1
3103	129	Senseval	[Senseval]	0.1	1	1	0	0
3104	129	scale	[scale]	0.1	1	1	0	0
3105	129	Fagan	[Fagan]	0.1	1	1	0	0
3106	129	PageRank	[PageRank]	0.1	1	1	0	0
3107	129	LIMSI	[LIMSI]	0.1	1	1	0	0
3108	129	success	[success]	0.1	1	2	1	1
3109	129	analyst	[Analyst's]	0.1	1	2	2	2
3110	129	channel	[channel]	0.1	1	1	0	0
3111	129	FSM	[FSM]	0.1	1	2	1	1
3112	129	automaton	[automaton]	0.1	1	2	2	2
3113	129	ATNs	[ATNs]	0.1	1	1	0	0
3114	129	native	[native]	0.1	1	1	0	0
3115	129	type-unification	[type-unification]	0.1	1	1	0	0
3116	129	discussion	[discussion]	0.1	1	2	1	1
3117	129	reranking	[reranking]	0.1	1	1	0	0
3118	129	k-means	[k-means]	0.1	1	1	0	0
3119	129	keyword	[keywords]	0.1	1	2	1	1
3120	129	Collins	[Collins]	0.1	1	1	0	0
3121	129	effort	[effort]	0.1	1	2	1	1
3122	129	syllogism	[syllogism]	0.1	1	1	0	0
3123	129	char	[char]	0.1	1	1	0	0
3124	129	DoPS	[DoPS]	0.1	1	1	0	0
3125	129	attendance	[attendance]	0.1	1	1	0	0
3126	129	advance	[advances]	0.1	1	2	2	2
3127	129	affix	[affix]	0.1	1	2	2	2
3128	129	dry-run	[dry-run]	0.1	1	1	0	0
3129	129	Buckley	[Buckley]	0.1	1	1	0	0
3130	129	log-likelihood	[log-likelihood]	0.1	1	1	0	0
3131	129	generality	[generality]	0.1	1	2	1	1
3132	129	memo	[memos]	0.1	1	1	0	0
3133	129	tracking	[tracking]	0.1	1	2	1	1
3134	129	Lincoln	[Lincoln]	0.1	1	2	1	1
3135	129	expressiveness	[expressiveness]	0.1	1	1	0	0
3136	129	magnitude	[magnitude]	0.1	1	1	0	0
3137	129	refinement	[refinements]	0.1	1	1	0	0
3138	129	(mt)	[(MT)]	0.1	1	2	2	2
3139	129	beam-search	[beam-search]	0.1	1	1	0	0
3140	129	Kyoto	[Kyoto]	0.1	1	1	0	0
3141	129	coordination	[coordination]	0.1	1	2	2	2
3142	129	presentation	[presentation]	0.1	1	1	0	0
3143	129	independency	[independency]	0.1	1	1	0	0
3144	129	word-trie	[word-trie]	0.1	1	1	0	0
3145	129	laboratory	[laboratory]	0.1	1	1	0	0
3146	129	certainty	[certainty]	0.1	1	2	2	2
3147	129	inflection	[inflection]	0.1	1	2	1	1
3148	129	Spanish	[Spanish]	0.1	1	1	0	0
3149	129	Unification	[Unification]	0.1	1	1	0	0
3150	129	concurrency	[concurrency]	0.1	1	2	1	1
3151	129	word-sense	[word-sense]	0.1	1	2	2	2
3152	129	characterization	[characterization]	0.1	1	1	0	0
3153	129	ontology	[ontology]	0.1	1	1	0	0
3154	129	MRD's	[MRD's]	0.1	1	1	0	0
3155	129	identify	[identifying]	0.1	1	1	0	0
3156	129	testing	[testing]	0.1	1	1	0	0
3157	129	synthesis	[synthesis]	0.1	1	2	2	2
3158	129	Transition	[Transition]	0.1	1	1	0	0
3159	129	predict	[predicting]	0.1	1	1	0	0
3160	129	reason	[reason]	0.1	1	2	2	2
3161	129	Zadeh	[Zadeh]	0.1	1	1	0	0
3162	129	notion	[notions]	0.1	1	2	2	2
3163	129	detection	[detection]	0.1	1	2	2	2
3164	129	bag-of-word	[bag-of-words]	0.1	1	2	1	1
3165	129	telephone	[telephone]	0.1	1	1	0	0
3166	129	difficulty	[difficulty]	0.1	1	2	1	1
3167	129	Linde	[Linde]	0.1	1	1	0	0
3168	129	vagueness	[vagueness]	0.1	1	2	2	2
3169	129	hardness	[hardness]	0.1	1	1	0	0
3170	129	forecaster'	[forecaster's]	0.1	1	1	0	0
3171	129	architecture	[architecture]	0.1	1	2	2	2
3172	129	grounding	[grounding]	0.1	1	2	1	1
3173	129	correction	[correction]	0.1	1	1	0	0
3174	129	limitation	[limitations]	0.1	1	2	1	1
3175	129	[1]	[[1]]	0.1	1	1	0	0
3176	129	quantifier	[quantifier]	0.1	1	2	1	1
3177	129	investigation	[investigation]	0.1	1	1	0	0
3178	129	programming	[programming]	0.1	1	1	0	0
3179	129	Freund	[Freund]	0.1	1	1	0	0
3180	129	perspective-taking	[perspective-taking]	0.1	1	1	0	0
3181	129	developer	[developers]	0.1	1	2	1	1
3182	129	subset	[subset]	0.1	1	2	1	1
3183	129	Arguello	[Arguello]	0.1	1	1	0	0
3184	129	corner	[corner]	0.1	1	1	0	0
3185	129	Parsing	[Parsing]	0.1	1	1	0	0
3186	129	report	[report]	0.1	1	2	1	1
3187	129	publication	[publication]	0.1	1	1	0	0
3188	129	Phonology	[Phonology]	0.1	1	1	0	0
3189	129	reconstruction	[reconstruction]	0.1	1	1	0	0
3190	129	Arabic-English	[Arabic-English]	0.1	1	1	0	0
3191	129	browsing	[browsing]	0.1	1	1	0	0
3192	129	/Mercer	[/Mercer]	0.1	1	1	0	0
3193	129	Shieber	[Shieber]	0.1	1	2	1	1
3194	129	variant	[variant]	0.1	1	2	2	2
3195	129	reach	[reach]	0.1	1	2	2	2
3196	129	EBMT	[EBMT]	0.1	1	2	1	1
3197	129	restart	[restarting]	0.1	1	1	0	0
3198	129	observation	[observation]	0.1	1	2	2	2
3199	129	passage	[passage]	0.1	1	1	0	0
3200	129	typicality	[typicality]	0.1	1	1	0	0
3201	129	match	[match]	0.1	1	2	1	1
3202	129	Czech	[Czech]	0.1	1	1	0	0
3203	129	pilot	[pilot]	0.1	1	2	1	1
3204	129	co-worker	[co-workers]	0.1	1	1	0	0
3205	129	implication	[implications]	0.1	1	2	2	2
3206	129	product	[products]	0.1	1	1	0	0
3207	129	law	[laws]	0.1	1	1	0	0
3208	129	condition	[condition]	0.1	1	2	2	2
3209	129	RINA	[RINA]	0.1	1	1	0	0
3210	129	supply	[supply]	0.1	1	1	0	0
3211	129	scenario	[scenario]	0.1	1	2	2	2
3212	129	graphic	[graphics]	0.1	1	1	0	0
3213	129	situation	[situation]	0.1	1	2	1	1
3214	129	dataset	[datasets]	0.1	1	2	1	1
3215	129	Bleu	[Bleu]	0.1	1	1	0	0
3216	129	doubt	[doubt]	0.1	1	1	0	0
3217	129	center	[centering]	0.1	1	1	0	0
3218	129	hardware	[hardware]	0.1	1	1	0	0
3219	129	indexing	[indexing]	0.1	1	2	2	2
3220	129	cohesion	[cohesion]	0.1	1	1	0	0
3221	129	coherence	[coherence]	0.1	1	2	1	1
3222	129	October	[October]	0.1	1	2	2	2
3223	129	structure-sharing	[structure-sharing]	0.1	1	2	1	1
3224	129	signal	[signals]	0.1	1	1	0	0
3225	129	importance	[importance]	0.1	1	2	2	2
3226	129	title	[title]	0.1	1	1	0	0
3227	129	harvesting	[harvesting]	0.1	1	1	0	0
3228	129	Montague	[Montague]	0.1	1	2	1	1
3229	129	message-passing	[message-passing]	0.1	1	1	0	0
3230	129	mention	[mentioning]	0.1	1	1	0	0
3231	129	map	[MAP]	0.1	1	1	0	0
3232	129	syllable	[syllables]	0.1	1	1	0	0
3233	129	cdhmm	[CDHMM]	0.1	1	1	0	0
3234	129	background	[backgrounds]	0.1	1	1	0	0
3235	129	repeat	[repeating]	0.1	1	1	0	0
3236	129	module	[module]	0.1	1	2	1	1
3237	129	SUMMIT	[SUMMIT]	0.1	1	1	0	0
3238	129	project	[project]	0.1	1	1	0	0
3239	129	enhancement	[enhancements]	0.1	1	1	0	0
3240	129	mle	[MLE]	0.1	1	1	0	0
3241	129	Penn	[Penn]	0.1	1	1	0	0
3242	129	speech-search	[speech-search]	0.1	1	1	0	0
3243	129	LNR	[LNR]	0.1	1	1	0	0
3244	129	LMs	[LMs]	0.1	1	2	2	2
3245	129	switchboard	[Switchboard]	0.1	1	1	0	0
3246	129	word-trigram	[word-trigram]	0.1	1	1	0	0
3247	129	benchmark	[benchmark]	0.1	1	2	1	1
3248	129	KANA-KANJI	[KANA-KANJI]	0.1	1	1	0	0
3249	129	mrr	[MRR]	0.1	1	1	0	0
3250	129	LRE	[LRE]	0.1	1	1	0	0
3251	129	voice	[voice]	0.1	1	1	0	0
3252	129	Landsbergen	[Landsbergen]	0.1	1	1	0	0
3253	129	sequence	[sequence]	0.1	1	2	1	1
3254	129	chunk	[chunk]	0.1	1	2	1	1
3255	129	annotate	[annotating]	0.1	1	1	0	0
3256	129	detail	[detail]	0.1	1	1	0	0
3257	129	CMU's	[CMU's]	0.1	1	1	0	0
3258	129	[rcg]	[[RCG]]	0.1	1	1	0	0
3259	129	discrimination	[discrimination]	0.1	1	1	0	0
3260	129	Hong	[Hong]	0.1	1	1	0	0
3261	129	p#p	[P#P]	0.1	1	1	0	0
3262	129	mapping	[mapping]	0.1	1	2	2	2
3263	129	Horn	[Horn]	0.1	1	1	0	0
3264	129	lexical-cohesion	[lexical-cohesion]	0.1	1	2	1	1
3265	129	boundary	[boundary]	0.1	1	2	2	2
3266	129	indexation	[indexation]	0.1	1	3	2	1
3267	129	engage	[engaging]	0.1	1	1	0	0
3268	129	hastiness	[hastiness]	0.1	1	1	0	0
3269	129	multiparty	[multiparty]	0.1	1	1	0	0
3270	129	subparagraph	[subparagraph]	0.1	1	1	0	0
3271	129	np-	[NP-]	0.1	1	1	0	0
3272	129	simplicity	[simplicity]	0.1	1	1	0	0
3273	129	statement	[statement]	0.1	1	2	2	2
3274	129	refer	[Referring]	0.1	1	1	0	0
3275	129	markov	[Markov]	0.1	1	1	0	0
3276	129	dimensionality	[dimensionality]	0.1	1	1	0	0
3277	129	[rcl]	[[RCL]]	0.1	1	1	0	0
3278	129	probability	[probabilities]	0.1	1	1	0	0
3279	129	terminal	[terminal]	0.1	1	1	0	0
3280	129	Lewis	[Lewis]	0.1	1	1	0	0
3281	129	rejoinder	[rejoinder]	0.1	1	2	1	1
3282	129	consideration	[consideration]	0.1	1	2	1	1
3283	129	valency	[valency]	0.1	1	2	1	1
3284	129	ordering	[ordering]	0.1	1	1	0	0
3285	129	decade	[decade]	0.1	1	2	2	2
3286	129	practice	[practice]	0.1	1	1	0	0
3287	129	edit	[edit]	0.1	1	1	0	0
3288	129	starting	[starting]	0.1	1	1	0	0
3289	129	co-occurrence	[co-occurrence]	0.1	1	1	0	0
3290	129	vector	[vector]	0.1	1	1	0	0
3291	129	heuristic	[heuristics]	0.1	1	1	0	0
3292	129	smta	[SmTA]	0.1	1	1	0	0
3293	129	predicate-argument	[predicate-argument]	0.1	1	2	2	2
3294	129	Korean-to-English	[Korean-to-English]	0.1	1	1	0	0
3295	129	location	[location]	0.1	1	1	0	0
3296	129	software	[software]	0.1	1	1	0	0
3297	129	repository	[repository]	0.1	1	2	2	2
3298	129	presence	[presence]	0.1	1	2	2	2
3299	129	FERRET	[FERRET]	0.1	1	2	1	1
3300	129	Case	[Case]	0.1	1	1	0	0
3301	129	copy	[copy]	0.1	1	2	1	1
3302	129	Schutze	[Schutze's]	0.1	1	1	0	0
3303	129	line	[line]	0.1	1	1	0	0
3304	129	recovery	[recovery]	0.1	1	1	0	0
3305	129	confidence	[confidence]	0.1	1	2	2	2
3306	129	Phrasal	[Phrasal]	0.1	1	1	0	0
3307	129	discuss	[discussing]	0.1	1	1	0	0
3308	129	Intel	[Intel]	0.1	1	1	0	0
3309	129	integration	[integration]	0.1	1	2	1	1
3310	129	manner	[manner]	0.1	1	2	2	2
3311	129	requestor	[Requestors]	0.1	1	1	0	0
3312	129	cooccurrence	[cooccurrence]	0.1	1	2	1	1
3313	129	formalisme	[formalism's]	0.1	1	1	0	0
3314	129	pos	[POS]	0.1	1	1	0	0
3315	129	scheduling	[scheduling]	0.1	1	1	0	0
3316	129	per)	[PER)]	0.1	1	1	0	0
3317	129	.0%	[.0%]	0.1	1	2	2	2
3318	129	well-formedness	[well-formedness]	0.1	1	1	0	0
3319	129	.2%	[.2%]	0.1	1	2	2	2
3320	129	extraction	[extraction]	0.1	1	2	1	1
3321	129	light	[light]	0.1	1	2	1	1
3322	129	choice	[choice]	0.1	1	2	2	2
3323	129	Kumar	[Kumar]	0.1	1	1	0	0
3324	129	pragmatics	[pragmatics]	0.1	1	1	0	0
3325	129	California	[California]	0.1	1	1	0	0
3326	129	series	[series]	0.1	1	3	2	1
3327	129	SPoT	[SPoT]	0.1	1	1	0	0
3328	129	session	[session]	0.1	1	1	0	0
3329	129	adjoin	[adjoining]	0.1	1	2	1	1
3330	129	material	[material]	0.1	1	2	2	2
3331	129	PER	[PER]	0.1	1	1	0	0
3332	129	hope	[hope]	0.1	1	1	0	0
3333	129	member	[member]	0.1	1	2	2	2
3334	129	patent	[patent]	0.1	1	1	0	0
3335	129	wh-question	[WH-questions]	0.1	1	1	0	0
3336	129	direction-givegivegivegivegivegivegivion-	[direction-giving]	0.1	1	1	0	0
3337	129	segmenter	[segmenter]	0.1	1	1	0	0
3338	129	deviation	[deviations]	0.1	1	1	0	0
3339	129	logon	[LOGON]	0.1	1	1	0	0
3340	129	people	[people]	0.1	1	1	0	0
3341	129	ranking	[rankings]	0.1	1	2	1	1
3342	129	operation	[operation]	0.1	1	2	2	2
3343	129	demonstrator	[demonstrator]	0.1	1	1	0	0
3344	129	span	[span]	0.1	1	2	2	2
3345	129	principle-and-parameters	[principle-and-parameters]	0.1	1	1	0	0
3346	129	extrapolation	[extrapolation]	0.1	1	1	0	0
3347	129	parser/generator	[parser/generator]	0.1	1	2	2	2
3348	129	score	[scores]	0.1	1	1	0	0
3349	129	TDMT	[TDMT]	0.1	1	2	2	2
3350	129	uncertainty	[uncertainty]	0.1	1	1	0	0
3351	129	bitext	[bitexts]	0.1	1	2	2	2
3352	129	prefix*-stem-suffix*	[prefix*-stem-suffix*]	0.1	1	1	0	0
3353	129	scope	[scope]	0.1	1	2	1	1
3354	129	message	[message]	0.1	1	2	1	1
3355	129	age	[age]	0.1	1	1	0	0
3356	129	tendency	[tendency]	0.1	1	2	1	1
3357	129	likelihood	[likelihood]	0.1	1	2	1	1
3358	129	spot	[SPoT]	0.1	1	2	1	1
3359	129	aid	[aid]	0.1	1	1	0	0
3360	129	other	[others]	0.1	1	1	0	0
3361	129	aim	[aim]	0.1	1	2	1	1
3362	129	movement	[movements]	0.1	1	2	2	2
3363	129	weakness	[weaknesses]	0.1	1	1	0	0
3364	129	.62%	[.62%]	0.1	1	1	0	0
3365	129	aps	[APs]	0.1	1	2	1	1
3366	129	sampling	[sampling]	0.1	1	1	0	0
3367	129	phone-string	[phone-string]	0.1	1	1	0	0
3368	129	strength	[strength]	0.1	1	5	4	1
3369	129	absence	[absence]	0.1	1	1	0	0
3370	129	pre-exist	[pre-existing]	0.1	1	1	0	0
3371	129	CCRs	[CCRs]	0.1	1	1	0	0
3372	129	modeling	[modeling]	0.1	1	2	1	1
3373	129	phone	[phone]	0.1	1	1	0	0
3374	129	story-telling	[story-telling]	0.1	1	1	0	0
3375	129	navy	[Navy]	0.1	1	1	0	0
3376	129	sublanguage	[sublanguage]	0.1	1	2	1	1
3377	129	typing-algorithm	[typing-algorithm]	0.1	1	1	0	0
3378	129	assignment	[assignment]	0.1	1	2	2	2
3379	129	increase	[increase]	0.1	1	2	2	2
3380	129	documentation	[documentation]	0.1	1	1	0	0
3381	129	deal	[deal]	0.1	1	2	2	2
3382	129	completion	[completion]	0.1	1	1	0	0
3383	129	NTHU's	[NTHU's]	0.1	1	1	0	0
3384	129	ACL	[ACL]	0.1	1	1	0	0
3385	129	procedure	[procedure]	0.1	1	1	0	0
3386	129	parallelism	[parallelism]	0.1	1	1	0	0
3387	129	usability	[usability]	0.1	1	1	0	0
3388	129	operate	[operating]	0.1	1	1	0	0
3389	129	paragraph	[paragraph]	0.1	1	1	0	0
3390	129	table	[tables]	0.1	1	1	0	0
3391	129	novice	[novice]	0.1	1	1	0	0
3392	129	eigenvector	[eigenvectors]	0.1	1	2	2	2
3393	129	APT	[APT]	0.1	1	1	0	0
3394	129	APs	[APs]	0.1	1	2	1	1
3395	129	cbmt	[CBMT]	0.1	1	1	0	0
3396	129	Paramax	[Paramax]	0.1	1	2	1	1
3397	129	Transfer-Driven	[Transfer-Driven]	0.1	1	1	0	0
3398	129	communicator	[Communicator]	0.1	1	2	1	1
3399	129	breaking-off	[breaking-off]	0.1	1	1	0	0
3400	129	weather	[weather]	0.1	1	2	2	2
3401	129	guide	[guide]	0.1	1	1	0	0
3402	129	concatenation	[concatenation]	0.1	1	2	2	2
3403	129	recall	[recall]	0.1	1	3	2	1
3404	129	token	[token]	0.1	1	2	1	1
3405	129	emotion	[Emotions]	0.1	1	1	0	0
3406	129	HTML	[HTML]	0.1	1	2	1	1
3407	129	NOMAD	[NOMAD]	0.1	1	1	0	0
3408	129	Gale	[Gale]	0.1	1	1	0	0
3409	129	non-NP-antecedent	[non-NP-antecedents]	0.1	1	1	0	0
3410	129	SCF	[SCF]	0.1	1	2	1	1
3411	130	minor extension of component	[minor extensions of components]	0.0	4	1	1	1
3412	130	use wsd evaluation methodology	[using WSD evaluation methodology]	0.0	4	1	1	1
3413	130	disambiguation scheme	[disambiguation scheme]	0.0	2	1	1	1
3414	130	term of simplicity	[terms of simplicity]	0.0	3	1	1	1
3415	130	achieve similar performance to complex mixture	[achieving similar performance to complex mixtures]	0.0	6	1	1	1
3416	130	descriptor	[descriptors]	0.0	1	1	1	1
3417	130	utilize example effectively	[utilizing examples effectively]	0.0	3	1	1	1
3418	130	use a set	[using a set]	0.0	3	1	1	1
3419	130	character level	[character level]	0.0	2	2	2	1
3420	130	several key factor	[several key factors]	0.0	3	1	1	1
3421	130	superiority of this method over method	[superiority of this method over methods]	0.0	6	1	1	1
3422	130	inflective language such as Czech	[inflective languages such as Czech]	0.0	5	1	1	1
3423	130	performance sentence	[performance sentences]	0.0	2	1	1	1
3424	130	frame encode the meaning	[frame encoding the meaning]	0.0	4	1	1	1
3425	130	integral part	[integral part]	0.0	2	1	1	1
3426	130	problem extract information	[problem extracting information]	0.0	3	1	1	1
3427	130	corpus analysis	[corpus analysis]	0.0	2	1	1	1
3428	130	cost of hand-crafting generation system	[cost of hand-crafting generation systems]	0.0	5	1	1	1
3429	130	utterance about a task	[utterance about a task]	0.0	4	1	1	1
3430	130	avoidance	[avoidance]	0.0	1	1	1	1
3431	130	social relationship email .g	[social relationship e .g]	0.0	4	1	1	1
3432	130	investigate the combinatorics of indexation	[investigating the combinatorics of indexation]	0.0	5	1	1	1
3433	130	structure string	[structures strings]	0.0	2	1	1	1
3434	130	extraction of translation lexicon	[extraction of translation lexicons]	0.0	4	1	1	1
3435	130	correct this error	[correcting these errors]	0.0	3	1	1	1
3436	130	parsing final language	[parsing final language]	0.0	3	1	1	1
3437	130	Representative sample	[Representative samples]	0.0	2	1	1	1
3438	130	approach induce a grammar	[approach inducing a grammar]	0.0	4	1	1	1
3439	130	13% relative decrease	[13% relative decrease]	0.0	3	1	1	1
3440	130	combination with SMT model	[combination with SMT models]	0.0	4	1	1	1
3441	130	c	[C]	0.0	1	1	1	1
3442	130	accurate disambiguation	[accurate disambiguation]	0.0	2	1	1	1
3443	130	system support	[system support]	0.0	2	1	1	1
3444	130	chart representation	[chart representations]	0.0	2	1	1	1
3445	130	amount of internal variation superior	[amount of internal variation superior]	0.0	5	1	1	1
3446	130	integral process in understanding	[integral process in understanding]	0.0	4	1	1	1
3447	130	character recognition model	[character recognition model]	0.0	3	1	1	1
3448	130	machine translation system CCLINC	[machine translation system CCLINC]	0.0	4	1	1	1
3449	130	domain model with classifier	[domain models with classifiers]	0.0	4	1	1	1
3450	130	evaluation result involve automatic extraction	[evaluation results involving automatic extraction]	0.0	5	1	1	1
3451	130	use processor aps for language translation	[using processors APs for language translation]	0.0	6	1	1	1
3452	130	various tagging strategy	[various tagging strategies]	0.0	3	1	1	1
3453	130	grammatical knowledge	[grammatical knowledge]	0.0	2	1	1	1
3454	130	i efficient parsing a verb language	[i efficient parsing a verb language]	0.0	6	1	1	1
3455	130	large-scale acquisition of information	[large-scale acquisition of information]	0.0	4	1	1	1
3456	130	news archive	[news archive]	0.0	2	1	1	1
3457	130	high cost of generation system	[high cost of generation systems]	0.0	5	1	1	1
3458	130	cover the domain	[covering the domain]	0.0	3	1	1	1
3459	130	identification of 2-character name without title	[identification of 2-character names without title]	0.0	6	1	1	1
3460	130	large text-understanding system	[large text-understanding systems]	0.0	3	1	1	1
3461	130	use associative processor for language translation	[using associative processors for language translation]	0.0	6	1	1	1
3462	130	Arabic word segmentation system	[Arabic word segmentation system]	0.0	4	1	1	1
3463	130	efficient parsing of Korean	[efficient parsing of Korean]	0.0	4	1	1	1
3464	130	edge with a valid interpretation	[edges with a valid interpretation]	0.0	5	1	1	1
3465	130	verb class	[verb classes]	0.0	2	1	1	1
3466	130	markov model cdhmm	[Markov models CDHMM]	0.0	3	1	1	1
3467	130	construction of model	[construction of models]	0.0	3	1	1	1
3468	130	97% match accuracy	[97% match accuracy]	0.0	3	1	1	1
3469	130	patent claim sentence	[patent claim sentences]	0.0	3	1	1	1
3470	130	classical grammatical formalism	[classical grammatical formalisms]	0.0	3	1	1	1
3471	130	syntactic structure for speech act	[syntactic structure for speech acts]	0.0	5	1	1	1
3472	130	idus	[IDUS]	0.0	1	1	1	1
3473	130	use include jointly conditioning on word	[use including jointly conditioning on words]	0.0	6	1	1	1
3474	130	discover parallel sentence in comparable corpus	[discovering parallel sentences in comparable corpora]	0.0	6	1	1	1
3475	130	area of research	[areas of research]	0.0	3	1	1	1
3476	130	start with a small corpus word	[starting with a small corpus words]	0.0	6	1	1	1
3477	130	numeric-valued attribute such as size	[numeric-valued attributes such as size]	0.0	5	1	1	1
3478	130	achieve good baseline result	[achieving good baseline results]	0.0	4	1	1	1
3479	130	topmost of the edge adjacent	[topmost of the edges adjacent]	0.0	5	1	1	1
3480	130	exploit a corpus	[exploiting a corpus]	0.0	3	1	1	1
3481	130	weighting of phrase translation	[weighting of phrase translations]	0.0	4	1	1	1
3482	130	candidate the correct structure	[candidates the correct structure]	0.0	4	1	1	1
3483	130	grammar TAG	[grammars TAG]	0.0	2	1	1	1
3484	130	source of constraint	[source of constraint]	0.0	3	1	1	1
3485	130	,449 word token	[,449 word tokens]	0.0	3	1	1	1
3486	130	method describe inflection	[method describing inflection]	0.0	3	1	1	1
3487	130	usually the beginning	[usually the beginning]	0.0	3	1	1	1
3488	130	performance enhancing technique	[performance enhancing techniques]	0.0	3	1	1	1
3489	130	segment of tape-recorded description	[segments of tape-recorded descriptions]	0.0	4	1	1	1
3490	130	usual linguistic introspection	[usual linguistic introspection]	0.0	3	1	1	1
3491	130	arbitrary set	[arbitrary set]	0.0	2	1	1	1
3492	130	sentence condensation	[sentence condensation]	0.0	2	1	1	1
3493	130	computational theory of plausible reasoning	[computational theory of plausible reasoning]	0.0	5	1	1	1
3494	130	use word-trigram recognition require manual transcription	[using word-trigram recognition requiring manual transcription]	0.0	6	1	1	1
3495	130	lexicon specification	[lexicon specification]	0.0	2	1	1	1
3496	130	SLM	[SLM]	0.0	1	1	1	1
3497	130	part of grammar parsing	[part of grammar parsing]	0.0	4	1	1	1
3498	130	efficiency gain	[efficiency gains]	0.0	2	1	1	1
3499	130	Air	[Air]	0.0	1	2	2	1
3500	130	temporal relation	[temporal relations]	0.0	2	1	1	1
3501	130	block identity	[block identities]	0.0	2	1	1	1
3502	130	purpose a node with great	[purposes a node with greater]	0.0	5	1	1	1
3503	130	woman	[women]	0.0	1	2	2	1
3504	130	effect on the cluster	[effect on the clusters]	0.0	4	1	1	1
3505	130	natural language in setting	[natural language in settings]	0.0	4	1	1	1
3506	130	mostly woman	[mostly women]	0.0	2	1	1	1
3507	130	such as size	[such as size]	0.0	3	1	1	1
3508	130	pint	[pints]	0.0	1	1	1	1
3509	130	increase of 35%	[increase of 35%]	0.0	3	1	1	1
3510	130	significant improvement for speaker adaptation	[significant improvement for speaker adaptation]	0.0	5	1	1	1
3511	130	correct paraphrase mean equivalence	[correct paraphrases meaning equivalence]	0.0	4	1	1	1
3512	130	focus on consideration	[focus on considerations]	0.0	3	1	1	1
3513	130	3 good indicator	[3 better indicators]	0.0	3	1	1	1
3514	130	task build the editor	[task building the editor]	0.0	4	1	1	1
3515	130	DCG formalism	[DCG formalism]	0.0	2	1	1	1
3516	130	generation of text through transformation	[generation of text through transformation]	0.0	5	1	1	1
3517	130	language learner	[language learners]	0.0	2	1	1	1
3518	130	two problem numeric-valued attribute	[two problems numeric-valued attributes]	0.0	4	1	1	1
3519	130	study on communication	[study on communications]	0.0	3	1	1	1
3520	130	state-of-the-art Chinese word sense	[state-of-the-art Chinese word sense]	0.0	4	1	1	1
3521	130	maximum a estimation	[maximum a estimation]	0.0	3	1	1	1
3522	130	new algorithm for the approach	[new algorithm for the approach]	0.0	5	1	1	1
3523	130	only area	[only area]	0.0	2	1	1	1
3524	130	new linguistic concept	[new linguistic concepts]	0.0	3	1	1	1
3525	130	topmost of the edge	[topmost of the edges]	0.0	4	1	1	1
3526	130	method construct word vector lsa-based method	[methods constructing word vectors LSA-based methods]	0.0	6	1	1	1
3527	130	qualitative constraint	[qualitative constraints]	0.0	2	1	1	1
3528	130	use parallel corpus a available resource	[using parallel corpora a available resource]	0.0	6	1	1	1
3529	130	include conditioning on multiple consecutive word	[including conditioning on multiple consecutive words]	0.0	6	1	1	1
3530	130	Markov probability of a chain	[Markov probability of a chain]	0.0	5	1	1	1
3531	130	phone n-gram	[phone n-gram]	0.0	2	1	1	1
3532	130	build spelling-checker	[building spelling-checkers]	0.0	2	1	1	1
3533	130	Prolog cf	[Prolog cf]	0.0	2	1	1	1
3534	130	conversational cue	[conversational cues]	0.0	2	1	1	1
3535	130	core module	[core modules]	0.0	2	1	1	1
3536	130	Rapid system development via acquisition	[Rapid system development via acquisition]	0.0	5	1	1	1
3537	130	repository of lexical information	[repository of lexical information]	0.0	4	1	1	1
3538	130	real-valued feature email .g	[real-valued features e .g]	0.0	4	1	1	1
3539	130	Rapid system development to domain	[Rapid system development to domains]	0.0	5	1	1	1
3540	130	unsegmented text	[unsegmented texts]	0.0	2	1	1	1
3541	130	repository of information	[repository of information]	0.0	3	1	1	1
3542	130	similarity-driven transfer System	[Similarity-driven Transfer System]	0.0	3	1	1	1
3543	130	task of interpretation	[task of interpretation]	0.0	3	1	1	1
3544	130	lexical approach	[lexical approach]	0.0	2	1	1	1
3545	130	two contribution	[two contributions]	0.0	2	1	1	1
3546	130	useful software infrastructure	[useful software infrastructure]	0.0	3	1	1	1
3547	130	overall summarization quality	[Overall summarization quality]	0.0	3	1	1	1
3548	130	SRH	[SRH]	0.0	1	1	1	1
3549	130	off-the-shelf classifier	[off-the-shelf classifiers]	0.0	2	1	1	1
3550	130	restriction on the category	[restrictions on the categories]	0.0	4	1	1	1
3551	130	generate the parse	[generating the parse]	0.0	3	1	1	1
3552	130	system output text	[systems outputting texts]	0.0	3	1	1	1
3553	130	underlying word alignment	[underlying word alignment]	0.0	3	1	1	1
3554	130	CSR speech Recognition	[CSR Speech Recognition]	0.0	3	1	1	1
3555	130	direction	[directions]	0.0	1	1	1	1
3556	130	human language technology	[human language technology]	0.0	3	1	1	1
3557	130	understanding of natural language	[understanding of natural language]	0.0	4	1	1	1
3558	130	qualitative user evaluation	[qualitative user evaluation]	0.0	3	1	1	1
3559	130	language application	[language applications]	0.0	2	1	1	1
3560	130	respond	[responding]	0.0	1	1	1	1
3561	130	October '91 test set	[October '91 test set]	0.0	4	1	1	1
3562	130	underspecified representation	[underspecified representation]	0.0	2	1	1	1
3563	130	idus document Understanding System	[IDUS Document Understanding System]	0.0	4	1	1	1
3564	130	algorithm for unification	[algorithms for unification]	0.0	3	1	1	1
3565	130	corpus of multiple translation	[corpus of multiple translations]	0.0	4	1	1	1
3566	130	collect statistic in a corpus	[collecting statistics in a corpus]	0.0	5	1	1	1
3567	130	utterance from the target speaker	[utterances from the target speaker]	0.0	5	1	1	1
3568	130	notation	[notations]	0.0	1	1	1	1
3569	130	result on addressee identification in four-participant	[results on addressee identification in four-participants]	0.0	6	1	1	1
3570	130	central problem of word sense	[central problem of word sense]	0.0	5	1	1	1
3571	130	simple paraphrasing technique	[simple paraphrasing technique]	0.0	3	1	1	1
3572	130	feature selection	[feature selection]	0.0	2	1	1	1
3573	130	take advantage of the pattern	[taking advantage of the patterns]	0.0	5	1	1	1
3574	130	gather information	[gathering information]	0.0	2	1	1	1
3575	130	identify a predicate	[identifying a predicate]	0.0	3	1	1	1
3576	130	Dynamic Phrasal Lexicon	[Dynamic Phrasal Lexicon]	0.0	3	1	1	1
3577	130	Recent advance	[Recent advances]	0.0	2	1	1	1
3578	130	single frame encode the meaning	[single frame encoding the meaning]	0.0	5	1	1	1
3579	130	high-precision rule	[high-precision rules]	0.0	2	1	1	1
3580	130	certain aspect of language learning	[certain aspects of language learning]	0.0	5	1	1	1
3581	130	advantage of detailed information	[advantage of detailed information]	0.0	4	1	1	1
3582	130	classical mle reestimation algorithm	[classical MLE reestimation algorithms]	0.0	4	1	1	1
3583	130	18 .3m character	[18 .3M characters]	0.0	3	1	1	1
3584	130	model use processor	[model using processors]	0.0	3	1	1	1
3585	130	gain from information about meeting context	[gain from information about meeting context]	0.0	6	1	1	1
3586	130	task-based evaluation	[Task-based evaluation]	0.0	2	1	1	1
3587	130	relationship rank	[relationship ranks]	0.0	2	1	1	1
3588	130	additional index	[additional indices]	0.0	2	1	1	1
3589	130	previous single tagging result	[previous single tagging result]	0.0	4	1	1	1
3590	130	make exploratory corpus analysis accessible	[making exploratory corpus analysis accessible]	0.0	5	1	1	1
3591	130	few reading	[fewer readings]	0.0	2	1	1	1
3592	130	such as size perspective-taking	[such as size perspective-taking]	0.0	4	1	1	1
3593	130	range of text	[range of texts]	0.0	3	1	1	1
3594	130	pattern of the parallelism	[patterns of the parallelism]	0.0	4	1	1	1
3595	130	reference resolution	[reference resolution]	0.0	2	1	1	1
3596	130	extract specific information	[extracting specific information]	0.0	3	1	1	1
3597	130	small size	[small size]	0.0	2	1	1	1
3598	130	Diego	[Diego]	0.0	1	1	1	1
3599	130	structure for speech act	[structure for speech acts]	0.0	4	1	1	1
3600	130	approach employ n-gram model	[approach employing n-gram models]	0.0	4	1	1	1
3601	130	improvement in precision	[improvement in precision]	0.0	3	1	1	1
3602	130	benefit from the integration	[benefit from the integration]	0.0	4	1	1	1
3603	130	depend on the requirement	[depending on the requirements]	0.0	4	1	1	1
3604	130	conventional word-trigram recognition	[conventional word-trigram recognition]	0.0	3	1	1	1
3605	130	noun phrase reiteration	[noun phrase reiteration]	0.0	3	1	1	1
3606	130	contribute match in the sentence	[contributing matches in the sentence]	0.0	5	1	1	1
3607	130	geometric model of the scene	[geometric model of the scene]	0.0	5	1	1	1
3608	130	noisy portion	[noisy portion]	0.0	2	1	1	1
3609	130	related area	[related areas]	0.0	2	1	1	1
3610	130	non-monotonic reasoning	[non-monotonic reasoning]	0.0	2	1	1	1
3611	130	neighbour method	[neighbour methods]	0.0	2	1	1	1
3612	130	Hansards	[Hansards]	0.0	1	1	1	1
3613	130	speaker leeway	[speaker leeway]	0.0	2	1	1	1
3614	130	face-to-face meeting use Network classifier	[face-to-face meetings using Network classifiers]	0.0	5	1	1	1
3615	130	expensive part of grammar parsing	[expensive part of grammar parsing]	0.0	5	1	1	1
3616	130	start	[starting]	0.0	1	1	1	1
3617	130	start with a corpus 100 word	[starting with a corpus 100 words]	0.0	6	1	1	1
3618	130	improvement over all	[improvement over all]	0.0	3	1	1	1
3619	130	problem numeric-valued attribute	[problems numeric-valued attributes]	0.0	3	1	1	1
3620	130	effect the reduction of analysis cost	[effects the reduction of analysis cost]	0.0	6	1	1	1
3621	130	result program	[resulting program]	0.0	2	1	1	1
3622	130	hand-crafted approach	[hand-crafted approaches]	0.0	2	1	1	1
3623	130	recognition accuracy	[recognition accuracy]	0.0	2	1	1	1
3624	130	enumerate all possible indexing	[enumerating all possible indexings]	0.0	4	1	1	1
3625	130	continuous speech recognition of a language	[continuous speech recognition of a language]	0.0	6	1	1	1
3626	130	plausible interpretation	[plausible interpretation]	0.0	2	1	1	1
3627	130	gather a number of abstract	[gathering a number of abstracts]	0.0	5	1	1	1
3628	130	unification algorithm avoidance	[unification algorithms avoidance]	0.0	3	1	1	1
3629	130	develop time approximation	[developing time approximations]	0.0	3	1	1	1
3630	130	Computer	[Computer]	0.0	1	1	1	1
3631	130	acquire sense-tagged datum	[acquiring sense-tagged data]	0.0	3	1	1	1
3632	130	enter	[entering]	0.0	1	1	1	1
3633	130	develop time approximation for this computation	[developing time approximations for these computations]	0.0	6	1	1	1
3634	130	represent the meaning of a proposition	[representing the meaning of a proposition]	0.0	6	1	1	1
3635	130	general domain	[general domains]	0.0	2	1	1	1
3636	130	two problem attribute	[two problems attributes]	0.0	3	1	1	1
3637	130	estimation of markov model	[estimation of Markov models]	0.0	4	1	1	1
3638	130	other nlp problem	[other NLP problems]	0.0	3	1	1	1
3639	130	part-of-speech tag sequence	[part-of-speech tag sequences]	0.0	3	1	1	1
3640	130	drastic generalization corresponding to arc	[drastic generalization corresponding to arcs]	0.0	5	1	1	1
3641	130	introspection	[introspection]	0.0	1	1	1	1
3642	130	distribution of grammatical knowledge	[distribution of grammatical knowledge]	0.0	4	1	1	1
3643	130	employ a approach	[employing a approach]	0.0	3	1	1	1
3644	130	interesting ambiguity	[interesting ambiguities]	0.0	2	1	1	1
3645	130	use class-dependent interpolation	[using class-dependent interpolation]	0.0	3	1	1	1
3646	130	large study of expression	[larger study of expressions]	0.0	4	1	1	1
3647	130	robust parsing language	[Robust parsing language]	0.0	3	1	1	1
3648	130	involve automatic extraction of translation lexicon	[involving automatic extraction of translation lexicons]	0.0	6	1	1	1
3649	130	datum object	[data objects]	0.0	2	1	1	1
3650	130	submanifold	[submanifold]	0.0	1	1	1	1
3651	130	relative decrease	[relative decrease]	0.0	2	1	1	1
3652	130	classical formalism	[classical formalisms]	0.0	2	1	1	1
3653	130	increase of 35% in mrr	[increase of 35% in MRR]	0.0	5	1	1	1
3654	130	contribution to vocabulary	[contributions to vocabulary]	0.0	3	1	1	1
3655	130	lr parser	[LR parser]	0.0	2	1	1	1
3656	130	more frequently individually good predictor	[more frequently individually good predictors]	0.0	5	1	1	1
3657	130	interesting combination	[interesting combination]	0.0	2	1	1	1
3658	130	ideal aim for interpretation	[ideal aiming for interpretations]	0.0	4	1	1	1
3659	130	approach to restricted-domain parsing	[approach to restricted-domain parsing]	0.0	4	1	1	1
3660	130	revision support	[revision support]	0.0	2	1	1	1
3661	130	theory 1988/	[theory 1988/]	0.0	2	1	1	1
3662	130	computer program SPRINT	[computer program SPRINT]	0.0	3	1	1	1
3663	130	word with meaning	[words with meanings]	0.0	3	1	1	1
3664	130	efficient parser	[efficient parser]	0.0	2	1	1	1
3665	130	treatment of Montague semantics	[treatment of Montague semantics]	0.0	4	1	1	1
3666	130	correct paraphrase	[correct paraphrases]	0.0	2	1	1	1
3667	130	iii Rapid system development	[iii Rapid system development]	0.0	4	1	1	1
3668	130	require less development time	[requiring less development time]	0.0	4	1	1	1
3669	130	generality available	[generality available]	0.0	2	1	1	1
3670	130	use a state-of-the-art word sense	[Using a state-of-the-art word sense]	0.0	5	1	1	1
3671	130	result edge	[resulting edges]	0.0	2	1	1	1
3672	130	contrast to work	[contrast to work]	0.0	3	1	1	1
3673	130	find the language	[Finding the language]	0.0	3	1	1	1
3674	130	accuracy rate of disambiguation	[accuracy rate of disambiguation]	0.0	4	1	1	1
3675	130	view for extensional reference	[view for extensional reference]	0.0	4	1	1	1
3676	130	entropy word	[entropy word]	0.0	2	1	1	1
3677	130	for scoring set of concept	[for scoring sets of concepts]	0.0	5	1	1	1
3678	130	Chinese non-parallel newspaper corpus	[Chinese non-parallel newspaper corpora]	0.0	4	1	1	1
3679	130	edge as in conventional treatment	[edges as in conventional treatments]	0.0	5	1	1	1
3680	130	system output unsegmented text	[systems outputting unsegmented texts]	0.0	4	1	1	1
3681	130	social relationship	[social relationship]	0.0	2	1	1	1
3682	130	avoidance of copying of subgraph	[avoidance of copying of subgraphs]	0.0	5	1	1	1
3683	130	entry from machine dictionary	[entries from machine dictionaries]	0.0	4	1	1	1
3684	130	language generation system	[language generation system]	0.0	3	1	1	1
3685	130	semantic domain	[semantic domains]	0.0	2	1	1	1
3686	130	identification of Chinese name without title	[identification of Chinese names without title]	0.0	6	1	1	1
3687	130	grammaticality sentence equivalence	[grammaticality sentences equivalence]	0.0	3	1	1	1
3688	130	goal sound dialog system	[goal sounding dialog systems]	0.0	4	1	1	1
3689	130	efficient Prolog cf	[efficient Prolog cf]	0.0	3	1	1	1
3690	130	process in understanding	[process in understanding]	0.0	3	1	1	1
3691	130	m-th	[m-th]	0.0	1	1	1	1
3692	130	confidence measure	[confidence measures]	0.0	2	1	1	1
3693	130	proper treatment	[proper treatment]	0.0	2	1	1	1
3694	130	third	[third]	0.0	1	1	1	1
3695	130	small corpus 100 word	[small corpus 100 words]	0.0	4	1	1	1
3696	130	utterance from five site	[utterances from five sites]	0.0	4	1	1	1
3697	130	approach to q/a	[approach to Q/A]	0.0	3	1	1	1
3698	130	CDD	[CDD]	0.0	1	1	1	1
3699	130	move in light	[move in light]	0.0	3	1	1	1
3700	130	many property	[many properties]	0.0	2	1	1	1
3701	130	efficient parsing language with case marker	[efficient parsing language with case markers]	0.0	6	1	1	1
3702	130	employ n-gram model	[employing n-gram models]	0.0	3	1	1	1
3703	130	state Arabic stemmer	[state Arabic stemmer]	0.0	3	1	1	1
3704	130	reliable basis for the acquisition	[reliable basis for the acquisition]	0.0	5	1	1	1
3705	130	plan COMPLEX a lexicon	[planning COMPLEX a lexicon]	0.0	4	1	1	1
3706	130	word-sense of word with meaning	[word-senses of words with meanings]	0.0	5	1	1	1
3707	130	Cooccurrence	[Cooccurrence]	0.0	1	1	1	1
3708	130	improvement in paraphrase classification accuracy	[improvement in paraphrase classification accuracy]	0.0	5	1	1	1
3709	130	unigram model	[unigram model]	0.0	2	1	1	1
3710	130	speech tagging	[speech tagging]	0.0	2	1	1	1
3711	130	translation example	[translation examples]	0.0	2	1	1	1
3712	130	SemCor	[SemCor]	0.0	1	1	1	1
3713	130	corpus exploration tool	[corpus exploration tool]	0.0	3	1	1	1
3714	130	Hierarchical Graph Kernel	[Hierarchical Graph Kernel]	0.0	3	1	1	1
3715	130	several experiment corresponding to domain	[Several experiments corresponding to domains]	0.0	5	1	1	1
3716	130	choice for the parser	[choice for the parser]	0.0	4	1	1	1
3717	130	honorifics word	[honorifics words]	0.0	2	1	1	1
3718	130	large list	[large list]	0.0	2	1	1	1
3719	130	apply statistical model	[applying statistical models]	0.0	3	1	1	1
3720	130	employ a traditional approach	[employing a traditional approach]	0.0	4	1	1	1
3721	130	information related to a scenario	[information related to a scenario]	0.0	5	1	1	1
3722	130	proprietary Arabic stemmer	[proprietary Arabic stemmer]	0.0	3	1	1	1
3723	130	speech recognition application	[speech recognition applications]	0.0	3	1	1	1
3724	130	apply a set	[applying a set]	0.0	3	1	1	1
3725	130	definition a concept	[definition a concept]	0.0	3	1	1	1
3726	130	pose problem	[posing problems]	0.0	2	1	1	1
3727	130	corresponding to arc in a chart	[corresponding to arcs in a chart]	0.0	6	1	1	1
3728	130	make corpus analysis accessible to researcher	[making corpus analysis accessible to researchers]	0.0	6	1	1	1
3729	130	retrieve example similar	[retrieving examples similar]	0.0	3	1	1	1
3730	130	key factor behind this surprising finding	[key factors behind this surprising finding]	0.0	6	1	1	1
3731	130	addressee identification in four-participant	[addressee identification in four-participants]	0.0	4	1	1	1
3732	130	linguistic parser/generator for LFG	[linguistic parser/generator for LFG]	0.0	4	1	1	1
3733	130	mixture observation density	[mixture observation densities]	0.0	3	1	1	1
3734	130	word-trigram recognition require manual transcription	[word-trigram recognition requiring manual transcription]	0.0	5	1	1	1
3735	130	broad-coverage principle-based parser	[broad-coverage principle-based parser]	0.0	3	1	1	1
3736	130	representation the Dynamic Phrasal Lexicon	[representation the Dynamic Phrasal Lexicon]	0.0	5	1	1	1
3737	130	transfer System	[Transfer System]	0.0	2	1	1	1
3738	130	node with in-degree great	[node with in-degree greater]	0.0	4	1	1	1
3739	130	represent the meaning	[representing the meaning]	0.0	3	1	1	1
3740	130	standard floppy	[standard floppy]	0.0	2	1	1	1
3741	130	web-based writing	[Web-based writing]	0.0	2	1	1	1
3742	130	also in the related area	[also in the related areas]	0.0	5	1	1	1
3743	130	University	[University]	0.0	1	1	1	1
3744	130	shallow linguistic feature	[shallow linguistic features]	0.0	3	1	1	1
3745	130	level of representation	[levels of representation]	0.0	3	1	1	1
3746	130	adequate explanation of discourse phenomenon	[adequate explanation of discourse phenomena]	0.0	5	1	1	1
3747	130	he/she/man/woman	[he/she/man/woman]	0.0	1	1	1	1
3748	130	knowledgebase	[knowledgebases]	0.0	1	1	1	1
3749	130	workability	[workability]	0.0	1	1	1	1
3750	130	gain from information	[gain from information]	0.0	3	1	1	1
3751	130	prior in conditional model	[priors in conditional models]	0.0	4	1	1	1
3752	130	decoder implementation	[decoder implementations]	0.0	2	1	1	1
3753	130	representation USR	[representation USR]	0.0	2	1	1	1
3754	130	predict boundary	[predicting boundaries]	0.0	2	1	1	1
3755	130	phrase translation from alignment	[phrase translations from alignments]	0.0	4	1	1	1
3756	130	yield a treatment	[yielding a treatment]	0.0	3	1	1	1
3757	130	role of information	[role of information]	0.0	3	1	1	1
3758	130	consistency problem	[consistency problem]	0.0	2	1	1	1
3759	130	attractive property	[attractive properties]	0.0	2	1	1	1
3760	130	i robust parsing	[i Robust parsing]	0.0	3	1	1	1
3761	130	only 12 training speaker	[only 12 training speakers]	0.0	4	1	1	1
3762	130	mutually equivalent reading	[mutually equivalent readings]	0.0	3	1	1	1
3763	130	evaluation result	[evaluation results]	0.0	2	1	1	1
3764	130	subset of database	[subsets of database]	0.0	3	1	1	1
3765	130	copying of subgraph	[copying of subgraphs]	0.0	3	1	1	1
3766	130	little corpus	[little corpus]	0.0	2	1	1	1
3767	130	adverb	[adverbs]	0.0	1	1	1	1
3768	130	Natural language processing nlp	[Natural Language Processing NLP]	0.0	4	1	1	1
3769	130	exist spelling-checker	[existing spelling-checkers]	0.0	2	1	1	1
3770	130	constraint as level	[constraints as levels]	0.0	3	1	1	1
3771	130	method discover parallel sentence	[method discovering parallel sentences]	0.0	4	1	1	1
3772	130	kind of role system	[kind of Role System]	0.0	4	1	1	1
3773	130	social rank	[social ranks]	0.0	2	1	1	1
3774	130	two grammatical formalism	[two grammatical formalisms]	0.0	3	1	1	1
3775	130	relationship rank of the referent	[relationship ranks of the referents]	0.0	5	1	1	1
3776	130	newspaper article on missile	[newspaper articles on missiles]	0.0	4	1	1	1
3777	130	mle reestimation algorithm	[MLE reestimation algorithms]	0.0	3	1	1	1
3778	130	evaluate answer to definition question	[evaluating answers to definition questions]	0.0	5	1	1	1
3779	130	concept of the linguistic theory	[concepts of the linguistic theory]	0.0	5	1	1	1
3780	130	interval	[intervals]	0.0	1	1	1	1
3781	130	wsi algorithm	[WSI algorithm]	0.0	2	1	1	1
3782	130	sparsity	[sparsity]	0.0	1	1	1	1
3783	130	use semantic category	[using semantic categories]	0.0	3	1	1	1
3784	130	Markov Model	[Markov Model]	0.0	2	1	1	1
3785	130	simple unsupervised technique	[simple unsupervised technique]	0.0	3	1	1	1
3786	130	amount of ambiguity	[amount of ambiguity]	0.0	3	1	1	1
3787	130	log(	[log(]	0.0	1	1	1	1
3788	130	automate aspect of language learning	[automating aspects of language learning]	0.0	5	1	1	1
3789	130	computer with graphic display	[computers with graphics displays]	0.0	4	1	1	1
3790	130	build corpus from HTML document	[building corpus from HTML documents]	0.0	5	1	1	1
3791	130	part of a large study	[part of a larger study]	0.0	5	1	1	1
3792	130	only 40 utterance	[only 40 utterances]	0.0	3	1	1	1
3793	130	capability in dialogue system	[capabilities in dialogue systems]	0.0	4	1	1	1
3794	130	acquire stem	[acquiring stems]	0.0	2	1	1	1
3795	130	scope information	[scope information]	0.0	2	1	1	1
3796	130	assess development	[assessing developments]	0.0	2	1	1	1
3797	130	equivalent reading	[equivalent readings]	0.0	2	1	1	1
3798	130	human understanding of language	[human understanding of language]	0.0	4	1	1	1
3799	130	output system entity annotation	[outputs systems entity annotations]	0.0	4	1	1	1
3800	130	.5% word	[.5% word]	0.0	2	1	1	1
3801	130	copying of unmodified subgraph	[copying of unmodified subgraphs]	0.0	4	1	1	1
3802	130	parsing accuracy rate	[parsing accuracy rate]	0.0	3	1	1	1
3803	130	polynomial time solution	[polynomial time solution]	0.0	3	1	1	1
3804	130	specific domain	[specific domain]	0.0	2	1	1	1
3805	130	two list	[two lists]	0.0	2	1	1	1
3806	130	application of the technique	[application of the techniques]	0.0	4	1	1	1
3807	130	advantage of this approach	[advantage of this approach]	0.0	4	1	1	1
3808	130	.0% improvement over system	[.0% improvement over system]	0.0	4	1	1	1
3809	130	log-linear model	[log-linear models]	0.0	2	1	1	1
3810	130	dialogue datum	[dialogue data]	0.0	2	1	1	1
3811	130	visualization method	[visualization method]	0.0	2	1	1	1
3812	130	work harvesting bitext	[work harvesting bitexts]	0.0	3	1	1	1
3813	130	interaction with datum source	[interaction with data sources]	0.0	4	1	1	1
3814	130	three technique	[three techniques]	0.0	2	1	1	1
3815	130	case-based reasoning	[case-based reasoning]	0.0	2	1	1	1
3816	130	Entity NE	[Entity NE]	0.0	2	1	1	1
3817	130	chart parsing instantiation	[chart Parsing instantiation]	0.0	3	1	1	1
3818	130	dynamic challenge	[dynamic challenge]	0.0	2	1	1	1
3819	130	effect the reduction	[effects the reduction]	0.0	3	1	1	1
3820	130	360K floppy	[360K floppy]	0.0	2	1	1	1
3821	130	paradigm	[paradigm]	0.0	1	2	2	1
3822	130	parallel sentence in corpus	[parallel sentences in corpora]	0.0	4	1	1	1
3823	130	great deal of language text	[great deal of language texts]	0.0	5	1	1	1
3824	130	12,000 utterance of spontaneous speech	[12,000 utterances of spontaneous speech]	0.0	5	1	1	1
3825	130	coverage grammar	[coverage grammar]	0.0	2	1	1	1
3826	130	unrestricted text	[unrestricted texts]	0.0	2	1	1	1
3827	130	Prolog subset	[Prolog subset]	0.0	2	1	1	1
3828	130	importantly text	[importantly texts]	0.0	2	1	1	1
3829	130	TDMT on APs	[TDMT on APs]	0.0	3	1	1	1
3830	130	more question	[more questions]	0.0	2	1	1	1
3831	130	statistical machine translation method	[statistical machine translation method]	0.0	4	1	1	1
3832	130	use processor	[using processors]	0.0	2	1	1	1
3833	130	way automatically identify argument structure	[way automatically identifying argument structures]	0.0	5	1	1	1
3834	130	reference word	[reference word]	0.0	2	1	1	1
3835	130	.e	[.e]	0.0	1	1	1	1
3836	130	access to a news archive	[access to a news archive]	0.0	5	1	1	1
3837	130	domain independent model	[domain independent models]	0.0	3	1	1	1
3838	130	12,000 utterance from site	[12,000 utterances from sites]	0.0	4	1	1	1
3839	130	use utterance from the target speaker	[Using utterances from the target speaker]	0.0	6	1	1	1
3840	130	neat relation to grammar	[neat relation to grammar]	0.0	4	1	1	1
3841	130	sample output	[sample output]	0.0	2	1	1	1
3842	130	novel method add linguistic annotation	[novel method adding linguistic annotation]	0.0	5	1	1	1
3843	130	log-linear interpolation	[log-linear interpolation]	0.0	2	1	1	1
3844	130	Syntactic analysis of the coordinate sentence	[Syntactic analysis of the coordinate sentences]	0.0	6	1	1	1
3845	130	robust parsing of Korean a verb	[Robust parsing of Korean a verb]	0.0	6	1	1	1
3846	130	model of language	[model of language]	0.0	3	1	1	1
3847	130	same node	[same nodes]	0.0	2	1	1	1
3848	130	new measure with judgment	[new measure with judgment]	0.0	4	1	1	1
3849	130	technique for syntactic analysis	[techniques for syntactic analysis]	0.0	4	1	1	1
3850	130	datum structure for machine translation	[data structure for machine translation]	0.0	5	1	1	1
3851	130	flexible planning-based architecture	[flexible planning-based architecture]	0.0	3	1	1	1
3852	130	perform cluster number estimation	[performing cluster number estimation]	0.0	4	1	1	1
3853	130	interpretation of natural language	[interpretation of natural language]	0.0	4	1	1	1
3854	130	algorithm acquire new stem	[algorithm acquiring new stems]	0.0	4	1	1	1
3855	130	shortcoming	[shortcomings]	0.0	1	1	1	1
3856	130	standard term	[standard term]	0.0	2	1	1	1
3857	130	parameter smoothing	[parameter smoothing]	0.0	2	1	1	1
3858	130	little corpus of word	[little corpus of words]	0.0	4	1	1	1
3859	130	human-computer interaction	[human-computer interaction]	0.0	2	1	1	1
3860	130	method build polarity-tagged corpus	[method building polarity-tagged corpus]	0.0	4	1	1	1
3861	130	use a fake symbol	[using a fake symbol]	0.0	4	1	1	1
3862	130	corpus of 2 srh	[corpus of 2 SRHs]	0.0	4	1	1	1
3863	130	Hierarchical Graph	[Hierarchical Graph]	0.0	2	1	1	1
3864	130	factor of speed-up	[factor of speed-up]	0.0	3	1	1	1
3865	130	provide useful insight	[providing useful insights]	0.0	3	1	1	1
3866	130	various kind of misunderstanding	[various kinds of misunderstandings]	0.0	4	1	1	1
3867	130	exist approach	[existing approaches]	0.0	2	1	1	1
3868	130	most man	[most men]	0.0	2	1	1	1
3869	130	word ellipsis	[words ellipsis]	0.0	2	1	1	1
3870	130	apply standard mt evaluation method	[applying standard MT evaluation methods]	0.0	5	1	1	1
3871	130	good result	[better results]	0.0	2	1	1	1
3872	130	incorporate the control structure	[incorporating the control structure]	0.0	4	1	1	1
3873	130	fake non-terminal symbol	[fake non-terminal symbol]	0.0	3	1	1	1
3874	130	nlp task	[NLP tasks]	0.0	2	1	1	1
3875	130	language in setting	[language in settings]	0.0	3	1	1	1
3876	130	use of semantics	[use of semantics]	0.0	3	1	1	1
3877	130	ellipse	[ellipses]	0.0	1	1	1	1
3878	130	negligible runtime	[negligible runtime]	0.0	2	1	1	1
3879	130	cover the domain of HK law	[covering the domain of HK laws]	0.0	6	1	1	1
3880	130	treatment for sentence analysis	[treatment for sentence analyses]	0.0	4	1	1	1
3881	130	strategy utilize mechanism	[strategies utilizing mechanisms]	0.0	3	1	1	1
3882	130	learning methodology applicable in domain	[learning methodology applicable in domains]	0.0	5	1	1	1
3883	130	5K	[5K]	0.0	1	1	1	1
3884	130	means learning of phrase translation	[means learning of phrase translations]	0.0	5	1	1	1
3885	130	problem of word sense WSD	[problem of word sense WSD]	0.0	5	1	1	1
3886	130	avoidance of copying	[avoidance of copying]	0.0	3	1	1	1
3887	130	information from a inflow of datum	[information from a inflow of data]	0.0	6	1	1	1
3888	130	grammaticality equivalence	[grammaticality equivalence]	0.0	2	1	1	1
3889	130	speed	[speed]	0.0	1	2	2	1
3890	130	alignment technique	[alignment techniques]	0.0	2	1	1	1
3891	130	resolve this ambiguity	[resolving this ambiguity]	0.0	3	1	1	1
3892	130	internal variation	[internal variation]	0.0	2	1	1	1
3893	130	representative SMT model	[representative SMT model]	0.0	3	1	1	1
3894	130	channel framework	[channel framework]	0.0	2	1	1	1
3895	130	q/a application	[Q/A applications]	0.0	2	1	1	1
3896	130	lack of datum	[lack of data]	0.0	3	1	1	1
3897	130	gather a number from the web	[gathering a number from the Web]	0.0	6	1	1	1
3898	130	algorithm for the approach	[algorithm for the approach]	0.0	4	1	1	1
3899	130	expressiveness of the two formalism	[expressiveness of the two formalisms]	0.0	5	1	1	1
3900	130	two strategy	[two strategies]	0.0	2	1	1	1
3901	130	efficient parsing	[efficient parsing]	0.0	2	1	1	1
3902	130	loss	[loss]	0.0	1	1	1	1
3903	130	language input	[language input]	0.0	2	2	2	1
3904	130	example from a fragment	[example from an fragment]	0.0	4	1	1	1
3905	130	consist of 126 sentence	[consisting of 126 sentences]	0.0	4	1	1	1
3906	130	State Model FSM	[State Model FSM]	0.0	3	1	1	1
3907	130	independent event-based summarization approach	[independent event-based summarization approaches]	0.0	4	1	1	1
3908	130	experience in the muc-3 evaluation	[experience in the MUC-3 evaluation]	0.0	5	1	1	1
3909	130	detection of that activity	[detection of those activities]	0.0	4	1	1	1
3910	130	list of possible sentence plan	[list of possible sentence plans]	0.0	5	1	1	1
3911	130	Representative sample from a language definition	[Representative samples from an language definition]	0.0	6	1	1	1
3912	130	considerable importance	[considerable importance]	0.0	2	1	1	1
3913	130	ne .g	[NE .g]	0.0	2	1	1	1
3914	130	relationship social rank	[relationship social ranks]	0.0	3	1	1	1
3915	130	translation into a RCG	[translation into an RCG]	0.0	4	1	1	1
3916	130	other syntax formalism	[other syntax formalisms]	0.0	3	1	1	1
3917	130	formal computation	[formal computation]	0.0	2	1	1	1
3918	130	underspecified semantic representation USR	[underspecified semantic representation USR]	0.0	4	1	1	1
3919	130	foundation	[foundation]	0.0	1	1	1	1
3920	130	probabilistic context-free grammar	[probabilistic context-free grammars]	0.0	3	1	1	1
3921	130	many lexical-syntactic form	[many lexical-syntactic forms]	0.0	3	1	1	1
3922	130	simple technique	[simple technique]	0.0	2	1	1	1
3923	130	use of parser evaluation method	[use of parser evaluation methods]	0.0	5	1	1	1
3924	130	contribution continuous speech recognition	[contributions continuous speech recognition]	0.0	4	1	1	1
3925	130	lexical variation in a set	[lexical variation in a set]	0.0	5	1	1	1
3926	130	96% paraphrase mean equivalence	[96% paraphrases meaning equivalence]	0.0	4	1	1	1
3927	130	1 a set	[1 an set]	0.0	3	1	1	1
3928	130	potential increase of 35%	[potential increase of 35%]	0.0	4	1	1	1
3929	130	extend the speech/language integration architecture	[extending the speech/language integration architecture]	0.0	5	1	1	1
3930	130	feature from the context	[features from the contexts]	0.0	4	1	1	1
3931	130	remote meteorological event	[remote meteorological events]	0.0	3	1	1	1
3932	130	force a interpretation	[forcing a interpretation]	0.0	3	1	1	1
3933	130	two problem attribute such as size	[two problems attributes such as size]	0.0	6	1	1	1
3934	130	numeric-valued attribute	[numeric-valued attributes]	0.0	2	1	1	1
3935	130	existence	[existence]	0.0	1	3	3	1
3936	130	parsing a verb final language	[parsing a verb final language]	0.0	5	1	1	1
3937	130	iii system development to new domain	[iii system development to new domains]	0.0	6	1	1	1
3938	130	pharmaceutical news archive	[pharmaceutical news archive]	0.0	3	1	1	1
3939	130	broad use include conditioning on word	[broad use including conditioning on words]	0.0	6	1	1	1
3940	130	ground in interaction	[ground in interaction]	0.0	3	1	1	1
3941	130	DCG	[DCG]	0.0	1	1	1	1
3942	130	model of the process	[model of the process]	0.0	4	1	1	1
3943	130	misunderstanding between speaker	[misunderstandings between speaker]	0.0	3	1	1	1
3944	130	suffix datum structure	[suffix data structure]	0.0	3	1	1	1
3945	130	document Understanding System	[Document Understanding System]	0.0	3	1	1	1
3946	130	insertion	[insertion]	0.0	1	1	1	1
3947	130	use in a multi-site evaluation	[use in a multi-site evaluation]	0.0	5	1	1	1
3948	130	processor aps	[processors APs]	0.0	2	1	1	1
3949	130	configuration	[configuration]	0.0	1	1	1	1
3950	130	exploratory corpus analysis accessible	[exploratory corpus analysis accessible]	0.0	4	1	1	1
3951	130	language understanding with speech recognition	[language understanding with speech recognition]	0.0	5	1	1	1
3952	130	relation between segment	[relations between segments]	0.0	3	1	1	1
3953	130	use a speech	[using a speech]	0.0	3	1	1	1
3954	130	also a generalization corresponding to arc	[also a generalization corresponding to arcs]	0.0	6	1	1	1
3955	130	relatively simple means heuristic learning	[relatively simple means heuristic learning]	0.0	5	1	1	1
3956	130	issue recently	[issue recently]	0.0	2	1	1	1
3957	130	effective use	[effective use]	0.0	2	1	1	1
3958	130	chemical warfare	[chemical warfare]	0.0	2	1	1	1
3959	130	theory of human reasoning	[theory of human reasoning]	0.0	4	1	1	1
3960	130	first empirical test	[first empirical test]	0.0	3	1	1	1
3961	130	DFA	[DFA]	0.0	1	1	1	1
3962	130	corpus a available resource	[corpora a available resource]	0.0	4	1	1	1
3963	130	robust parsing language with case marker	[Robust parsing language with case markers]	0.0	6	1	1	1
3964	130	coordinate structure	[coordinate structure]	0.0	2	1	1	1
3965	130	apply mutual information	[applying mutual information]	0.0	3	1	1	1
3966	130	literal meaning	[literal meaning]	0.0	2	1	1	1
3967	130	relatively free word	[relatively free word]	0.0	3	1	1	1
3968	130	multiple English translation	[multiple English translations]	0.0	3	1	1	1
3969	130	.5% agreement	[.5% agreement]	0.0	2	1	1	1
3970	130	claim to the contrary	[claims to the contrary]	0.0	4	1	1	1
3971	130	chart-based phrase structure parsing	[chart-based phrase structure parsing]	0.0	4	1	1	1
3972	130	analysis cost	[analysis cost]	0.0	2	1	1	1
3973	130	community	[communities]	0.0	1	1	1	1
3974	130	version	[version]	0.0	1	1	1	1
3975	130	stop	[stop]	0.0	1	1	1	1
3976	130	dominant part	[dominant part]	0.0	2	1	1	1
3977	130	mt evaluation method	[MT evaluation methods]	0.0	3	1	1	1
3978	130	meaning of input	[meaning of input]	0.0	3	1	1	1
3979	130	score speech recognition hypothese in term	[scoring speech recognition hypotheses in terms]	0.0	6	1	1	1
3980	130	interactive corpus exploration tool	[interactive corpus exploration tool]	0.0	4	1	1	1
3981	130	concept of the underlying theory	[concepts of the underlying theory]	0.0	5	1	1	1
3982	130	Phrasal Lexicon [Zernik88]	[Phrasal Lexicon [Zernik88]]	0.0	3	1	1	1
3983	130	most man like young woman	[most men like young women]	0.0	5	1	1	1
3984	130	feature selection method	[feature selection methods]	0.0	3	1	1	1
3985	130	problem include unlimited vocabulary	[problems including unlimited vocabulary]	0.0	4	1	1	1
3986	130	answer in corpus	[answers in corpora]	0.0	3	1	1	1
3987	130	BU	[BU]	0.0	1	1	1	1
3988	130	cluster verb	[clustering verbs]	0.0	2	1	1	1
3989	130	extensional equivalence	[extensional equivalence]	0.0	2	1	1	1
3990	130	certain kind	[certain kinds]	0.0	2	1	1	1
3991	130	syntax-based constraint	[syntax-based constraint]	0.0	2	1	1	1
3992	130	form a utterance	[forming an utterance]	0.0	3	1	1	1
3993	130	Basque	[Basque]	0.0	1	1	1	1
3994	130	query optimisation	[query optimisation]	0.0	2	1	1	1
3995	130	condition for this test suite	[condition for this test suite]	0.0	5	1	1	1
3996	130	computational system speech recognition	[computational system speech recognition]	0.0	4	1	1	1
3997	130	reduction of analysis cost	[reduction of analysis cost]	0.0	4	1	1	1
3998	130	contribution to vocabulary speech recognition	[contributions to vocabulary speech recognition]	0.0	5	1	1	1
3999	130	standard wsd evaluation methodology	[standard WSD evaluation methodology]	0.0	4	1	1	1
4000	130	assessment	[assessment]	0.0	1	1	1	1
4001	130	plausible reasoning	[plausible reasoning]	0.0	2	1	1	1
4002	130	standard output from information extraction	[standard outputs from information extraction]	0.0	5	1	1	1
4003	130	deal with complexity	[dealing with complexity]	0.0	3	1	1	1
4004	130	meaning ambiguity	[meanings ambiguity]	0.0	2	1	1	1
4005	130	use complicated knowledgebase	[using complicated knowledgebases]	0.0	3	1	1	1
4006	130	means learning	[means learning]	0.0	2	1	1	1
4007	130	Marker	[Marker]	0.0	1	1	1	1
4008	130	mixture of this technique	[mixture of these techniques]	0.0	4	1	1	1
4009	130	syntax-based machine translation system	[syntax-based machine translation system]	0.0	4	1	1	1
4010	130	biase parsing toward meaning	[biasing parsing toward meanings]	0.0	4	1	1	1
4011	130	linguistic parser/generator	[linguistic parser/generator]	0.0	2	1	1	1
4012	130	i robust parsing language	[i Robust parsing language]	0.0	4	1	1	1
4013	130	i parsing of Korean a verb	[i parsing of Korean a verb]	0.0	6	1	1	1
4014	130	relation to categorial grammar	[relation to categorial grammar]	0.0	4	1	1	1
4015	130	useful feature	[useful features]	0.0	2	1	1	1
4016	130	computer text generation system	[computer text generation system]	0.0	4	1	1	1
4017	130	Hierarchical Phrasal Lexicon DHPL [Zernik88]	[Hierarchical Phrasal Lexicon DHPL [Zernik88]]	0.0	5	1	1	1
4018	130	Research method course	[Research Methods course]	0.0	3	1	1	1
4019	130	achieve robustness	[achieving robustness]	0.0	2	1	1	1
4020	130	five site	[five sites]	0.0	2	1	1	1
4021	130	spot in language processing	[spots in language processing]	0.0	4	1	1	1
4022	130	parallel sentence in comparable corpus	[parallel sentences in comparable corpora]	0.0	5	1	1	1
4023	130	disambiguation system one	[disambiguation systems one]	0.0	3	1	1	1
4024	130	unknown word feature	[unknown word features]	0.0	3	1	1	1
4025	130	three method construct word vector	[three methods constructing word vectors]	0.0	5	1	1	1
4026	130	still satisfy communication need	[still satisfying communication needs]	0.0	4	1	1	1
4027	130	equivalence classification	[equivalence classification]	0.0	2	1	1	1
4028	130	language model LMs	[language models LMs]	0.0	3	1	1	1
4029	130	constraint for word alignment	[constraint for word alignment]	0.0	4	1	1	1
4030	130	SENSEVAL-2 lexical sample task	[SENSEVAL-2 lexical sample task]	0.0	4	1	1	1
4031	130	adjacency	[adjacency]	0.0	1	1	1	1
4032	130	subset .0%	[subset .0%]	0.0	2	1	1	1
4033	130	possible variation	[possible variations]	0.0	2	1	1	1
4034	130	numerical constraint	[numerical constraints]	0.0	2	1	1	1
4035	130	old resource management rm corpus	[old Resource Management RM corpus]	0.0	5	1	1	1
4036	130	word unsegmented corpus	[word unsegmented corpus]	0.0	3	1	1	1
4037	130	most important task	[most important task]	0.0	3	1	1	1
4038	130	recent debate	[recent debate]	0.0	2	1	1	1
4039	130	link of entry	[linking of entries]	0.0	3	1	1	1
4040	130	parser for a language	[parser for a language]	0.0	4	1	1	1
4041	130	method for speech understanding	[method for speech understanding]	0.0	4	1	1	1
4042	130	identify hub	[identifying hubs]	0.0	2	1	1	1
4043	130	research on new problem	[research on new problems]	0.0	4	1	1	1
4044	130	ne he/she/man/woman	[NE he/she/man/woman]	0.0	2	1	1	1
4045	130	material the Hansards	[material the Hansards]	0.0	3	1	1	1
4046	130	system combination for wsd	[system combination for WSD]	0.0	4	1	1	1
4047	130	document descriptor	[document descriptors]	0.0	2	1	1	1
4048	130	different lexical-syntactic form	[different lexical-syntactic forms]	0.0	3	1	1	1
4049	130	'synchronous' way	['synchronous' way]	0.0	2	1	1	1
4050	130	average precision	[average precision]	0.0	2	1	1	1
4051	130	new view for reference	[new view for reference]	0.0	4	1	1	1
4052	130	three word from word-level alignment model	[three words from word-level alignment models]	0.0	6	1	1	1
4053	130	interpretation of ellipse	[interpretation of ellipses]	0.0	3	1	1	1
4054	130	event-based mutli-document summarization approach	[event-based mutli-document summarization approaches]	0.0	4	1	1	1
4055	130	non-parallel newspaper corpus	[non-parallel newspaper corpora]	0.0	3	1	1	1
4056	130	syntax-based machine translation mt	[Syntax-based machine translation MT]	0.0	4	2	2	1
4057	130	datum relevant	[data relevant]	0.0	2	1	1	1
4058	130	c code	[C code]	0.0	2	1	1	1
4059	130	combine language model LMs	[combining language models LMs]	0.0	4	1	1	1
4060	130	chart parsing instantiation of clause	[chart Parsing instantiation of clauses]	0.0	5	1	1	1
4061	130	saturation	[saturation]	0.0	1	1	1	1
4062	130	exploitation	[exploitation]	0.0	1	1	1	1
4063	130	robust parsing method	[robust parsing method]	0.0	3	1	1	1
4064	130	visualize	[visualizing]	0.0	1	1	1	1
4065	130	superiority over method	[superiority over methods]	0.0	3	1	1	1
4066	130	integrate language understanding	[integrating language understanding]	0.0	3	1	1	1
4067	130	in-domain corpus	[in-domain corpus]	0.0	2	1	1	1
4068	130	quality translation	[quality translation]	0.0	2	1	1	1
4069	130	dialog model	[dialog model]	0.0	2	1	1	1
4070	130	corresponding with the good confidence	[corresponding with the best confidence]	0.0	5	1	1	1
4071	130	method for analysis	[method for analysis]	0.0	3	1	1	1
4072	130	particular scenario	[particular scenario]	0.0	2	1	1	1
4073	130	word-semantic information	[word-semantic information]	0.0	2	1	1	1
4074	130	tree-to-tree translation	[tree-to-tree translation]	0.0	2	1	1	1
4075	130	center on the construction	[centering on the construction]	0.0	4	1	1	1
4076	130	performance of model	[performance of models]	0.0	3	1	1	1
4077	130	abstracting	[abstracting]	0.0	1	1	1	1
4078	130	move structure	[move structures]	0.0	2	1	1	1
4079	130	representation of document	[representation of documents]	0.0	3	2	2	1
4080	130	articulation	[articulation]	0.0	1	1	1	1
4081	130	appealing alternative—in term	[appealing alternative—in terms]	0.0	3	1	1	1
4082	130	use of substitution	[use of substitutions]	0.0	3	1	1	1
4083	130	monitoring of lack	[monitoring of lack]	0.0	3	1	1	1
4084	130	use feature of the tree	[using features of the tree]	0.0	5	1	1	1
4085	130	Understanding	[Understanding]	0.0	1	1	1	1
4086	130	efficient principle-based parser for English	[efficient principle-based parser for English]	0.0	5	1	1	1
4087	130	test of a speculative claim	[test of an speculative claim]	0.0	5	1	1	1
4088	130	abstraction	[abstraction]	0.0	1	1	1	1
4089	130	subject of debate	[subject of debate]	0.0	3	1	1	1
4090	130	method produce phrase	[method producing phrases]	0.0	3	1	1	1
4091	130	own hierarchy	[own hierarchy]	0.0	2	1	1	1
4092	130	high quality translation via word sense	[High quality translation via word sense]	0.0	6	1	1	1
4093	130	i robust parsing of Korean	[i Robust parsing of Korean]	0.0	5	1	1	1
4094	130	numerous aspect	[numerous aspects]	0.0	2	1	1	1
4095	130	saturation of the structure	[saturation of the structure]	0.0	4	1	1	1
4096	130	bear on commonsense reasoning	[bearing on commonsense reasoning]	0.0	4	1	1	1
4097	130	even in a way	[even in a way]	0.0	4	1	1	1
4098	130	Role System	[Role System]	0.0	2	1	1	1
4099	130	past year	[past years]	0.0	2	1	1	1
4100	130	answer agent search for answer	[answering agents searching for answers]	0.0	5	1	1	1
4101	130	identify predicate argument structure	[identifying predicate argument structures]	0.0	4	1	1	1
4102	130	effective CFG filter	[effective CFG filter]	0.0	3	1	1	1
4103	130	contain 28 word token	[containing 28 word tokens]	0.0	4	1	1	1
4104	130	Acyclic Graph HDAG	[Acyclic Graph HDAG]	0.0	3	1	1	1
4105	130	large unsegmented corpus	[large unsegmented corpus]	0.0	3	1	1	1
4106	130	monitoring	[monitoring]	0.0	1	1	1	1
4107	130	therefore a choice	[therefore a choice]	0.0	3	1	1	1
4108	130	eye	[eye]	0.0	1	1	1	1
4109	130	progress from generation of text	[progressing from generation of text]	0.0	5	1	1	1
4110	130	use processor for language translation	[using processors for language translation]	0.0	5	1	1	1
4111	130	machine translation evaluation	[machine translation evaluation]	0.0	3	1	1	1
4112	130	powerful computer with graphic display	[powerful computers with graphics displays]	0.0	5	1	1	1
4113	130	long span	[longest span]	0.0	2	1	1	1
4114	130	hand-crafting generation system	[hand-crafting generation systems]	0.0	3	1	1	1
4115	130	SA	[SA]	0.0	1	1	1	1
4116	130	two output system entity annotation	[two outputs systems entity annotations]	0.0	5	1	1	1
4117	130	SL	[SL]	0.0	1	1	1	1
4118	130	approach to rank problem	[approach to ranking problems]	0.0	4	1	1	1
4119	130	backbone	[backbone]	0.0	1	1	1	1
4120	130	model’	[model’s]	0.0	1	1	1	1
4121	130	f-measure a relative decrease	[F-measure a relative decrease]	0.0	4	1	1	1
4122	130	cohesion-based approach	[cohesion-based approach]	0.0	2	1	1	1
4123	130	interpretation for value	[interpretation for values]	0.0	3	1	1	1
4124	130	large Arabic corpus	[large Arabic corpus]	0.0	3	1	1	1
4125	130	two hardness result	[Two hardness results]	0.0	3	1	1	1
4126	130	use in mt	[use in MT]	0.0	3	1	1	1
4127	130	new technology of computer	[new technology of computers]	0.0	4	1	1	1
4128	130	model of wh-question	[models of WH-questions]	0.0	3	1	1	1
4129	130	member of the set	[member of the set]	0.0	4	1	1	1
4130	130	robust parsing of Korean	[Robust parsing of Korean]	0.0	4	1	1	1
4131	130	special problem for reader	[special problems for readers]	0.0	4	1	1	1
4132	130	.8% improvement	[.8% improvement]	0.0	2	1	1	1
4133	130	index such as the dominance distribution	[indices such as the dominance distribution]	0.0	6	1	1	1
4134	130	text in the domain	[texts in the domain]	0.0	4	1	1	1
4135	130	performance of the word-sense	[performance of the word-sense]	0.0	4	1	1	1
4136	130	language model of a corpus	[language model of an corpus]	0.0	5	1	1	1
4137	130	sense per collocation observation	[sense per collocation observation]	0.0	4	2	2	1
4138	130	limited domain semantics	[limited domain semantics]	0.0	3	1	1	1
4139	130	pair of grammar work	[pairs of grammars working]	0.0	4	1	1	1
4140	130	particular interpretation	[particular interpretation]	0.0	2	1	1	1
4141	130	result on addressee identification face-to-face meeting	[results on addressee identification face-to-face meetings]	0.0	6	1	1	1
4142	130	increase the dialogue duration	[increasing the dialogue duration]	0.0	4	1	1	1
4143	130	coordinate structure analysis model	[coordinate structure analysis model]	0.0	4	1	1	1
4144	130	approach to parsing	[approach to parsing]	0.0	3	1	1	1
4145	130	lexical variation	[lexical variation]	0.0	2	1	1	1
4146	130	pronoun in sentence	[pronoun in sentences]	0.0	3	1	1	1
4147	130	system entity annotation	[systems entity annotations]	0.0	3	1	1	1
4148	130	language technology	[language technology]	0.0	2	1	1	1
4149	130	meta-inference type	[meta-inference types]	0.0	2	1	1	1
4150	130	[Zernik88]	[[Zernik88]]	0.0	1	1	1	1
4151	130	inflow	[inflow]	0.0	1	1	1	1
4152	130	system with mobile agent	[system with mobile agents]	0.0	4	1	1	1
4153	130	statistical pos tagger	[statistical POS tagger]	0.0	3	1	1	1
4154	130	gold standard	[gold standard]	0.0	2	1	1	1
4155	130	Phonology structure Grammar	[Phonology Structure Grammar]	0.0	3	1	1	1
4156	130	achieve performance to mixture	[achieving performance to mixtures]	0.0	4	1	1	1
4157	130	speech recognition hypothese SRH	[speech recognition hypotheses SRH]	0.0	4	1	1	1
4158	130	WI	[WI]	0.0	1	1	1	1
4159	130	fine-grained modeling	[fine-grained modeling]	0.0	2	1	1	1
4160	130	version of grammar	[version of grammars]	0.0	3	1	1	1
4161	130	condensation	[condensation]	0.0	1	1	1	1
4162	130	anaphora resolution module	[anaphora resolution module]	0.0	3	1	1	1
4163	130	employ a phonological approach	[employing a phonological approach]	0.0	4	1	1	1
4164	130	mutli-document summarization	[mutli-document summarization]	0.0	2	1	1	1
4165	130	representation the Hierarchical Phrasal Lexicon	[representation the Hierarchical Phrasal Lexicon]	0.0	5	1	1	1
4166	130	statistical Machine Translation	[Statistical Machine Translation]	0.0	3	1	1	1
4167	130	news	[news]	0.0	1	1	1	1
4168	130	speech Recognition technology	[Speech Recognition technology]	0.0	3	1	1	1
4169	130	exist probabilistic parsing model	[existing probabilistic parsing models]	0.0	4	1	1	1
4170	130	Chomsky program	[Chomsky's program]	0.0	2	1	1	1
4171	130	large study	[larger study]	0.0	2	1	1	1
4172	130	Senseval-3 lexical sample task	[Senseval-3 lexical sample task]	0.0	4	1	1	1
4173	130	learning Model	[Learning Model]	0.0	2	1	1	1
4174	130	communication between APs	[communications between APs]	0.0	3	1	1	1
4175	130	accord to character bigram	[according to character bigrams]	0.0	4	1	1	1
4176	130	practical approach	[practical approach]	0.0	2	1	1	1
4177	130	simple means	[simple means]	0.0	2	1	1	1
4178	130	effect use a corpus	[effect using an corpus]	0.0	4	1	1	1
4179	130	many formalism	[many formalisms]	0.0	2	1	1	1
4180	130	example most similar	[examples most similar]	0.0	3	1	1	1
4181	130	plan	[plan]	0.0	1	1	1	1
4182	130	agenda-based scheduling parser	[agenda-based scheduling parser]	0.0	3	1	1	1
4183	130	weather datum	[weather data]	0.0	2	1	1	1
4184	130	answer candidate	[answer candidates]	0.0	2	1	1	1
4185	130	language learning process	[language learning process]	0.0	3	1	1	1
4186	130	response adaptive	[responses adaptive]	0.0	2	1	1	1
4187	130	portable understanding system	[portable understanding system]	0.0	3	1	1	1
4188	130	corpus datum	[corpus data]	0.0	2	1	1	1
4189	130	commonly available resource	[commonly available resource]	0.0	3	1	1	1
4190	130	superset	[superset]	0.0	1	1	1	1
4191	130	.24% accuracy	[.24% accuracy]	0.0	2	1	1	1
4192	130	gain	[gain]	0.0	1	1	1	1
4193	130	technology of computer	[technology of computers]	0.0	3	1	1	1
4194	130	adopt a paraphrasing technique	[adopting a paraphrasing technique]	0.0	4	1	1	1
4195	130	context-dependent modelling	[context-dependent modelling]	0.0	2	1	1	1
4196	130	FlexP	[FlexP]	0.0	1	1	1	1
4197	130	candidate the structure	[candidates the structure]	0.0	3	1	1	1
4198	130	introduction to machine translation	[introduction to machine translation]	0.0	4	1	1	1
4199	130	tree Grammar	[Tree Grammars]	0.0	2	1	1	1
4200	130	include inherent limitation of mt architecture	[including inherent limitations of MT architectures]	0.0	6	1	1	1
4201	130	generative model of natural language	[generative model of natural language]	0.0	5	1	1	1
4202	130	annotation performance	[annotation performance]	0.0	2	1	1	1
4203	130	different strategy utilize mechanism	[different strategies utilizing mechanisms]	0.0	4	1	1	1
4204	130	distinction	[distinction]	0.0	1	1	1	1
4205	130	new view	[new view]	0.0	2	1	1	1
4206	130	text-understanding system	[text-understanding systems]	0.0	2	1	1	1
4207	130	most young man	[most young men]	0.0	3	1	1	1
4208	130	robust parsing a verb language	[Robust parsing a verb language]	0.0	5	1	1	1
4209	130	meeting situation	[meeting situation]	0.0	2	1	1	1
4210	130	two style perform inference in network	[Two styles performing inference in networks]	0.0	6	1	1	1
4211	130	common attribute sequence	[common attribute sequences]	0.0	3	1	1	1
4212	130	interpretation from a chunk	[interpretation from a chunk]	0.0	4	1	1	1
4213	130	improvement over word alignment technique	[improvement over word alignment techniques]	0.0	5	1	1	1
4214	130	f-measure a 13% decrease	[F-measure a 13% decrease]	0.0	4	1	1	1
4215	130	quantifier most in the proposition	[quantifier most in the proposition]	0.0	5	1	1	1
4216	130	two standard output from information extraction	[two standard outputs from information extraction]	0.0	6	1	1	1
4217	130	hold multiple candidate	[holding multiple candidates]	0.0	3	2	2	1
4218	130	efficient Prolog	[efficient Prolog]	0.0	2	1	1	1
4219	130	knowledge base system	[knowledge base system]	0.0	3	1	1	1
4220	130	use include conditioning on multiple word	[use including conditioning on multiple words]	0.0	6	1	1	1
4221	130	use a speech from many speaker	[using a speech from many speakers]	0.0	6	1	1	1
4222	130	research on problem include vocabulary	[research on problems including vocabulary]	0.0	5	1	1	1
4223	130	dimension of user model	[dimensions of user models]	0.0	4	1	1	1
4224	130	contribute	[contributing]	0.0	1	1	1	1
4225	130	possible candidate	[possible candidates]	0.0	2	1	1	1
4226	130	two word-sense disambiguation system one	[two word-sense disambiguation systems one]	0.0	5	1	1	1
4227	130	explication	[explication]	0.0	1	1	1	1
4228	130	special type of automaton	[special type of automaton]	0.0	4	1	1	1
4229	130	apply model to datum	[applying models to data]	0.0	4	1	1	1
4230	130	valuable method	[valuable methods]	0.0	2	1	1	1
4231	130	frequently predictor of document descriptor	[frequently predictors of document descriptors]	0.0	5	1	1	1
4232	130	apply mt evaluation method	[applying MT evaluation methods]	0.0	4	1	1	1
4233	130	State Model	[State Model]	0.0	2	1	1	1
4234	130	Stabler'	[Stabler's]	0.0	1	1	1	1
4235	130	category on the terminal	[categories on the terminal]	0.0	4	1	1	1
4236	130	cluster number estimation	[cluster number estimation]	0.0	3	1	1	1
4237	130	simple means learning	[simple means learning]	0.0	3	1	1	1
4238	130	repository of information for use	[repository of information for use]	0.0	5	1	1	1
4239	130	fulfillment	[fulfillment]	0.0	1	1	1	1
4240	130	subjective judgment	[subjective judgments]	0.0	2	1	1	1
4241	130	amount of work	[amount of work]	0.0	3	1	1	1
4242	130	use utterance for adaptation	[Using utterances for adaptation]	0.0	4	1	1	1
4243	130	predict the right translation	[predicting the right translation]	0.0	4	1	1	1
4244	130	efficient parsing a verb	[efficient parsing a verb]	0.0	4	1	1	1
4245	130	special problem	[special problems]	0.0	2	1	1	1
4246	130	requirement of language translation	[requirements of language translation]	0.0	4	1	1	1
4247	130	monolingual material	[monolingual material]	0.0	2	1	1	1
4248	130	efficient Prolog query optimisation	[efficient Prolog query optimisation]	0.0	4	1	1	1
4249	130	employ a traditional generative approach	[employing a traditional generative approach]	0.0	5	1	1	1
4250	130	tag sequence	[tag sequences]	0.0	2	1	1	1
4251	130	37% reduction	[37% reduction]	0.0	2	1	1	1
4252	130	estimation of markov model cdhmm	[estimation of Markov models CDHMM]	0.0	5	1	1	1
4253	130	ground in human-computer interaction	[ground in human-computer interaction]	0.0	4	1	1	1
4254	130	idea of this method	[idea of this method]	0.0	4	1	1	1
4255	130	discover parallel sentence in corpus	[discovering parallel sentences in corpora]	0.0	5	1	1	1
4256	130	parsing of Korean a verb language	[parsing of Korean a verb language]	0.0	6	1	1	1
4257	130	high quality translation	[High quality translation]	0.0	3	1	1	1
4258	130	small 10k sentence parallel corpus	[small 10K sentences parallel corpus]	0.0	5	1	1	1
4259	130	muc-3 evaluation	[MUC-3 evaluation]	0.0	2	1	1	1
4260	130	use asr output	[using ASR output]	0.0	3	1	1	1
4261	130	redundancy elimination	[redundancy elimination]	0.0	2	1	1	1
4262	130	dimension of user model skill level	[dimensions of user models skill level]	0.0	6	1	1	1
4263	130	feature a language model score	[features a language model score]	0.0	5	1	1	1
4264	130	area of Human-Machine Communication	[areas of Human-Machine Communication]	0.0	4	1	1	1
4265	130	difficult problem	[difficult problems]	0.0	2	1	1	1
4266	130	database query	[database query]	0.0	2	1	1	1
4267	130	also a drastic generalization roughly corresponding	[also a drastic generalization roughly corresponding]	0.0	6	1	1	1
4268	130	Graph Kernel	[Graph Kernel]	0.0	2	1	1	1
4269	130	noun group	[noun group]	0.0	2	1	1	1
4270	130	large knowledgebase	[large knowledgebases]	0.0	2	1	1	1
4271	130	actual description	[actual descriptions]	0.0	2	1	1	1
4272	130	font	[fonts]	0.0	1	1	1	1
4273	130	transducer	[transducer]	0.0	1	1	1	1
4274	130	apply a set of extraction	[applying a set of extraction]	0.0	5	1	1	1
4275	130	accuracy in language identification	[accuracy in language identification]	0.0	4	1	1	1
4276	130	user of tool	[users of tool]	0.0	3	1	1	1
4277	130	vital requirement	[vital requirements]	0.0	2	1	1	1
4278	130	representational format	[representational format]	0.0	2	1	1	1
4279	130	application of Unification Grammar	[application of Unification Grammar]	0.0	4	1	1	1
4280	130	evaluate the summarization quality	[evaluating the summarization quality]	0.0	4	1	1	1
4281	130	similarity between word	[similarity between words]	0.0	3	1	1	1
4282	130	piece	[piece]	0.0	1	1	1	1
4283	130	role in syntax	[role in syntax]	0.0	3	1	1	1
4284	130	.0% improvement	[.0% improvement]	0.0	2	1	1	1
4285	130	representation of a scope ambiguity	[representation of a scope ambiguity]	0.0	5	1	1	1
4286	130	question answer capability on text	[question answering capability on text]	0.0	5	1	1	1
4287	130	describe the layout	[describing the layout]	0.0	3	1	1	1
4288	130	language understand technology	[language understanding technology]	0.0	3	1	1	1
4289	130	involve extraction from text	[involving extraction from text]	0.0	4	1	1	1
4290	130	relationship email .g rank	[relationship e .g ranks]	0.0	4	1	1	1
4291	130	skill	[skill]	0.0	1	1	1	1
4292	130	response adaptive to user	[responses adaptive to users]	0.0	4	1	1	1
4293	130	use a a corpus	[using a a corpus]	0.0	4	1	1	1
4294	130	instantiation scheme	[instantiation scheme]	0.0	2	1	1	1
4295	130	new part-of-speech tagger use	[new part-of-speech tagger use]	0.0	4	1	1	1
4296	130	system development via acquisition of grammar	[system development via acquisition of grammars]	0.0	6	1	1	1
4297	130	guidance for novice user	[guidance for novice users]	0.0	4	1	1	1
4298	130	web-based academic writing	[Web-based academic writing]	0.0	3	1	1	1
4299	130	improvement for speaker adaptation SA	[improvement for speaker adaptation SA]	0.0	5	1	1	1
4300	130	further reduction	[further reduction]	0.0	2	1	1	1
4301	130	several factor include inherent limitation	[several factors including inherent limitations]	0.0	5	1	1	1
4302	130	find datum	[finding data]	0.0	2	1	1	1
4303	130	Begin/After	[Begin/After]	0.0	1	1	1	1
4304	130	promising approach to writing	[promising approach to writing]	0.0	4	1	1	1
4305	130	language parsing nlp	[language parsing NLP]	0.0	3	1	1	1
4306	130	evaluate a SMT model	[evaluating a SMT model]	0.0	4	1	1	1
4307	130	precise description	[precise description]	0.0	2	1	1	1
4308	130	certain class	[certain classes]	0.0	2	1	1	1
4309	130	5 year	[5 years]	0.0	2	1	1	1
4310	130	grammaticality correct sentence equivalence	[grammaticality correct sentences equivalence]	0.0	4	1	1	1
4311	130	construction of a corpus resource	[construction of a corpus resource]	0.0	5	1	1	1
4312	130	Chinese-English text	[Chinese-English texts]	0.0	2	1	1	1
4313	130	New experimental result on all application	[New experimental results on all applications]	0.0	6	1	1	1
4314	130	possible structure	[possible structures]	0.0	2	1	1	1
4315	130	meaning of a sentence	[meaning of a sentence]	0.0	4	1	1	1
4316	130	contrast	[contrast]	0.0	1	2	2	1
4317	130	sense of as much	[sense of as much]	0.0	4	1	1	1
4318	130	sentence co-occurrence	[sentence co-occurrences]	0.0	2	1	1	1
4319	130	28 word token	[28 word tokens]	0.0	3	1	1	1
4320	130	tagging result	[tagging result]	0.0	2	1	1	1
4321	130	include language processing	[including Language Processing]	0.0	3	1	1	1
4322	130	occurrence	[occurrences]	0.0	1	1	1	1
4323	130	method .e lsa-based method	[methods .e LSA-based methods]	0.0	4	1	1	1
4324	130	reflect the relationship social rank	[reflecting the relationship social ranks]	0.0	5	1	1	1
4325	130	scheme'	[scheme's]	0.0	1	1	1	1
4326	130	predictor of document descriptor	[predictors of document descriptors]	0.0	4	1	1	1
4327	130	text retrieval	[text retrieval]	0.0	2	1	1	1
4328	130	accord to simple character bigram	[according to simple character bigrams]	0.0	5	1	1	1
4329	130	arbiter-based combination	[arbiter-based combination]	0.0	2	1	1	1
4330	130	person's knowledge	[person's knowledge]	0.0	2	1	1	1
4331	130	discover parallel sentence	[discovering parallel sentences]	0.0	3	1	1	1
4332	130	accomplishment	[accomplishments]	0.0	1	1	1	1
4333	130	lm	[LM]	0.0	1	1	1	1
4334	130	evaluation use information retrieval	[evaluation using information retrieval]	0.0	4	1	1	1
4335	130	different algorithm	[different algorithms]	0.0	2	1	1	1
4336	130	good guidance	[good guidance]	0.0	2	1	1	1
4337	130	add annotation	[adding annotation]	0.0	2	1	1	1
4338	130	personal computer with graphic display	[personal computers with graphics displays]	0.0	5	1	1	1
4339	130	output of ocr system	[output of OCR systems]	0.0	4	1	1	1
4340	130	extend activity	[extending activity]	0.0	2	1	1	1
4341	130	plan COMPLEX	[planning COMPLEX]	0.0	2	1	1	1
4342	130	mt output	[MT output]	0.0	2	1	1	1
4343	130	amount of lexical variation superior	[amount of lexical variation superior]	0.0	5	1	1	1
4344	130	domain acoustic model with off-the-shelf classifier	[domain acoustic models with off-the-shelf classifiers]	0.0	6	1	1	1
4345	130	induce a grammar	[inducing a grammar]	0.0	3	1	1	1
4346	130	generation of expression	[Generation of Expressions]	0.0	3	1	1	1
4347	130	syntactic construction	[syntactic construction]	0.0	2	1	1	1
4348	130	partial instantiation	[partial instantiation]	0.0	2	1	1	1
4349	130	Phonology structure Grammar KPSG	[Phonology Structure Grammar KPSG]	0.0	4	1	1	1
4350	130	superiority of this method	[superiority of this method]	0.0	4	1	1	1
4351	130	such as editing term	[such as editing terms]	0.0	4	1	1	1
4352	130	method of inference	[method of inference]	0.0	3	1	1	1
4353	130	importance to Machine Translation SMT	[importance to Machine Translation SMT]	0.0	5	1	1	1
4354	130	frequent omission	[frequent omissions]	0.0	2	1	1	1
4355	130	efficient broad-coverage principle-based parser	[efficient broad-coverage principle-based parser]	0.0	4	1	1	1
4356	130	great benefit	[great benefit]	0.0	2	1	1	1
4357	130	broad use of feature include conditioning	[broad use of features including conditioning]	0.0	6	1	1	1
4358	130	context-free grammar work	[context-free grammars working]	0.0	3	1	1	1
4359	130	contribution speech recognition	[contributions speech recognition]	0.0	3	1	1	1
4360	130	use Bayesian Network classifier	[using Bayesian Network classifiers]	0.0	4	1	1	1
4361	130	many different lexical-syntactic form	[many different lexical-syntactic forms]	0.0	4	1	1	1
4362	130	automatic word sense induction	[automatic word sense induction]	0.0	4	1	1	1
4363	130	generative probabilistic character recognition model	[generative probabilistic character recognition model]	0.0	5	1	1	1
4364	130	also a generalization corresponding	[also a generalization corresponding]	0.0	4	1	1	1
4365	130	candidate the syntactic structure	[candidates the syntactic structure]	0.0	4	1	1	1
4366	130	Senseval series	[Senseval series]	0.0	2	1	1	1
4367	130	tdmt the part	[TDMT the part]	0.0	3	1	1	1
4368	130	suppress the quantifier	[suppressing the quantifier]	0.0	3	1	1	1
4369	130	2-character name without title	[2-character names without title]	0.0	4	1	1	1
4370	130	Slavonic language	[Slavonic languages]	0.0	2	1	1	1
4371	130	representation the Phrasal Lexicon [Zernik88]	[representation the Phrasal Lexicon [Zernik88]]	0.0	5	1	1	1
4372	130	adaptation	[adaptation]	0.0	1	1	1	1
4373	130	train learner	[training learners]	0.0	2	1	1	1
4374	130	model use processor aps	[model using processors APs]	0.0	4	1	1	1
4375	130	system user	[system users]	0.0	2	1	1	1
4376	130	semantically classify SCF datum	[semantically classifying SCF data]	0.0	4	1	1	1
4377	130	relative decrease in f-measure error	[relative decrease in F-measure error]	0.0	5	1	1	1
4378	130	least 96% paraphrase mean equivalence	[least 96% paraphrases meaning equivalence]	0.0	5	1	1	1
4379	130	judicious mixture	[judicious mixture]	0.0	2	1	1	1
4380	130	Korean newspaper article	[Korean newspaper articles]	0.0	3	1	1	1
4381	130	tree-based model	[tree-based models]	0.0	2	1	1	1
4382	130	correctness of answer	[correctness of answers]	0.0	3	1	1	1
4383	130	corpus analysis accessible	[corpus analysis accessible]	0.0	3	1	1	1
4384	130	presence of word	[presence of words]	0.0	3	1	1	1
4385	130	new word-sense disambiguation system one	[new word-sense disambiguation systems one]	0.0	5	1	1	1
4386	130	encode the meaning of the sentence	[encoding the meaning of the sentence]	0.0	6	1	1	1
4387	130	computational model	[computational model]	0.0	2	1	1	1
4388	130	substantial improvement over all	[substantial improvement over all]	0.0	4	1	1	1
4389	130	good translation quality	[better translation quality]	0.0	3	1	1	1
4390	130	rm	[RM]	0.0	1	1	1	1
4391	130	flexible architecture with a variety	[flexible architecture with a variety]	0.0	5	1	1	1
4392	130	abduction of relation between segment	[abduction of relations between segments]	0.0	5	1	1	1
4393	130	repository of datum object	[repository of data objects]	0.0	4	1	1	1
4394	130	parsing a verb language	[parsing a verb language]	0.0	4	1	1	1
4395	130	comparison of CFG filter technique	[comparison of CFG filtering techniques]	0.0	5	1	1	1
4396	130	lack of feedback	[lack of feedback]	0.0	3	1	1	1
4397	130	generative optical character recognition model	[generative optical character recognition model]	0.0	5	1	1	1
4398	130	Stabler' formalization	[Stabler's formalization]	0.0	2	1	1	1
4399	130	language with little inflection	[languages with little inflection]	0.0	4	1	1	1
4400	130	new disambiguation scheme	[new disambiguation scheme]	0.0	3	1	1	1
4401	130	Communication	[Communication]	0.0	1	1	1	1
4402	130	volume from the web	[volume from the Web]	0.0	4	1	1	1
4403	130	Phrasal Lexicon	[Phrasal Lexicon]	0.0	2	1	1	1
4404	130	collection of 12,000 utterance	[collection of 12,000 utterances]	0.0	4	1	1	1
4405	130	other clue	[Other clues]	0.0	2	1	1	1
4406	130	mt architecture	[MT architectures]	0.0	2	1	1	1
4407	130	analyze English sentence	[analyzing English sentences]	0.0	3	1	1	1
4408	130	one language	[one language]	0.0	2	1	1	1
4409	130	translation equivalent	[translation equivalents]	0.0	2	1	1	1
4410	130	distinct domain tutorial dialogue	[distinct domains tutorial dialogue]	0.0	4	1	1	1
4411	130	Graph Kernel for language datum	[Graph Kernel for language data]	0.0	5	1	1	1
4412	130	several experiment corresponding	[Several experiments corresponding]	0.0	3	1	1	1
4413	130	equivalence of intensional concept	[equivalence of intensional concepts]	0.0	4	1	1	1
4414	130	example from a english-spanish fragment	[example from an English-Spanish fragment]	0.0	5	1	1	1
4415	130	negative impact	[negative impact]	0.0	2	1	1	1
4416	130	social relationship email .g rank	[social relationship e .g ranks]	0.0	5	1	1	1
4417	130	grammaticality equivalence in meaning	[grammaticality equivalence in meaning]	0.0	4	1	1	1
4418	130	information from a inflow	[information from a inflow]	0.0	4	1	1	1
4419	130	search for answer	[searching for answers]	0.0	3	1	1	1
4420	130	other method	[other methods]	0.0	2	1	1	1
4421	130	achieve performance to more complex mixture	[achieving performance to more complex mixtures]	0.0	6	1	1	1
4422	130	benefit to language pair	[benefit to language pairs]	0.0	4	1	1	1
4423	130	semantics of natural language application	[semantics of natural language applications]	0.0	5	1	1	1
4424	130	four effect	[four effects]	0.0	2	1	1	1
4425	130	problem a attribute such as size	[problems a attributes such as size]	0.0	6	1	1	1
4426	130	necessary condition for the use	[necessary conditions for the use]	0.0	5	1	1	1
4427	130	organizational strategy	[organizational strategies]	0.0	2	1	1	1
4428	130	new disambiguation system	[new disambiguation systems]	0.0	3	1	1	1
4429	130	theory about this rule	[theory about these rules]	0.0	4	1	1	1
4430	130	multiple word	[multiple words]	0.0	2	1	1	1
4431	130	monolingual text	[Monolingual text]	0.0	2	1	1	1
4432	130	deictic reference	[Deictic reference]	0.0	2	1	1	1
4433	130	candidate for the understanding result	[candidates for the understanding result]	0.0	5	1	1	1
4434	130	user study	[user studies]	0.0	2	1	1	1
4435	130	anaphora resolution module of Lucy	[anaphora resolution module of Lucy]	0.0	5	1	1	1
4436	130	extract information from text	[extracting information from texts]	0.0	4	1	1	1
4437	130	reading	[readings]	0.0	1	1	1	1
4438	130	domain semantics	[domain semantics]	0.0	2	1	1	1
4439	130	potentially large list	[potentially large list]	0.0	3	1	1	1
4440	130	correlation between the parse-based evaluation	[correlation between the parse-based evaluation]	0.0	5	1	1	1
4441	130	equivalence of concept	[equivalence of concepts]	0.0	3	1	1	1
4442	130	reduce the amount of ambiguity	[reducing the amount of ambiguity]	0.0	5	1	1	1
4443	130	strategy utilize knowledge-based mechanism	[strategies utilizing knowledge-based mechanisms]	0.0	4	1	1	1
4444	130	construct a phonological system	[constructing a phonological system]	0.0	4	1	1	1
4445	130	use a lm	[using a LM]	0.0	3	1	1	1
4446	130	point of view	[point of view]	0.0	3	1	1	1
4447	130	recall of 49%	[recall of 49%]	0.0	3	1	1	1
4448	130	probabilistic dependency insertion grammar	[probabilistic dependency insertion grammar]	0.0	4	1	1	1
4449	130	.3m Chinese character	[.3M Chinese characters]	0.0	3	1	1	1
4450	130	factor behind this finding	[factors behind this finding]	0.0	4	1	1	1
4451	130	phase	[phases]	0.0	1	1	1	1
4452	130	principled technique for analysis	[principled techniques for analysis]	0.0	4	1	1	1
4453	130	University at San Diego	[University at San Diego]	0.0	4	1	1	1
4454	130	train two learner	[training two learners]	0.0	3	1	1	1
4455	130	new word-sense disambiguation system	[new word-sense disambiguation systems]	0.0	4	1	1	1
4456	130	grammaticality at least 99% sentence	[grammaticality at least 99% sentences]	0.0	5	1	1	1
4457	130	scenic description	[scenic descriptions]	0.0	2	1	1	1
4458	130	rank problem	[ranking problems]	0.0	2	1	1	1
4459	130	detailed information	[detailed information]	0.0	2	1	1	1
4460	130	recent work harvesting English-Chinese bitext	[recent work harvesting English-Chinese bitexts]	0.0	5	1	1	1
4461	130	robust interactive method	[robust interactive method]	0.0	3	1	1	1
4462	130	system development via knowledge-based acquisition	[system development via knowledge-based acquisition]	0.0	5	1	1	1
4463	130	find more datum relevant	[finding more data relevant]	0.0	4	1	1	1
4464	130	score speech recognition hypothese SRH	[scoring speech recognition hypotheses SRH]	0.0	5	1	1	1
4465	130	exploitation of domain semantics	[exploitation of domain semantics]	0.0	4	1	1	1
4466	130	kind of knowledge	[kinds of knowledge]	0.0	3	1	1	1
4467	130	knowledge about own knowledge	[knowledge about own knowledge]	0.0	4	1	1	1
4468	130	parsing algorithm	[parsing algorithms]	0.0	2	1	1	1
4469	130	basis for a useful tool	[basis for a useful tool]	0.0	5	1	1	1
4470	130	result of a qualitative user evaluation	[results of a qualitative user evaluation]	0.0	6	1	1	1
4471	130	speech/language integration	[speech/language integration]	0.0	2	1	1	1
4472	130	Roget's Thesaurus	[Roget's Thesaurus]	0.0	2	1	1	1
4473	130	explicit information from dictionary	[explicit information from dictionaries]	0.0	4	1	1	1
4474	130	result of a preliminary user evaluation	[results of a preliminary user evaluation]	0.0	6	1	1	1
4475	130	corpus 100 ,000 word	[corpus 100 ,000 words]	0.0	4	1	1	1
4476	130	application to the framework	[application to the framework]	0.0	4	1	1	1
4477	130	extract specific information from text	[extracting specific information from texts]	0.0	5	1	1	1
4478	130	maximum-entropy model for output selection	[maximum-entropy model for output selection]	0.0	5	1	1	1
4479	130	novel method build corpus	[novel method building corpus]	0.0	4	1	1	1
4480	130	expression include a word	[expressions including a word]	0.0	4	1	1	1
4481	130	Intel i860 chip	[Intel i860 chip]	0.0	3	1	1	1
4482	130	translate	[translates]	0.0	1	1	1	1
4483	130	grammar coverage	[grammar coverage]	0.0	2	1	1	1
4484	130	implication for discourse processing algorithm	[implications for discourse processing algorithms]	0.0	5	1	1	1
4485	130	extend activity usually in a manner	[extending activity usually in a manner]	0.0	6	1	1	1
4486	130	scope information of the correct structure	[scope information of the correct structure]	0.0	6	1	1	1
4487	130	sophisticated unit	[sophisticated units]	0.0	2	1	1	1
4488	130	method acquire datum	[method acquiring data]	0.0	3	1	1	1
4489	130	SMT algorithm	[SMT algorithms]	0.0	2	1	1	1
4490	130	linguistic introspection	[linguistic introspection]	0.0	2	1	1	1
4491	130	evaluation methodology	[evaluation methodology]	0.0	2	1	1	1
4492	130	use conventional word-trigram recognition require transcription	[using conventional word-trigram recognition requiring transcription]	0.0	6	1	1	1
4493	130	chart-based algorithm	[chart-based algorithms]	0.0	2	1	1	1
4494	130	category Cooccurrence Restriction	[Category Cooccurrence Restrictions]	0.0	3	1	1	1
4495	130	match accuracy	[match accuracy]	0.0	2	1	1	1
4496	130	non-literal aspect such as communication procedure	[non-literal aspects such as communication procedures]	0.0	6	1	1	1
4497	130	reformulation	[reformulation]	0.0	1	1	1	1
4498	130	research on major new problem	[research on major new problems]	0.0	5	1	1	1
4499	130	predict concept	[predicting concepts]	0.0	2	1	1	1
4500	130	identification of 2-character name	[identification of 2-character names]	0.0	4	1	1	1
4501	130	drastic generalization corresponding	[drastic generalization corresponding]	0.0	3	1	1	1
4502	130	evaluate word sense WSD model	[evaluating word sense WSD models]	0.0	5	1	1	1
4503	130	entropy word alignment algorithm	[entropy word alignment algorithm]	0.0	4	1	1	1
4504	130	human performance	[human performance]	0.0	2	1	1	1
4505	130	achieve performance to complex mixture	[achieving performance to complex mixtures]	0.0	5	1	1	1
4506	130	worst-case parsing time complexity	[worst-case parsing time complexity]	0.0	4	1	1	1
4507	130	complete distribution of knowledge	[Complete distribution of knowledge]	0.0	4	1	1	1
4508	130	build a language model	[building a language model]	0.0	4	1	1	1
4509	130	good predictor	[good predictors]	0.0	2	1	1	1
4510	130	cast doubt	[casting doubt]	0.0	2	1	1	1
4511	130	use expectation	[using expectations]	0.0	2	1	1	1
4512	130	evidence from a ,000 feature	[evidence from an ,000 features]	0.0	5	1	1	1
4513	130	97% exact match accuracy	[97% exact match accuracy]	0.0	4	1	1	1
4514	130	use in a common evaluation	[use in a common evaluation]	0.0	5	1	1	1
4515	130	explicit use of tag context	[explicit use of tag contexts]	0.0	5	1	1	1
4516	130	promising approach to web-based writing	[promising approach to Web-based writing]	0.0	5	1	1	1
4517	130	current discourse processing algorithm	[current discourse processing algorithms]	0.0	4	1	1	1
4518	130	word classification	[word classification]	0.0	2	1	1	1
4519	130	ONTOSCORE a system	[ONTOSCORE a system]	0.0	3	1	1	1
4520	130	extend activity in a rightward manner	[extending activity in a rightward manner]	0.0	6	1	1	1
4521	130	generic formalism for the combination	[generic formalism for the combination]	0.0	5	1	1	1
4522	130	Lucy	[Lucy]	0.0	1	1	1	1
4523	130	industry watch function	[industry watch function]	0.0	3	1	1	1
4524	130	grammar formalism such as system	[grammar formalisms such as systems]	0.0	5	1	1	1
4525	130	translation of language	[translation of language]	0.0	3	1	1	1
4526	130	understanding of text	[understanding of texts]	0.0	3	1	1	1
4527	130	12,000 utterance of speech	[12,000 utterances of speech]	0.0	4	1	1	1
4528	130	retrieve example	[retrieving examples]	0.0	2	1	1	1
4529	130	machine translation SMT	[machine translation SMT]	0.0	3	2	2	1
4530	130	use associative processor aps	[using associative processors APs]	0.0	4	1	1	1
4531	130	automatic evaluation measure	[automatic evaluation measures]	0.0	3	1	1	1
4532	130	fundamentally different strategy utilize knowledge-based mechanism	[fundamentally different strategies utilizing knowledge-based mechanisms]	0.0	6	1	1	1
4533	130	respectful word	[respectful words]	0.0	2	1	1	1
4534	130	use additional feature as evidence	[using additional features as evidence]	0.0	5	1	1	1
4535	130	promising approach to academic writing	[promising approach to academic writing]	0.0	5	1	1	1
4536	130	utterance of speech	[utterances of speech]	0.0	3	1	1	1
4537	130	work harvesting English-Chinese bitext	[work harvesting English-Chinese bitexts]	0.0	4	1	1	1
4538	130	class-dependent interpolation	[class-dependent interpolation]	0.0	2	1	1	1
4539	130	character recognition ocr model	[character recognition OCR model]	0.0	4	1	1	1
4540	130	system non-monotonic reasoning	[system non-monotonic reasoning]	0.0	3	1	1	1
4541	130	very simple sentence-plan-generator	[very simple sentence-plan-generator]	0.0	3	1	1	1
4542	130	logical definition of grammar	[logical definition of grammars]	0.0	4	1	1	1
4543	130	Network classifier	[Network classifiers]	0.0	2	1	1	1
4544	130	primarily on information from dictionary MRD's	[primarily on information from dictionaries MRD's]	0.0	6	1	1	1
4545	130	information bottleneck	[Information Bottleneck]	0.0	2	1	1	1
4546	130	important task build the editor	[important task building the editor]	0.0	5	1	1	1
4547	130	acquisition of information .g	[acquisition of information .g]	0.0	4	1	1	1
4548	130	problem for reader	[problems for readers]	0.0	3	1	1	1
4549	130	translation of the source text	[translations of the source text]	0.0	5	1	1	1
4550	130	top-level prediction task	[top-level prediction task]	0.0	3	1	1	1
4551	130	apply language	[applying language]	0.0	2	1	1	1
4552	130	machine learning	[machine learning]	0.0	2	1	1	1
4553	130	semantics of language application	[semantics of language applications]	0.0	4	1	1	1
4554	130	explore bitext in a large volume	[exploring bitexts in a larger volume]	0.0	6	1	1	1
4555	130	more commonly available resource	[more commonly available resource]	0.0	4	1	1	1
4556	130	contribute match	[contributing matches]	0.0	2	1	1	1
4557	130	issue in Natural language processing	[issue in Natural Language Processing]	0.0	5	1	1	1
4558	130	translation accuracy	[translation accuracy]	0.0	2	1	1	1
4559	130	two kind	[two kinds]	0.0	2	1	1	1
4560	130	clue such as editing term	[clues such as editing terms]	0.0	5	1	1	1
4561	130	make analysis robust---an agenda-based scheduling parser	[making analysis robust---an agenda-based scheduling parser]	0.0	6	1	1	1
4562	130	reader	[readers]	0.0	1	1	1	1
4563	130	problem include vocabulary	[problems including vocabulary]	0.0	3	1	1	1
4564	130	semantic equivalence classification	[semantic equivalence classification]	0.0	3	1	1	1
4565	130	include jointly conditioning	[including jointly conditioning]	0.0	3	1	1	1
4566	130	Hierarchical Phrasal Lexicon DHPL	[Hierarchical Phrasal Lexicon DHPL]	0.0	4	1	1	1
4567	130	word from alignment model	[words from alignment models]	0.0	4	1	1	1
4568	130	machine readable dictionary	[machine readable dictionaries]	0.0	3	1	1	1
4569	130	source interval	[source interval]	0.0	2	1	1	1
4570	130	newspaper article	[newspaper articles]	0.0	2	1	1	1
4571	130	material the Canadian Hansards	[material the Canadian Hansards]	0.0	4	1	1	1
4572	130	use in case-based mt	[use in case-based MT]	0.0	4	1	1	1
4573	130	efficient parsing of Korean final language	[efficient parsing of Korean final language]	0.0	6	1	1	1
4574	130	use tagging strategy	[using tagging strategies]	0.0	3	1	1	1
4575	130	State	[State]	0.0	1	1	1	1
4576	130	reach into other language	[reach into other languages]	0.0	4	1	1	1
4577	130	sentence in corpus	[sentences in corpora]	0.0	3	1	1	1
4578	130	validity of such database	[validity of such databases]	0.0	4	1	1	1
4579	130	human reasoning	[human reasoning]	0.0	2	1	1	1
4580	130	NE system	[NE system]	0.0	2	1	1	1
4581	130	addressee identification face-to-face meeting	[addressee identification face-to-face meetings]	0.0	4	1	1	1
4582	130	Prolog query optimisation	[Prolog query optimisation]	0.0	3	1	1	1
4583	130	broad use of feature	[broad use of features]	0.0	4	1	1	1
4584	130	scope ambiguity	[scope ambiguity]	0.0	2	1	1	1
4585	130	(BLEU	[(BLEU]	0.0	1	1	1	1
4586	130	make exploratory corpus analysis	[making exploratory corpus analysis]	0.0	4	1	1	1
4587	130	goal build that editor	[goal building that editor]	0.0	4	1	1	1
4588	130	previous tagging result	[previous tagging result]	0.0	3	1	1	1
4589	130	full analysis component	[full analysis component]	0.0	3	1	1	1
4590	130	99% sentence	[99% sentences]	0.0	2	1	1	1
4591	130	easy design	[easy design]	0.0	2	1	1	1
4592	130	three type	[three types]	0.0	2	1	1	1
4593	130	indifference	[indifference]	0.0	1	1	1	1
4594	130	weak equivalence of the formalism	[weak equivalence of the formalisms]	0.0	5	1	1	1
4595	130	lexical rule	[lexical rules]	0.0	2	1	1	1
4596	130	result expression	[resulting expression]	0.0	2	1	1	1
4597	130	linguistic feature	[linguistic features]	0.0	2	1	1	1
4598	130	speech dialogue	[speech dialogue]	0.0	2	1	1	1
4599	130	million word	[million word]	0.0	2	1	1	1
4600	130	convey the meaning	[conveying the meaning]	0.0	3	1	1	1
4601	130	kind of misunderstanding between speaker	[kinds of misunderstandings between speaker]	0.0	5	1	1	1
4602	130	understanding system	[understanding system]	0.0	2	1	1	1
4603	130	head-oriented notion	[head-oriented notions]	0.0	2	1	1	1
4604	130	interesting combination of resource	[interesting combination of resources]	0.0	4	1	1	1
4605	130	domain dependence	[domain dependence]	0.0	2	1	1	1
4606	130	million	[millions]	0.0	1	1	1	1
4607	130	experiment corresponding	[experiments corresponding]	0.0	2	1	1	1
4608	130	many form	[many forms]	0.0	2	1	1	1
4609	130	interactive disambiguation scheme	[interactive disambiguation scheme]	0.0	3	1	1	1
4610	130	substring parsing	[substring parsing]	0.0	2	1	1	1
4611	130	output selection	[output selection]	0.0	2	1	1	1
4612	130	way deal with complexity	[ways dealing with complexity]	0.0	4	1	1	1
4613	130	significant efficiency gain	[significant efficiency gains]	0.0	3	1	1	1
4614	130	information .g	[information .g]	0.0	2	1	1	1
4615	130	find the candidate	[Finding the candidates]	0.0	3	1	1	1
4616	130	machine translation (smt)	[machine translation (SMT)]	0.0	3	1	1	1
4617	130	harvesting English-Chinese bitext	[harvesting English-Chinese bitexts]	0.0	3	1	1	1
4618	130	performance in application	[performance in applications]	0.0	3	1	1	1
4619	130	discover sentence in corpus	[discovering sentences in corpora]	0.0	4	1	1	1
4620	130	performance enhancing technique include part	[performance enhancing techniques including part]	0.0	5	1	1	1
4621	130	only edge with a semantic interpretation	[only edges with a semantic interpretation]	0.0	6	1	1	1
4622	130	theory of reasoning	[theory of reasoning]	0.0	3	1	1	1
4623	130	speech recognition of a language	[speech recognition of a language]	0.0	5	1	1	1
4624	130	very small corpus word	[very small corpus words]	0.0	4	1	1	1
4625	130	sense of much of the sentence	[sense of much of the sentence]	0.0	6	1	1	1
4626	130	two standard output system entity annotation	[two standard outputs systems entity annotations]	0.0	6	1	1	1
4627	130	fundamentally different strategy	[fundamentally different strategies]	0.0	3	1	1	1
4628	130	conglomeration of lexical rule	[conglomeration of lexical rules]	0.0	4	1	1	1
4629	130	practical way deal	[practical ways dealing]	0.0	3	1	1	1
4630	130	coherent speech recognition hypothese	[coherent speech recognition hypotheses]	0.0	4	1	1	1
4631	130	variation between the training corpus	[variations between the training corpus]	0.0	5	1	1	1
4632	130	learning phrase long than	[learning phrases longer than]	0.0	4	1	1	1
4633	130	grolier'	[Grolier's]	0.0	1	1	1	1
4634	130	parallel corpus word	[parallel corpus words]	0.0	3	1	1	1
4635	130	basic	[basics]	0.0	1	2	2	1
4636	130	SI corpus	[SI corpus]	0.0	2	1	1	1
4637	130	index such as the time	[indices such as the time]	0.0	5	1	1	1
4638	130	nlp component	[NLP components]	0.0	2	1	1	1
4639	130	condition for the use of expression	[conditions for the use of expressions]	0.0	6	1	1	1
4640	130	inferential theory 1988/	[inferential theory 1988/]	0.0	3	1	1	1
4641	130	quantifier most	[quantifier most]	0.0	2	1	1	1
4642	130	exist parser	[existing parser]	0.0	2	1	1	1
4643	130	preliminary user evaluation	[preliminary user evaluation]	0.0	3	1	1	1
4644	130	Korean Phonology structure Grammar	[Korean Phonology Structure Grammar]	0.0	4	1	1	1
4645	130	tier	[tiers]	0.0	1	1	1	1
4646	130	generate from redundancy	[generating from redundancy]	0.0	3	1	1	1
4647	130	academic writing	[academic writing]	0.0	2	1	1	1
4648	130	Lucy a portable understanding system	[Lucy a portable understanding system]	0.0	5	1	1	1
4649	130	additional ,000 feature	[additional ,000 features]	0.0	3	1	1	1
4650	130	SENSEVAL-2 English sample task	[SENSEVAL-2 English sample task]	0.0	4	1	1	1
4651	130	construct a system speech recognition	[constructing a system speech recognition]	0.0	5	1	1	1
4652	130	method visualize	[method visualizing]	0.0	2	1	1	1
4653	130	chart-based phrase structure	[chart-based phrase structure]	0.0	3	1	1	1
4654	130	still valuable method	[still valuable methods]	0.0	3	1	1	1
4655	130	ii quality translation via word sense	[ii quality translation via word sense]	0.0	6	1	1	1
4656	130	appealing alternative—in term of simplicity	[appealing alternative—in terms of simplicity]	0.0	5	1	1	1
4657	130	include part	[including part]	0.0	2	1	1	1
4658	130	current capability	[current capabilities]	0.0	2	1	1	1
4659	130	describe the processing in a discourse	[describing the processing in a discourse]	0.0	6	1	1	1
4660	130	frequently good predictor of document descriptor	[frequently good predictors of document descriptors]	0.0	6	1	1	1
4661	130	isomorphic Grammar approach	[Isomorphic Grammars approach]	0.0	3	1	1	1
4662	130	general-purpose language	[general-purpose language]	0.0	2	1	1	1
4663	130	one superiority over method	[one superiority over methods]	0.0	4	1	1	1
4664	130	complete lexical distribution	[Complete lexical distribution]	0.0	3	1	1	1
4665	130	corresponding to arc	[corresponding to arcs]	0.0	3	1	1	1
4666	130	flexible inference method	[flexible inference method]	0.0	3	1	1	1
4667	130	lexical distribution	[lexical distribution]	0.0	2	1	1	1
4668	130	SENSEVAL-2 noun	[SENSEVAL-2 nouns]	0.0	2	1	1	1
4669	130	error correction of input	[error correction of input]	0.0	4	1	1	1
4670	130	layout of a apartment	[layout of an apartment]	0.0	4	1	1	1
4671	130	identification from a corpus	[identification from a corpus]	0.0	4	1	1	1
4672	130	meaning of the sentence	[meaning of the sentence]	0.0	4	1	1	1
4673	130	common evaluation	[common evaluation]	0.0	2	1	1	1
4674	130	output of a probabilistic parser	[output of an probabilistic parser]	0.0	5	1	1	1
4675	130	lexicon with entry	[lexicon with entries]	0.0	3	1	1	1
4676	130	technique for training module	[Techniques for training modules]	0.0	4	1	1	1
4677	130	pooling	[pooling]	0.0	1	1	1	1
4678	130	zero pronoun	[zero pronouns]	0.0	2	1	1	1
4679	130	phonetic modelling	[phonetic modelling]	0.0	2	1	1	1
4680	130	output sentence plan	[output sentence plans]	0.0	3	1	1	1
4681	130	pattern in translation datum	[patterns in translation data]	0.0	4	1	1	1
4682	130	interactive method for speech understanding	[interactive method for speech understanding]	0.0	5	1	1	1
4683	130	acquire new stem from a word	[acquiring new stems from a word]	0.0	6	1	1	1
4684	130	descriptor from phrase	[descriptors from phrases]	0.0	3	1	1	1
4685	130	proposition p	[proposition p]	0.0	2	1	1	1
4686	130	i robust efficient parsing a verb	[i Robust efficient parsing a verb]	0.0	6	1	1	1
4687	130	complexity of some	[complexity of some]	0.0	3	1	1	1
4688	130	chinese-to-english SMT model	[Chinese-to-English SMT model]	0.0	3	1	1	1
4689	130	linguistic feature of question	[linguistic features of questions]	0.0	4	1	1	1
4690	130	belief about the world	[belief about the world]	0.0	4	1	1	1
4691	130	tree-to-tree transducer	[tree-to-tree transducer]	0.0	2	1	1	1
4692	130	diverse pool	[diverse pool]	0.0	2	1	1	1
4693	130	number of evaluation	[number of evaluations]	0.0	3	1	1	1
4694	130	inherently a conglomeration	[inherently a conglomeration]	0.0	3	1	1	1
4695	130	produce such phrase from a corpus	[producing such phrases from a corpora]	0.0	6	1	1	1
4696	130	use of prior	[use of priors]	0.0	3	1	1	1
4697	130	time solution	[time solution]	0.0	2	1	1	1
4698	130	prediction of fragment	[predictions of fragments]	0.0	3	1	1	1
4699	130	framework restrictive than early one	[framework restrictive than earlier ones]	0.0	5	1	1	1
4700	130	use interpolation of n-gram	[using interpolation of N-grams]	0.0	4	1	1	1
4701	130	use a binary calculus notation	[using a binary calculus notation]	0.0	5	1	1	1
4702	130	such mt	[such MT]	0.0	2	1	1	1
4703	130	evidence from a additional feature	[evidence from an additional features]	0.0	5	1	1	1
4704	130	New result	[New results]	0.0	2	1	1	1
4705	130	interpretation of language	[interpretation of language]	0.0	3	1	1	1
4706	130	5 speed-up over a SUN	[5 speed-up over a SUN]	0.0	5	1	1	1
4707	130	place of a semantics	[place of a semantics]	0.0	4	1	1	1
4708	130	input expression	[input expression]	0.0	2	1	1	1
4709	130	interactive question-answering system	[interactive question-answering system]	0.0	3	1	1	1
4710	130	explicit development model	[explicit development model]	0.0	3	1	1	1
4711	130	transfer component	[transfer component]	0.0	2	1	1	1
4712	130	new part-of-speech tagger	[new part-of-speech tagger]	0.0	3	1	1	1
4713	130	response adaptive to individual user	[responses adaptive to individual users]	0.0	5	1	1	1
4714	130	human understanding	[human understanding]	0.0	2	1	1	1
4715	130	validity of such complex database	[validity of such complex databases]	0.0	5	1	1	1
4716	130	BU recognition system	[BU recognition system]	0.0	3	1	1	1
4717	130	actual tape-recorded description	[actual tape-recorded descriptions]	0.0	3	1	1	1
4718	130	learning phrase word from alignment model	[learning phrases words from alignment models]	0.0	6	1	1	1
4719	130	2 decision tree learning	[2 decision tree learning]	0.0	4	1	1	1
4720	130	recovery technique	[recovery technique]	0.0	2	1	1	1
4721	130	performance gain	[performance gains]	0.0	2	1	1	1
4722	130	FERRET a question-answering system	[FERRET an question-answering system]	0.0	4	1	1	1
4723	130	gather statistic	[gathering statistics]	0.0	2	1	1	1
4724	130	new inference method	[new inference method]	0.0	3	1	1	1
4725	130	extraposition	[extraposition]	0.0	1	1	1	1
4726	130	contextual clue	[contextual clues]	0.0	2	1	1	1
4727	130	amount of internal lexical variation	[amount of internal lexical variation]	0.0	5	1	1	1
4728	130	cast doubt on this assumption	[casting doubt on this assumption]	0.0	5	1	1	1
4729	130	parsing language with case marker	[parsing language with case markers]	0.0	5	1	1	1
4730	130	grammaticality 99% correct sentence	[grammaticality 99% correct sentences]	0.0	4	1	1	1
4731	130	one superiority over exist method	[one superiority over existing methods]	0.0	5	1	1	1
4732	130	apply mt evaluation method (BLEU	[applying MT evaluation methods (BLEU]	0.0	5	1	1	1
4733	130	property of tree-adjoining grammar	[properties of tree-adjoining grammars]	0.0	4	1	1	1
4734	130	framework of collaboration	[framework of collaborations]	0.0	3	1	1	1
4735	130	alternative—in term	[alternative—in terms]	0.0	2	1	1	1
4736	130	use in such mt cbmt	[use in such MT CBMT]	0.0	5	1	1	1
4737	130	most dominant part	[most dominant part]	0.0	3	1	1	1
4738	130	common Coalition Language System	[Common Coalition Language System]	0.0	4	1	1	1
4739	130	kind of linguistic knowledge	[kinds of linguistic knowledge]	0.0	4	1	1	1
4740	130	average precision over text	[average precision over text]	0.0	4	1	1	1
4741	130	avoid failure	[avoiding failures]	0.0	2	1	1	1
4742	130	good single tagging result	[best single tagging result]	0.0	4	1	1	1
4743	130	most probable morpheme sequence	[most probable morpheme sequence]	0.0	4	1	1	1
4744	130	corpus of 100 word	[corpus of 100 words]	0.0	4	1	1	1
4745	130	traditional word alignment technique	[traditional word alignment techniques]	0.0	4	1	1	1
4746	130	independently valuable nlp component	[independently valuable NLP components]	0.0	4	1	1	1
4747	130	achieve performance to mixture of technique	[achieving performance to mixtures of techniques]	0.0	6	1	1	1
4748	130	basis for the acquisition	[basis for the acquisition]	0.0	4	1	1	1
4749	130	set of rule	[sets of rules]	0.0	3	1	1	1
4750	130	efficient parsing final language	[efficient parsing final language]	0.0	4	1	1	1
4751	130	lexical gap	[lexical gap]	0.0	2	1	1	1
4752	130	correct chain	[correct chain]	0.0	2	1	1	1
4753	130	structure contain set	[structures containing sets]	0.0	3	1	1	1
4754	130	output unsegmented text	[outputting unsegmented texts]	0.0	3	1	1	1
4755	130	concept-based seed	[concept-based seeds]	0.0	2	1	1	1
4756	130	amount superior	[amount superior]	0.0	2	1	1	1
4757	130	machine translation test	[machine translation tests]	0.0	3	1	1	1
4758	130	simple application	[simple application]	0.0	2	1	1	1
4759	130	potential increase	[potential increase]	0.0	2	1	1	1
4760	130	conceptual Dependency diagram	[Conceptual Dependency Diagram]	0.0	3	1	1	1
4761	130	type of deterministic automaton	[type of deterministic automaton]	0.0	4	1	1	1
4762	130	CCLINC Korean-to-English	[CCLINC Korean-to-English]	0.0	2	1	1	1
4763	130	system with intelligent mobile agent	[system with intelligent mobile agents]	0.0	5	1	1	1
4764	130	common ground	[common ground]	0.0	2	1	1	1
4765	130	bear	[bearing]	0.0	1	1	1	1
4766	130	NE performance	[NE performance]	0.0	2	1	1	1
4767	130	use of precede tag context	[use of preceding tag contexts]	0.0	5	1	1	1
4768	130	language text email .g	[language texts e .g]	0.0	4	1	1	1
4769	130	especially name	[especially names]	0.0	2	1	1	1
4770	130	small corpus of word	[small corpus of words]	0.0	4	1	1	1
4771	130	bilingual parallel corpus a available resource	[bilingual parallel corpora a available resource]	0.0	6	1	1	1
4772	130	number of metaphor mapping	[number of metaphor mappings]	0.0	4	1	1	1
4773	130	runtime	[runtime]	0.0	1	1	1	1
4774	130	include unlimited vocabulary	[including unlimited vocabulary]	0.0	3	1	1	1
4775	130	minor error	[minor errors]	0.0	2	1	1	1
4776	130	high-accuracy alignment model	[high-accuracy alignment models]	0.0	3	1	1	1
4777	130	construction of lexica	[construction of lexica]	0.0	3	1	1	1
4778	130	bilingual corpus a available resource	[bilingual corpora a available resource]	0.0	5	1	1	1
4779	130	three method .e lsa-based method	[three methods .e LSA-based methods]	0.0	5	1	1	1
4780	130	tagger use	[tagger use]	0.0	2	1	1	1
4781	130	answer to everyday question	[answers to everyday questions]	0.0	4	1	1	1
4782	130	more datum relevant to task	[more data relevant to task]	0.0	5	1	1	1
4783	130	two phase	[two phases]	0.0	2	1	1	1
4784	130	Hierarchical Acyclic Graph Kernel	[Hierarchical Acyclic Graph Kernel]	0.0	4	1	1	1
4785	130	two new system one	[two new systems one]	0.0	4	1	1	1
4786	130	learning algorithm for identification	[learning algorithm for identification]	0.0	4	1	1	1
4787	130	determine the meaning of input	[determining the meaning of input]	0.0	5	1	1	1
4788	130	COMPLEX	[COMPLEX]	0.0	1	1	1	1
4789	130	primarily on information from dictionary	[primarily on information from dictionaries]	0.0	5	1	1	1
4790	130	compare the linguistic expressiveness	[comparing the linguistic expressiveness]	0.0	4	1	1	1
4791	130	top-down scope information	[top-down scope information]	0.0	3	1	1	1
4792	130	tagger explicit use	[tagger explicit use]	0.0	3	1	1	1
4793	130	alphabet	[alphabets]	0.0	1	1	1	1
4794	130	interaction requirement	[interaction requirements]	0.0	2	1	1	1
4795	130	possible sentence plan	[possible sentence plans]	0.0	3	1	1	1
4796	130	key factor include inherent limitation	[key factors including inherent limitations]	0.0	5	1	1	1
4797	130	NIST mt evaluation software	[NIST MT evaluation software]	0.0	4	1	1	1
4798	130	develop polynomial time approximation	[developing polynomial time approximations]	0.0	4	1	1	1
4799	130	alignment quality	[alignment quality]	0.0	2	1	1	1
4800	130	domain model	[domain models]	0.0	2	1	1	1
4801	130	two problem a attribute	[two problems a attributes]	0.0	4	1	1	1
4802	130	use a symbol	[using a symbol]	0.0	3	1	1	1
4803	130	three component	[three components]	0.0	2	1	1	1
4804	130	meeting use Network classifier	[meetings using Network classifiers]	0.0	4	1	1	1
4805	130	iii system development	[iii system development]	0.0	3	1	1	1
4806	130	multiple meaning ambiguity	[multiple meanings ambiguity]	0.0	3	1	1	1
4807	130	result of this experiment	[results of this experiment]	0.0	4	1	1	1
4808	130	frame semantics paradigm	[frame semantics paradigm]	0.0	3	1	1	1
4809	130	difficult SENSEVAL-2 noun	[difficult SENSEVAL-2 nouns]	0.0	3	1	1	1
4810	130	identify word in newspaper	[identifying words in newspapers]	0.0	4	1	1	1
4811	130	commercial system	[commercial systems]	0.0	2	1	1	1
4812	130	gather information related	[gathering information related]	0.0	3	1	1	1
4813	130	correct homophone error	[correcting homophone errors]	0.0	3	1	1	1
4814	130	frequently individually good predictor	[frequently individually good predictors]	0.0	4	1	1	1
4815	130	good baseline result	[good baseline results]	0.0	3	1	1	1
4816	130	characteristic	[characteristics]	0.0	1	1	1	1
4817	130	role of purpose	[role of purpose]	0.0	3	1	1	1
4818	130	problem attribute	[problems attributes]	0.0	2	1	1	1
4819	130	underspecified semantic representation	[underspecified semantic representation]	0.0	3	1	1	1
4820	130	language datum	[language data]	0.0	2	1	1	1
4821	130	method of natural language interface	[methods of natural language interfaces]	0.0	5	1	1	1
4822	130	abstracting industry	[abstracting industry]	0.0	2	1	1	1
4823	130	language essay	[language essays]	0.0	2	1	1	1
4824	130	representational format for meaning	[representational format for meaning]	0.0	4	1	1	1
4825	130	iv modeling of unknown word feature	[iv modeling of unknown word features]	0.0	6	1	1	1
4826	130	three capability	[three capabilities]	0.0	2	1	1	1
4827	130	key feature	[key features]	0.0	2	1	1	1
4828	130	numerous aspect of communication	[numerous aspects of communication]	0.0	4	1	1	1
4829	130	cluster number	[cluster number]	0.0	2	1	1	1
4830	130	form language	[form language]	0.0	2	1	1	1
4831	130	use 40 utterance for adaptation	[Using 40 utterances for adaptation]	0.0	5	1	1	1
4832	130	,000 feature	[,000 features]	0.0	2	1	1	1
4833	130	more precise description	[more precise description]	0.0	3	1	1	1
4834	130	approach to machine translation	[approach to machine translation]	0.0	4	1	1	1
4835	130	statistical mt system	[statistical MT systems]	0.0	3	1	1	1
4836	130	COMPLEX a computational lexicon	[COMPLEX a computational lexicon]	0.0	4	1	1	1
4837	130	sequence in the pattern	[sequence in the pattern]	0.0	4	1	1	1
4838	130	representation for document classification	[representations for document classification]	0.0	4	1	1	1
4839	130	normal form	[normal form]	0.0	2	1	1	1
4840	130	case study	[case study]	0.0	2	1	1	1
4841	130	2 a response	[2 a response]	0.0	3	1	1	1
4842	130	include jointly conditioning on word	[including jointly conditioning on words]	0.0	5	1	1	1
4843	130	typical SMT model	[typical SMT models]	0.0	3	1	1	1
4844	130	training datum suitable	[training data suitable]	0.0	3	1	1	1
4845	130	Grammar approach to mt	[Grammars approach to MT]	0.0	4	1	1	1
4846	130	paradigm for speaker-independent SI training	[paradigm for speaker-independent SI training]	0.0	5	1	1	1
4847	130	training set	[training set]	0.0	2	1	1	1
4848	130	beginning	[beginning]	0.0	1	1	1	1
4849	130	choice of structure	[choice of structure]	0.0	3	1	1	1
4850	130	two contribution to large vocabulary	[two contributions to large vocabulary]	0.0	5	1	1	1
4851	130	parsing toward meaning	[parsing toward meanings]	0.0	3	1	1	1
4852	130	aggregation	[aggregation]	0.0	1	1	1	1
4853	130	recent claim to the contrary	[recent claims to the contrary]	0.0	5	1	1	1
4854	130	increasingly common claim	[increasingly common claim]	0.0	3	1	1	1
4855	130	Language	[Language]	0.0	1	1	1	1
4856	130	inherently a conglomeration of rule	[inherently a conglomeration of rules]	0.0	5	1	1	1
4857	130	strong impact	[strong impact]	0.0	2	1	1	1
4858	130	English analyst	[English analyst]	0.0	2	1	1	1
4859	130	multiple corpus	[multiple corpora]	0.0	2	1	1	1
4860	130	specification of a grammar model	[specification of an grammar model]	0.0	5	1	1	1
4861	130	application of metric	[application of metrics]	0.0	3	1	1	1
4862	130	question-answering q/a system	[question-answering Q/A system]	0.0	3	1	1	1
4863	130	maximum a posteriori estimation	[maximum a posteriori estimation]	0.0	4	1	1	1
4864	130	classify undisambiguated SCF datum	[classifying undisambiguated SCF data]	0.0	4	1	1	1
4865	130	automatic q/a application	[automatic Q/A applications]	0.0	3	1	1	1
4866	130	correlation of the measure	[correlation of the measure]	0.0	4	1	1	1
4867	130	parser evaluation method	[parser evaluation methods]	0.0	3	1	1	1
4868	130	110 word	[110 words]	0.0	2	1	1	1
4869	130	node with great than one	[node with greater than one]	0.0	5	1	1	1
4870	130	mt-based paraphrasing technique	[MT-based paraphrasing technique]	0.0	3	1	1	1
4871	130	different information	[different information]	0.0	2	1	1	1
4872	130	construct a system	[constructing a system]	0.0	3	1	1	1
4873	130	good predictor of document descriptor keyword	[good predictors of document descriptors keywords]	0.0	6	1	1	1
4874	130	'synchronous'	['synchronous']	0.0	1	1	1	1
4875	130	relevant aspect	[relevant aspects]	0.0	2	1	1	1
4876	130	new disambiguation system one	[new disambiguation systems one]	0.0	4	1	1	1
4877	130	problem a attribute	[problems a attributes]	0.0	3	1	1	1
4878	130	independent summarization approach	[independent summarization approaches]	0.0	3	1	1	1
4879	130	Graph	[Graph]	0.0	1	1	1	1
4880	130	start with a parallel corpus	[starting with a parallel corpus]	0.0	5	1	1	1
4881	130	research on major problem	[research on major problems]	0.0	4	1	1	1
4882	130	floppy	[floppy]	0.0	1	1	1	1
4883	130	predict top-level topic shift	[predicting top-level topic shifts]	0.0	4	1	1	1
4884	130	top-level boundary	[top-level boundaries]	0.0	2	1	1	1
4885	130	new methodology train spot	[new methodology training SPoT]	0.0	4	1	1	1
4886	130	morpheme sequence for a input	[morpheme sequence for a input]	0.0	5	1	1	1
4887	130	grammaticality 99% correct sentence equivalence	[grammaticality 99% correct sentences equivalence]	0.0	5	1	1	1
4888	130	separate component the structure	[separate components the structure]	0.0	4	1	1	1
4889	130	such edge as in treatment	[such edges as in treatments]	0.0	5	1	1	1
4890	130	though rarely in the context	[though rarely in the context]	0.0	5	1	1	1
4891	130	.0% improvement in the number	[.0% improvement in the number]	0.0	5	1	1	1
4892	130	acquisition of information	[acquisition of information]	0.0	3	1	1	1
4893	130	word-sense system	[word-sense systems]	0.0	2	1	1	1
4894	130	use a calculus notation	[using a calculus notation]	0.0	4	1	1	1
4895	130	.1% a 45% reduction	[.1% a 45% reduction]	0.0	4	1	1	1
4896	130	experiment corresponding to semantic domain	[experiments corresponding to semantic domains]	0.0	5	1	1	1
4897	130	purpose a node	[purposes a node]	0.0	3	1	1	1
4898	130	also a generalization	[also a generalization]	0.0	3	1	1	1
4899	130	vagueness in annotation	[vagueness in annotation]	0.0	3	1	1	1
4900	130	speech information	[speech information]	0.0	2	1	1	1
4901	130	explicit information from machine-readable dictionary	[explicit information from machine-readable dictionaries]	0.0	5	1	1	1
4902	130	case of summarization	[case of summarization]	0.0	3	1	1	1
4903	130	discourse through the operation	[discourse through the operation]	0.0	4	1	1	1
4904	130	particular class	[particular class]	0.0	2	1	1	1
4905	130	two stage	[two stages]	0.0	2	1	1	1
4906	130	important bear	[important bearing]	0.0	2	1	1	1
4907	130	word n-gram	[word n-grams]	0.0	2	1	1	1
4908	130	meaning representation	[meaning representation]	0.0	2	1	1	1
4909	130	exist method	[existing methods]	0.0	2	1	1	1
4910	130	non-terminal symbol	[non-terminal symbol]	0.0	2	1	1	1
4911	130	adaptive word formation	[adaptive word formation]	0.0	3	1	1	1
4912	130	Recent advance in speech Recognition technology	[Recent advances in Speech Recognition technology]	0.0	6	1	1	1
4913	130	commercial system output text	[commercial systems outputting texts]	0.0	4	1	1	1
4914	130	several factor	[several factors]	0.0	2	1	1	1
4915	130	biological warfare	[biological warfare]	0.0	2	1	1	1
4916	130	candidate for the homophone error	[candidates for the homophone errors]	0.0	5	1	1	1
4917	130	list of sentence plan	[list of sentence plans]	0.0	4	1	1	1
4918	130	unsupervised technique learn morphology	[unsupervised technique learning morphology]	0.0	4	1	1	1
4919	130	parallel corpus a more available resource	[parallel corpora a more available resource]	0.0	6	1	1	1
4920	130	conglomeration of rule	[conglomeration of rules]	0.0	3	1	1	1
4921	130	string with the performance	[string with the performance]	0.0	4	1	1	1
4922	130	answer agent search	[answering agents searching]	0.0	3	1	1	1
4923	130	grammar tailoring	[grammar tailoring]	0.0	2	1	1	1
4924	130	discover parallel sentence in non-parallel corpus	[discovering parallel sentences in non-parallel corpora]	0.0	6	1	1	1
4925	130	Parsing Model	[Parsing Model]	0.0	2	1	1	1
4926	130	entity-oriented approach	[entity-oriented approach]	0.0	2	1	1	1
4927	130	common Coalition	[Common Coalition]	0.0	2	1	1	1
4928	130	real-valued feature a language model score	[real-valued features a language model score]	0.0	6	1	1	1
4929	130	produce phrase from a word-aligned corpus	[producing phrases from a word-aligned corpora]	0.0	6	1	1	1
4930	130	prior in loglinear model	[priors in loglinear models]	0.0	4	1	1	1
4931	130	discover sentence in non-parallel corpus	[discovering sentences in non-parallel corpora]	0.0	5	1	1	1
4932	130	class-dependent interpolation of n-gram	[class-dependent interpolation of N-grams]	0.0	4	1	1	1
4933	130	plausible interpretation from a chunk	[plausible interpretation from a chunk]	0.0	5	1	1	1
4934	130	domain independent strategy	[domain independent strategy]	0.0	3	1	1	1
4935	130	make corpus analysis accessible	[making corpus analysis accessible]	0.0	4	1	1	1
4936	130	Coalition Language System at Lincoln Laboratory	[Coalition Language System at Lincoln Laboratory]	0.0	6	1	1	1
4937	130	relationship between the grammatical formalism	[relationship between the grammatical formalisms]	0.0	5	1	1	1
4938	130	promising approach to computer-assisted writing	[promising approach to computer-assisted writing]	0.0	5	1	1	1
4939	130	different aspect of the performance	[different aspects of the performance]	0.0	5	1	1	1
4940	130	transcription of training datum	[transcription of training data]	0.0	4	1	1	1
4941	130	state proprietary stemmer	[state proprietary stemmer]	0.0	3	1	1	1
4942	130	corpus exploration	[corpus exploration]	0.0	2	1	1	1
4943	130	available resource	[available resource]	0.0	2	1	1	1
4944	130	identify pattern	[identifying patterns]	0.0	2	1	1	1
4945	130	experience especially in the muc-3 evaluation	[experience especially in the MUC-3 evaluation]	0.0	6	1	1	1
4946	130	novice user	[novice users]	0.0	2	1	1	1
4947	130	substantial improvement	[substantial improvement]	0.0	2	1	1	1
4948	130	statistical technique	[statistical techniques]	0.0	2	1	1	1
4949	130	numerous non-literal aspect	[numerous non-literal aspects]	0.0	3	1	1	1
4950	130	least 96% correct paraphrase	[least 96% correct paraphrases]	0.0	4	1	1	1
4951	130	powerful computer	[powerful computers]	0.0	2	1	1	1
4952	130	also a drastic generalization	[also a drastic generalization]	0.0	4	1	1	1
4953	130	answer session	[answering session]	0.0	2	1	1	1
4954	130	Research method	[Research Methods]	0.0	2	1	1	1
4955	130	rule-based sentence planner	[rule-based sentence planners]	0.0	3	1	1	1
4956	130	maximum likelihood method	[maximum likelihood method]	0.0	3	1	1	1
4957	130	datum collection paradigm	[data collection paradigm]	0.0	3	1	1	1
4958	130	use corpus a available resource	[using corpora a available resource]	0.0	5	1	1	1
4959	130	likewise automatic evaluation method	[likewise automatic evaluation method]	0.0	4	1	1	1
4960	130	additional feature of the tree	[additional features of the tree]	0.0	5	1	1	1
4961	130	need for answer	[need for answers]	0.0	3	1	1	1
4962	130	property of grammar TAG	[properties of grammars TAG]	0.0	4	1	1	1
4963	130	corpus consist	[corpus consisting]	0.0	2	1	1	1
4964	130	Syntactic analysis	[Syntactic analysis]	0.0	2	1	1	1
4965	130	scoring system	[scoring system]	0.0	2	1	1	1
4966	130	Graph HDAG	[Graph HDAG]	0.0	2	1	1	1
4967	130	combine the style	[combining the styles]	0.0	3	1	1	1
4968	130	skill level	[skill level]	0.0	2	1	1	1
4969	130	output of other tool	[output of other tools]	0.0	4	1	1	1
4970	130	grammaticality sentence	[grammaticality sentences]	0.0	2	1	1	1
4971	130	non-English stemmer	[non-English stemmer]	0.0	2	1	1	1
4972	130	three method construct word vector method	[three methods constructing word vectors methods]	0.0	6	1	1	1
4973	130	500 ,000 feature	[500 ,000 features]	0.0	3	1	1	1
4974	130	source of training datum	[Sources of training data]	0.0	4	1	1	1
4975	130	experiment in a series	[experiment in a series]	0.0	4	1	1	1
4976	130	German corpus	[German corpus]	0.0	2	1	1	1
4977	130	close correlation	[close correlation]	0.0	2	1	1	1
4978	130	new word-sense system one	[new word-sense systems one]	0.0	4	1	1	1
4979	130	optical character recognition model	[optical character recognition model]	0.0	4	1	1	1
4980	130	significantly less development time	[significantly less development time]	0.0	4	1	1	1
4981	130	word from word-level alignment model	[words from word-level alignment models]	0.0	5	1	1	1
4982	130	learning phrase long than word	[learning phrases longer than words]	0.0	5	1	1	1
4983	130	portion use a fake symbol	[portion using a fake symbol]	0.0	5	1	1	1
4984	130	model of the global scene	[model of the global scene]	0.0	5	1	1	1
4985	130	parse record	[parse record]	0.0	2	1	1	1
4986	130	name without title	[names without title]	0.0	3	1	1	1
4987	130	relatively simple means learning	[relatively simple means learning]	0.0	4	1	1	1
4988	130	obviate the need	[obviating the need]	0.0	3	1	1	1
4989	130	translingual reach into language	[translingual reach into languages]	0.0	4	1	1	1
4990	130	maximization	[maximization]	0.0	1	1	1	1
4991	130	result on all application	[results on all applications]	0.0	4	1	1	1
4992	130	direct imitation of performance	[direct imitation of performance]	0.0	4	1	1	1
4993	130	conversation analysis	[conversation analysis]	0.0	2	1	1	1
4994	130	score alternative speech recognition hypothese SRH	[scoring alternative speech recognition hypotheses SRH]	0.0	6	1	1	1
4995	130	new type	[new type]	0.0	2	1	1	1
4996	130	high performance	[high performance]	0.0	2	1	1	1
4997	130	fully-connected first-order class grammar	[fully-connected first-order class grammar]	0.0	4	1	1	1
4998	130	find datum relevant to task	[finding data relevant to task]	0.0	5	1	1	1
4999	130	TAP-XL analyst assistant	[TAP-XL Analyst's Assistant]	0.0	3	1	1	1
5000	130	however the area	[However the area]	0.0	3	1	1	1
5001	130	develop the grammar of the source	[developing the grammars of the Source]	0.0	6	1	1	1
5002	130	reflect the relationship email .g	[reflecting the relationship e .g]	0.0	5	1	1	1
5003	130	contribution to large vocabulary	[contributions to large vocabulary]	0.0	4	1	1	1
5004	130	monitor the collection for use	[monitoring the collection for use]	0.0	5	1	1	1
5005	130	factor include limitation of mt architecture	[factors including limitations of MT architectures]	0.0	6	1	1	1
5006	130	predict the translation	[predicting the translation]	0.0	3	1	1	1
5007	130	track the discourse	[tracking the discourse]	0.0	3	1	1	1
5008	130	sophisticated annotation	[sophisticated annotation]	0.0	2	1	1	1
5009	130	good hypothesis	[best hypothesis]	0.0	2	1	1	1
5010	130	start with a very small corpus	[starting with a very small corpus]	0.0	6	1	1	1
5011	130	i robust parsing a verb language	[i Robust parsing a verb language]	0.0	6	1	1	1
5012	130	paradigm for SI training	[paradigm for SI training]	0.0	4	1	1	1
5013	130	construct a computational system speech recognition	[constructing a computational system speech recognition]	0.0	6	1	1	1
5014	130	12 training speaker	[12 training speakers]	0.0	3	1	1	1
5015	130	such as rewrite system	[such as rewriting systems]	0.0	4	1	1	1
5016	130	stemmer above	[stemmer above]	0.0	2	1	1	1
5017	130	polarization	[polarization]	0.0	1	1	1	1
5018	130	time approximation	[time approximations]	0.0	2	1	1	1
5019	130	combination of various structure	[combination of various structures]	0.0	4	1	1	1
5020	130	finite automaton	[finite automaton]	0.0	2	1	1	1
5021	130	Chinese newspaper corpus	[Chinese newspaper corpora]	0.0	3	1	1	1
5022	130	lsa-based method	[LSA-based methods]	0.0	2	1	1	1
5023	130	imitation of human performance	[imitation of human performance]	0.0	4	1	1	1
5024	130	learning procedure	[learning procedure]	0.0	2	1	1	1
5025	130	impediment	[impediment]	0.0	1	1	1	1
5026	130	use Arabic information retrieval	[using Arabic information retrieval]	0.0	4	1	1	1
5027	130	follow development	[Following developments]	0.0	2	1	1	1
5028	130	method identify pattern	[method identifying patterns]	0.0	3	1	1	1
5029	130	factor of 2	[factor of 2]	0.0	3	1	1	1
5030	130	analysis robust---an scheduling parser	[analysis robust---an scheduling parser]	0.0	4	1	1	1
5031	130	way improve system performance	[way improving system performance]	0.0	4	1	1	1
5032	130	achieve performance	[achieving performance]	0.0	2	1	1	1
5033	130	portion use a non-terminal symbol	[portion using a non-terminal symbol]	0.0	5	1	1	1
5034	130	various training factor	[various training factors]	0.0	3	1	1	1
5035	130	use large knowledgebase	[using large knowledgebases]	0.0	3	1	1	1
5036	130	most expensive part	[most expensive part]	0.0	3	1	1	1
5037	130	process translate a sentence	[process translating a sentence]	0.0	4	1	1	1
5038	130	sentence parallel corpus	[sentences parallel corpus]	0.0	3	1	1	1
5039	130	block pair	[blocks pairs]	0.0	2	1	1	1
5040	130	state of focus	[state of focus]	0.0	3	1	1	1
5041	130	use parallel corpus	[using parallel corpora]	0.0	3	1	1	1
5042	130	repair	[repair]	0.0	1	1	1	1
5043	130	employ a generative phonological approach	[employing a generative phonological approach]	0.0	5	1	1	1
5044	130	remote event	[remote events]	0.0	2	1	1	1
5045	130	predictor of document descriptor keyword	[predictors of document descriptors keywords]	0.0	5	1	1	1
5046	130	sentence utilize example	[sentence utilizing examples]	0.0	3	1	1	1
5047	130	information email .g	[information e .g]	0.0	3	1	1	1
5048	130	close correlation between the evaluation	[close correlation between the evaluation]	0.0	5	1	1	1
5049	130	bootstrapping approach tag	[bootstrapping approach tagging]	0.0	3	1	1	1
5050	130	lexical weighting	[lexical weighting]	0.0	2	1	1	1
5051	130	answer capability	[answering capability]	0.0	2	1	1	1
5052	130	efficient parser for English	[efficient parser for English]	0.0	4	1	1	1
5053	130	several field exercise	[several field exercises]	0.0	3	1	1	1
5054	130	Model for language understanding	[Model for language understanding]	0.0	4	1	1	1
5055	130	Dynamic Hierarchical Phrasal Lexicon	[Dynamic Hierarchical Phrasal Lexicon]	0.0	4	1	1	1
5056	130	quality Ie result	[quality IE results]	0.0	3	1	1	1
5057	130	approach to statistical machine translation	[approach to statistical machine translation]	0.0	5	1	1	1
5058	130	method for the reranking task	[method for the reranking task]	0.0	5	1	1	1
5059	130	model of syntax-based machine translation	[model of syntax-based machine translation]	0.0	5	1	1	1
5060	130	add linguistic annotation	[adding linguistic annotation]	0.0	3	1	1	1
5061	130	summarization quality of the system	[summarization quality of the system]	0.0	5	1	1	1
5062	130	multiple candidate for understanding result	[multiple candidates for understanding results]	0.0	5	1	1	1
5063	130	traditional information	[Traditional information]	0.0	2	1	1	1
5064	130	resolution	[resolution]	0.0	1	2	2	1
5065	130	less development time	[less development time]	0.0	3	1	1	1
5066	130	pose special problem	[posing special problems]	0.0	3	1	1	1
5067	130	valuable nlp component	[valuable NLP components]	0.0	3	1	1	1
5068	130	grammar of the source	[grammars of the Source]	0.0	4	1	1	1
5069	130	small 10k sentence corpus	[small 10K sentences corpus]	0.0	4	1	1	1
5070	130	method construct word vector .e	[methods constructing word vectors .e]	0.0	5	1	1	1
5071	130	pool	[pool]	0.0	1	1	1	1
5072	130	describe inflection	[describing inflection]	0.0	2	1	1	1
5073	130	resolve pronoun	[resolving pronouns]	0.0	2	1	1	1
5074	130	paraphrase ambiguous sentence	[paraphrasing ambiguous sentences]	0.0	3	1	1	1
5075	130	high level	[highest levels]	0.0	2	1	1	1
5076	130	HPM	[HPM]	0.0	1	1	1	1
5077	130	execution at the database	[execution at the database]	0.0	4	1	1	1
5078	130	word with multiple meaning	[words with multiple meanings]	0.0	4	1	1	1
5079	130	Paul	[Paul]	0.0	1	1	1	1
5080	130	scarce resource	[scarce resources]	0.0	2	1	1	1
5081	130	exploratory corpus analysis	[exploratory corpus analysis]	0.0	3	1	1	1
5082	130	experience plan COMPLEX a lexicon	[experience planning COMPLEX a lexicon]	0.0	5	1	1	1
5083	130	standard mt evaluation method	[standard MT evaluation methods]	0.0	4	1	1	1
5084	130	construct a computational system	[constructing a computational system]	0.0	4	1	1	1
5085	130	amount from the target speaker	[amount from the target speaker]	0.0	5	1	1	1
5086	130	conglomeration of individual rule	[conglomeration of individual rules]	0.0	4	1	1	1
5087	130	identify unknown word in newspaper	[identifying unknown words in newspapers]	0.0	5	1	1	1
5088	130	efficient parsing of Korean a verb	[efficient parsing of Korean a verb]	0.0	6	1	1	1
5089	130	explore bitext	[exploring bitexts]	0.0	2	1	1	1
5090	130	disambiguation tool	[disambiguation tool]	0.0	2	1	1	1
5091	130	describe a case study	[describing a case study]	0.0	4	1	1	1
5092	130	acquisition of constraint	[acquisition of constraints]	0.0	3	1	1	1
5093	130	target word	[target word]	0.0	2	1	1	1
5094	130	phrasal translation	[phrasal translation]	0.0	2	1	1	1
5095	130	information redundancy	[information redundancy]	0.0	2	1	1	1
5096	130	algorithm for the boost approach	[algorithm for the boosting approach]	0.0	5	1	1	1
5097	130	logical model in place	[logical model in place]	0.0	4	1	1	1
5098	130	interpolation method like log-linear interpolation	[interpolation methods like log-linear interpolation]	0.0	5	1	1	1
5099	130	standard output from information extraction Ie	[standard outputs from information extraction IE]	0.0	6	1	1	1
5100	130	18 .6% improvement	[18 .6% improvement]	0.0	3	1	1	1
5101	130	importantly Chinese-English text	[importantly Chinese-English texts]	0.0	3	1	1	1
5102	130	extract of newswire text	[extracts of newswire text]	0.0	4	2	2	1
5103	130	automatic extraction of translation lexicon	[automatic extraction of translation lexicons]	0.0	5	1	1	1
5104	130	precise description of feature structure	[precise description of feature structures]	0.0	5	1	1	1
5105	130	standard algorithm	[standard algorithms]	0.0	2	1	1	1
5106	130	respect	[respect]	0.0	1	1	1	1
5107	130	kind of user	[kinds of users]	0.0	3	1	1	1
5108	130	ongoing construction as basis	[ongoing construction as basis]	0.0	4	1	1	1
5109	130	Acyclic Graph	[Acyclic Graph]	0.0	2	1	1	1
5110	130	DARPA speech recognition	[DARPA speech recognition]	0.0	3	2	2	1
5111	130	unification-based grammar parsing	[unification-based grammar parsing]	0.0	3	1	1	1
5112	130	enter through asynchronous message passing	[entering through asynchronous message passing]	0.0	5	1	1	1
5113	130	three word	[three words]	0.0	2	1	1	1
5114	130	superset of L	[superset of L]	0.0	3	1	1	1
5115	130	example paraphrase sentence	[examples paraphrasing sentences]	0.0	3	1	1	1
5116	130	Stabler' formalization of Chomsky program	[Stabler's formalization of Chomsky's program]	0.0	5	1	1	1
5117	130	language generator	[language generator]	0.0	2	1	1	1
5118	130	reach into language	[reach into languages]	0.0	3	1	1	1
5119	130	relational calculus notation	[relational calculus notation]	0.0	3	1	1	1
5120	130	parsing mechanism	[parsing mechanism]	0.0	2	1	1	1
5121	130	first test of a claim	[first test of an claim]	0.0	5	1	1	1
5122	130	method discover sentence	[method discovering sentences]	0.0	3	1	1	1
5123	130	communication need	[communication needs]	0.0	2	1	1	1
5124	130	query optimisation in a database	[query optimisation in a database]	0.0	5	1	1	1
5125	130	such as the time	[such as the time]	0.0	4	1	1	1
5126	130	automatically acquire stem from a word	[automatically acquiring stems from a word]	0.0	6	1	1	1
5127	130	use machine learning technique	[using machine learning techniques]	0.0	4	1	1	1
5128	130	derivation tree	[derivation tree]	0.0	2	1	1	1
5129	130	baseline of 54	[baseline of 54]	0.0	3	1	1	1
5130	130	i robust efficient parsing of Korean	[i Robust efficient parsing of Korean]	0.0	6	1	1	1
5131	130	summarization approach	[summarization approaches]	0.0	2	1	1	1
5132	130	technology of powerful personal computer	[technology of powerful personal computers]	0.0	5	1	1	1
5133	130	reader such as word	[readers such as words]	0.0	4	1	1	1
5134	130	October test set	[October test set]	0.0	3	1	1	1
5135	130	algorithm for unification of feature structure	[algorithms for unification of feature structures]	0.0	6	1	1	1
5136	130	fuzzy quantifier most	[fuzzy quantifier most]	0.0	3	1	1	1
5137	130	presence of a unknown	[presence of a unknown]	0.0	4	1	1	1
5138	130	semantic constraint in domain	[semantic constraints in domains]	0.0	4	1	1	1
5139	130	two new system	[two new systems]	0.0	3	1	1	1
5140	130	viewpoint	[viewpoint]	0.0	1	1	1	1
5141	130	many grammar formalism such as system	[many grammar formalisms such as systems]	0.0	6	1	1	1
5142	130	machine translation method	[machine translation method]	0.0	3	1	1	1
5143	130	broad use include conditioning	[broad use including conditioning]	0.0	4	1	1	1
5144	130	planning-based architecture with a variety	[planning-based architecture with a variety]	0.0	5	1	1	1
5145	130	87 .5% agreement	[87 .5% agreement]	0.0	3	1	1	1
5146	130	automatically train spot	[automatically training SPoT]	0.0	3	1	1	1
5147	130	process in human understanding	[process in human understanding]	0.0	4	1	1	1
5148	130	city	[city]	0.0	1	1	1	1
5149	130	real-time language translation	[real-time language translation]	0.0	3	1	1	1
5150	130	inferential theory	[inferential theory]	0.0	2	1	1	1
5151	130	proposition with quantifier	[proposition with quantifiers]	0.0	3	1	1	1
5152	130	several extension	[Several extensions]	0.0	2	1	1	1
5153	130	understand system with agent	[understanding system with agents]	0.0	4	1	1	1
5154	130	generalization of chart parsing	[generalization of chart Parsing]	0.0	4	1	1	1
5155	130	amount of variation slightly superior	[amount of variation slightly superior]	0.0	5	1	1	1
5156	130	combine the style in a system	[combining the styles in a system]	0.0	6	1	1	1
5157	130	Rapid system development	[Rapid system development]	0.0	3	1	1	1
5158	130	inflow of datum	[inflow of data]	0.0	3	1	1	1
5159	130	optical character recognition ocr model	[optical character recognition OCR model]	0.0	5	1	1	1
5160	130	wsd accuracy	[WSD accuracy]	0.0	2	1	1	1
5161	130	english-spanish fragment	[English-Spanish fragment]	0.0	2	1	1	1
5162	130	include limitation	[including limitations]	0.0	2	1	1	1
5163	130	satisfy communication need	[satisfying communication needs]	0.0	3	1	1	1
5164	130	natural language to semantics	[natural languages to semantics]	0.0	4	1	1	1
5165	130	partial theory	[partial theories]	0.0	2	1	1	1
5166	130	computational phonological system	[computational phonological system]	0.0	3	1	1	1
5167	130	automatic extraction	[automatic extraction]	0.0	2	1	1	1
5168	130	use word-trigram recognition require transcription	[using word-trigram recognition requiring transcription]	0.0	5	1	1	1
5169	130	novel bootstrapping approach	[novel bootstrapping approach]	0.0	3	1	1	1
5170	130	use bilingual parallel corpus	[using bilingual parallel corpora]	0.0	4	1	1	1
5171	130	discourse processing algorithm	[discourse processing algorithms]	0.0	3	1	1	1
5172	130	implication automate aspect	[Implications automating aspects]	0.0	3	1	1	1
5173	130	apply language understand technology	[applying language understanding technology]	0.0	4	1	1	1
5174	130	accuracy difference	[accuracy difference]	0.0	2	1	1	1
5175	130	tree-adjoining grammar	[tree-adjoining grammars]	0.0	2	1	1	1
5176	130	small corpus word	[small corpus words]	0.0	3	1	1	1
5177	130	hypothesis corresponding	[hypothesis corresponding]	0.0	2	1	1	1
5178	130	dependency insertion	[dependency insertion]	0.0	2	1	1	1
5179	130	precise temporal adverb	[precise temporal adverbs]	0.0	3	1	1	1
5180	130	component the structure	[components the structure]	0.0	3	1	1	1
5181	130	design in particular	[designing in particular]	0.0	3	1	1	1
5182	130	efficient parsing a verb language	[efficient parsing a verb language]	0.0	5	1	1	1
5183	130	problem of SMT	[problems of SMT]	0.0	3	1	1	1
5184	130	two output	[two outputs]	0.0	2	1	1	1
5185	130	big a corpus	[bigger a corpus]	0.0	3	1	1	1
5186	130	valuable general-purpose nlp component	[valuable general-purpose NLP components]	0.0	4	1	1	1
5187	130	validity of entry	[validity of entries]	0.0	3	1	1	1
5188	130	baseline sentence planner	[baseline sentence planners]	0.0	3	1	1	1
5189	130	czech-english machine translation system	[Czech-English machine translation system]	0.0	4	1	1	1
5190	130	dynamic word formation	[dynamic word formation]	0.0	3	1	1	1
5191	130	text mining	[text mining]	0.0	2	1	1	1
5192	130	language to semantics	[languages to semantics]	0.0	3	1	1	1
5193	130	use of lexical feature include conditioning	[use of lexical features including conditioning]	0.0	6	1	1	1
5194	130	lexical-syntactic form	[lexical-syntactic forms]	0.0	2	1	1	1
5195	130	training reference	[training reference]	0.0	2	1	1	1
5196	130	unification of feature structure	[unification of feature structures]	0.0	4	1	1	1
5197	130	weak equivalence	[weak equivalence]	0.0	2	1	1	1
5198	130	minute	[minutes]	0.0	1	1	1	1
5199	130	Graph HDAG Kernel	[Graph HDAG Kernel]	0.0	3	1	1	1
5200	130	particular representational format	[particular representational format]	0.0	3	1	1	1
5201	130	complete distribution	[Complete distribution]	0.0	2	1	1	1
5202	130	aim for interpretation	[aiming for interpretations]	0.0	3	1	1	1
5203	130	acquisition of constraint in domain	[acquisition of constraints in domains]	0.0	5	1	1	1
5204	130	Chomsky minimalist program	[Chomsky's minimalist program]	0.0	3	1	1	1
5205	130	use of prior in model	[use of priors in models]	0.0	5	1	1	1
5206	130	top-level subtopic boundary	[top-level subtopic boundaries]	0.0	3	1	1	1
5207	130	new system one	[new systems one]	0.0	3	1	1	1
5208	130	training speaker for SI recognition	[training speakers for SI recognition]	0.0	5	1	1	1
5209	130	model of interaction	[models of interaction]	0.0	3	1	1	1
5210	130	Korean knowledge base system	[Korean knowledge base system]	0.0	4	1	1	1
5211	130	Dependency	[Dependency]	0.0	1	1	1	1
5212	130	newspaper corpus	[newspaper corpora]	0.0	2	1	1	1
5213	130	personal computer	[personal computers]	0.0	2	1	1	1
5214	130	typical IBM mt system	[typical IBM MT system]	0.0	4	1	1	1
5215	130	human-machine interaction	[human-machine interactions]	0.0	2	1	1	1
5216	130	English translation	[English translations]	0.0	2	1	1	1
5217	130	other area	[other areas]	0.0	2	1	1	1
5218	130	approach to writing	[approach to writing]	0.0	3	1	1	1
5219	130	application beyond the confines	[application beyond the confines]	0.0	4	1	1	1
5220	130	(given	[(given]	0.0	1	1	1	1
5221	130	discover sentence	[discovering sentences]	0.0	2	1	1	1
5222	130	require transcription	[requiring transcription]	0.0	2	1	1	1
5223	130	articulation of answer	[articulation of answers]	0.0	3	1	1	1
5224	130	follow recent development	[Following recent developments]	0.0	3	1	1	1
5225	130	quality of summary	[quality of summaries]	0.0	3	1	1	1
5226	130	three component the structure	[three components the structure]	0.0	4	1	1	1
5227	130	Arabic information retrieval	[Arabic information retrieval]	0.0	3	1	1	1
5228	130	linguistic information	[linguistic information]	0.0	2	1	1	1
5229	130	improvement over the baseline	[improvement over the baseline]	0.0	4	1	1	1
5230	130	output from information extraction	[outputs from information extraction]	0.0	4	1	1	1
5231	130	structure for elementary speech act	[structure for elementary speech acts]	0.0	5	1	1	1
5232	130	word sense disambiguation model	[word sense disambiguation model]	0.0	4	1	1	1
5233	130	description draw a space	[descriptions drawing a space]	0.0	4	1	1	1
5234	130	reranking task	[reranking task]	0.0	2	1	1	1
5235	130	Markov chain model	[Markov chain model]	0.0	3	1	1	1
5236	130	language such as Czech	[languages such as Czech]	0.0	4	1	1	1
5237	130	one author	[one author]	0.0	2	1	1	1
5238	130	debate	[debate]	0.0	1	1	1	1
5239	130	link of entry to translation equivalent	[linking of entries to translation equivalents]	0.0	6	1	1	1
5240	130	drastic generalization	[drastic generalization]	0.0	2	1	1	1
5241	130	result on all four application	[results on all four applications]	0.0	5	1	1	1
5242	130	free word order	[free word order]	0.0	3	1	1	1
5243	130	advantage of the strategy	[advantages of the strategies]	0.0	4	1	1	1
5244	130	element in the design	[element in the design]	0.0	4	1	1	1
5245	130	valency constraint	[valency constraints]	0.0	2	1	1	1
5246	130	bootstrapping approach to Entity tag	[bootstrapping approach to Entity tagging]	0.0	5	1	1	1
5247	130	use a word sense disambiguation model	[Using a word sense disambiguation model]	0.0	6	1	1	1
5248	130	particular format	[particular format]	0.0	2	1	1	1
5249	130	,449 word	[,449 word]	0.0	2	1	1	1
5250	130	person's knowledge about knowledge	[person's knowledge about knowledge]	0.0	4	1	1	1
5251	130	pair of probabilistic context-free grammar	[pairs of probabilistic context-free grammars]	0.0	5	1	1	1
5252	130	distance	[distance]	0.0	1	1	1	1
5253	130	system one	[systems one]	0.0	2	1	1	1
5254	130	treatment of syntax	[treatment of syntax]	0.0	3	1	1	1
5255	130	underlying computation model	[underlying computation model]	0.0	3	1	1	1
5256	130	sound dialog system	[sounding dialog systems]	0.0	3	1	1	1
5257	130	relationship between the two formalism	[relationship between the two formalisms]	0.0	5	1	1	1
5258	130	problem attribute such as size perspective-taking	[problems attributes such as size perspective-taking]	0.0	6	1	1	1
5259	130	28 ,449 word token	[28 ,449 word tokens]	0.0	4	1	1	1
5260	130	dictionary on the basis	[dictionary on the basis]	0.0	4	1	1	1
5261	130	style perform inference in network	[styles performing inference in networks]	0.0	5	1	1	1
5262	130	such edge	[such edges]	0.0	2	1	1	1
5263	130	speaker adaptation SA	[speaker adaptation SA]	0.0	3	1	1	1
5264	130	use this model	[using these models]	0.0	3	1	1	1
5265	130	guidance	[guidance]	0.0	1	1	1	1
5266	130	instance to the task	[instance to the task]	0.0	4	1	1	1
5267	130	phrase-based translation model	[phrase-based translation models]	0.0	3	1	1	1
5268	130	improvement in precision over text	[improvement in precision over text]	0.0	5	1	1	1
5269	130	ne rule	[NE rules]	0.0	2	1	1	1
5270	130	Unification categorial Grammar	[Unification Categorial Grammar]	0.0	3	1	1	1
5271	130	language text	[language texts]	0.0	2	2	2	1
5272	130	roughly corresponding to arc	[roughly corresponding to arcs]	0.0	4	1	1	1
5273	130	error reduction	[error reduction]	0.0	2	1	1	1
5274	130	case-based mt	[case-based MT]	0.0	2	1	1	1
5275	130	method add linguistic annotation	[method adding linguistic annotation]	0.0	4	1	1	1
5276	130	simple sentence-plan-generator	[simple sentence-plan-generator]	0.0	2	1	1	1
5277	130	sound dialog system within reach	[sounding dialog systems within reach]	0.0	5	1	1	1
5278	130	current syntax formalism	[current syntax formalisms]	0.0	3	1	1	1
5279	130	use in mt cbmt	[use in MT CBMT]	0.0	4	1	1	1
5280	130	link of lexical entry	[linking of lexical entries]	0.0	4	1	1	1
5281	130	machine translation evaluation measure	[machine translation evaluation measures]	0.0	4	1	1	1
5282	130	Acyclic Graph HDAG Kernel	[Acyclic Graph HDAG Kernel]	0.0	4	1	1	1
5283	130	analyze sentence	[analyzing sentences]	0.0	2	1	1	1
5284	130	Unification Grammar	[Unification Grammar]	0.0	2	1	1	1
5285	130	good choice	[good choice]	0.0	2	1	1	1
5286	130	vehicle	[vehicle]	0.0	1	1	1	1
5287	130	explanation of discourse phenomenon	[explanation of discourse phenomena]	0.0	4	1	1	1
5288	130	authoritative text collection	[authoritative text collection]	0.0	3	1	1	1
5289	130	score speech recognition hypothese	[scoring speech recognition hypotheses]	0.0	4	1	1	1
5290	130	word like sentence	[words like sentence]	0.0	3	1	1	1
5291	130	global meaning	[global meaning]	0.0	2	1	1	1
5292	130	bottom-up pattern-matching parser	[bottom-up pattern-matching parser]	0.0	3	1	1	1
5293	130	flexible architecture	[flexible architecture]	0.0	2	1	1	1
5294	130	Dynamic Phrasal Lexicon DHPL	[Dynamic Phrasal Lexicon DHPL]	0.0	4	1	1	1
5295	130	algorithm acquire stem	[algorithm acquiring stems]	0.0	3	1	1	1
5296	130	parse of a sentence	[parse of a sentence]	0.0	4	1	1	1
5297	130	.0% improvement over baseline system	[.0% improvement over baseline system]	0.0	5	1	1	1
5298	130	Parsing Model HPM	[Parsing Model HPM]	0.0	3	1	1	1
5299	130	extension of this idea	[extensions of this idea]	0.0	4	1	1	1
5300	130	different aspect include the influence	[different aspects including the influence]	0.0	5	1	1	1
5301	130	only edge with a valid interpretation	[only edges with a valid interpretation]	0.0	6	1	1	1
5302	130	multimedia datum	[multimedia data]	0.0	2	1	1	1
5303	130	corpus 100 word	[corpus 100 words]	0.0	3	1	1	1
5304	130	voice system	[voice system]	0.0	2	1	1	1
5305	130	approach to computer-assisted writing	[approach to computer-assisted writing]	0.0	4	1	1	1
5306	130	frequently predictor of document descriptor keyword	[frequently predictors of document descriptors keywords]	0.0	6	1	1	1
5307	130	challenge integrate q/a application	[challenges integrating Q/A applications]	0.0	4	1	1	1
5308	130	model of natural language	[model of natural language]	0.0	4	1	1	1
5309	130	Hierarchical Phrasal Lexicon [Zernik88]	[Hierarchical Phrasal Lexicon [Zernik88]]	0.0	4	1	1	1
5310	130	English understanding system	[English understanding system]	0.0	3	1	1	1
5311	130	word formation	[word formation]	0.0	2	1	1	1
5312	130	very small parallel corpus	[very small parallel corpus]	0.0	4	1	1	1
5313	130	follow development in the evaluation	[Following developments in the evaluation]	0.0	5	1	1	1
5314	130	coverage English grammar	[coverage English grammar]	0.0	3	1	1	1
5315	130	issue in language processing	[issue in Language Processing]	0.0	4	1	1	1
5316	130	use additional feature	[using additional features]	0.0	3	1	1	1
5317	130	term on feature selection method	[terms on feature selection methods]	0.0	5	1	1	1
5318	130	engine machine translation system	[engine machine translation system]	0.0	4	1	1	1
5319	130	Model for natural language understanding	[Model for natural language understanding]	0.0	5	1	1	1
5320	130	one element in the design	[one element in the design]	0.0	5	1	1	1
5321	130	standard output	[standard outputs]	0.0	2	1	1	1
5322	130	identify systematic pattern	[identifying systematic patterns]	0.0	3	1	1	1
5323	130	Chinese newspaper	[Chinese newspapers]	0.0	2	1	1	1
5324	130	2-character Chinese name	[2-character Chinese names]	0.0	3	1	1	1
5325	130	combiner with decision	[combiner with decisions]	0.0	3	1	1	1
5326	130	top-level topic shift	[top-level topic shifts]	0.0	3	1	1	1
5327	130	use include jointly conditioning	[use including jointly conditioning]	0.0	4	1	1	1
5328	130	first experiment	[first experiment]	0.0	2	1	1	1
5329	130	imitation	[imitation]	0.0	1	1	1	1
5330	130	linguistic representation the Phrasal Lexicon	[linguistic representation the Phrasal Lexicon]	0.0	5	1	1	1
5331	130	protocol from answer to question	[protocols from answers to questions]	0.0	5	1	1	1
5332	130	collect statistic	[collecting statistics]	0.0	2	1	1	1
5333	130	cooperative response adaptive to user	[cooperative responses adaptive to users]	0.0	5	1	1	1
5334	130	protocol from people' answer	[protocols from people's answers]	0.0	4	1	1	1
5335	130	term of grammar	[terms of Grammars]	0.0	3	1	1	1
5336	130	cohesion-based approach alone	[cohesion-based approach alone]	0.0	3	1	1	1
5337	130	combination of resource	[combination of resources]	0.0	3	1	1	1
5338	130	different strategy utilize knowledge-based mechanism	[different strategies utilizing knowledge-based mechanisms]	0.0	5	1	1	1
5339	130	syntactic disambiguation	[syntactic disambiguation]	0.0	2	1	1	1
5340	130	several factor behind this surprising finding	[several factors behind this surprising finding]	0.0	6	1	1	1
5341	130	event-based extractive summarization approach	[event-based extractive summarization approaches]	0.0	4	1	1	1
5342	130	case marker	[case markers]	0.0	2	1	1	1
5343	130	paraphrase sentence	[paraphrasing sentences]	0.0	2	1	1	1
5344	130	evaluation disambiguation algorithm	[evaluation disambiguation algorithms]	0.0	3	1	1	1
5345	130	University of California	[University of California]	0.0	3	1	1	1
5346	130	automate aspect	[automating aspects]	0.0	2	1	1	1
5347	130	space include a variety	[space including a variety]	0.0	4	1	1	1
5348	130	shortcoming of strategy	[shortcomings of strategies]	0.0	3	1	1	1
5349	130	proposition with fuzzy quantifier	[proposition with fuzzy quantifiers]	0.0	4	1	1	1
5350	130	,610 sentence	[,610 sentences]	0.0	2	1	1	1
5351	130	hand-built resource	[hand-built resources]	0.0	2	1	1	1
5352	130	independent event-based mutli-document summarization approach	[independent event-based mutli-document summarization approaches]	0.0	5	1	1	1
5353	130	inferential theory for language presupposition	[inferential theory for language presuppositions]	0.0	5	1	1	1
5354	130	view for reference	[view for reference]	0.0	3	1	1	1
5355	130	few noun	[few noun]	0.0	2	1	1	1
5356	130	now old resource management corpus	[now old Resource Management corpus]	0.0	5	1	1	1
5357	130	tdmt the dominant part	[TDMT the dominant part]	0.0	4	1	1	1
5358	130	information redundancy in input	[information redundancy in input]	0.0	4	1	1	1
5359	130	corpus a resource	[corpora a resource]	0.0	3	1	1	1
5360	130	tree-adjoining grammar TAG	[tree-adjoining grammars TAG]	0.0	3	1	1	1
5361	130	parallel corpus a available resource	[parallel corpora a available resource]	0.0	5	1	1	1
5362	130	primarily on information from machine-readable dictionary	[primarily on information from machine-readable dictionaries]	0.0	6	1	1	1
5363	130	map estimation	[MAP estimation]	0.0	2	1	1	1
5364	130	coordinate conjunction	[coordinate conjunction]	0.0	2	1	1	1
5365	130	reproducability	[reproducability]	0.0	1	1	1	1
5366	130	40 utterance	[40 utterances]	0.0	2	1	1	1
5367	130	least 99% sentence	[least 99% sentences]	0.0	3	1	1	1
5368	130	resource management corpus	[Resource Management corpus]	0.0	3	1	1	1
5369	130	top-down scope information of the structure	[top-down scope information of the structure]	0.0	6	1	1	1
5370	130	ne email .g	[NE e .g]	0.0	3	1	1	1
5371	130	intelligent system	[intelligent systems]	0.0	2	1	1	1
5372	130	validity	[validity]	0.0	1	2	2	1
5373	130	new technology	[new technology]	0.0	2	1	1	1
5374	130	precede	[preceding]	0.0	1	1	1	1
5375	130	use this tree-based model	[using these tree-based models]	0.0	4	1	1	1
5376	130	conventional SMT model	[conventional SMT models]	0.0	3	1	1	1
5377	130	effective way	[effective way]	0.0	2	1	1	1
5378	130	transfer System SimTran	[Transfer System SimTran]	0.0	3	1	1	1
5379	130	improvement in performance	[improvements in performance]	0.0	3	1	1	1
5380	130	ideal aim	[ideal aiming]	0.0	2	1	1	1
5381	130	more than 99% accuracy	[more than 99% accuracy]	0.0	4	1	1	1
5382	130	function word	[function words]	0.0	2	1	1	1
5383	130	knowledge level	[knowledge level]	0.0	2	1	1	1
5384	130	effect of polysemy	[effect of polysemy]	0.0	3	1	1	1
5385	130	aspect such as robust communication procedure	[aspects such as robust communication procedures]	0.0	6	1	1	1
5386	130	scheduling parser a recovery technique	[scheduling parser a recovery technique]	0.0	5	1	1	1
5387	130	primarily pass some issue	[primarily passing some issues]	0.0	4	1	1	1
5388	130	assign rank	[assigning ranks]	0.0	2	1	1	1
5389	130	include a variety	[including a variety]	0.0	3	1	1	1
5390	130	grammaticality sentence equivalence in meaning	[grammaticality sentences equivalence in meaning]	0.0	5	1	1	1
5391	130	various structure	[various structures]	0.0	2	1	1	1
5392	130	relative effect	[relative effects]	0.0	2	1	1	1
5393	130	person ne	[PERSON NE]	0.0	2	1	1	1
5394	130	separation margin	[separation margin]	0.0	2	1	1	1
5395	130	automatically identify argument structure	[automatically identifying argument structures]	0.0	4	1	1	1
5396	130	novel method add annotation	[novel method adding annotation]	0.0	4	1	1	1
5397	130	SI training	[SI training]	0.0	2	1	1	1
5398	130	result on addressee identification	[results on addressee identification]	0.0	4	1	1	1
5399	130	human translation	[human translations]	0.0	2	1	1	1
5400	130	light of various function	[light of various functions]	0.0	4	1	1	1
5401	130	generative character recognition model	[generative character recognition model]	0.0	4	1	1	1
5402	130	translation output	[translation output]	0.0	2	1	1	1
5403	130	robust method	[robust method]	0.0	2	1	1	1
5404	130	distinct phase	[distinct phases]	0.0	2	1	1	1
5405	130	human communication	[human communication]	0.0	2	1	1	1
5406	130	research topic	[research topics]	0.0	2	1	1	1
5407	130	pos tagger	[POS tagger]	0.0	2	1	1	1
5408	130	only edge with a interpretation	[only edges with a interpretation]	0.0	5	1	1	1
5409	130	mt cbmt	[MT CBMT]	0.0	2	1	1	1
5410	130	three method .e	[three methods .e]	0.0	3	1	1	1
5411	130	manual acquisition	[Manual acquisition]	0.0	2	1	1	1
5412	130	language with case marker	[language with case markers]	0.0	4	1	1	1
5413	130	pair of grammar	[pairs of grammars]	0.0	3	1	1	1
5414	130	subjective human judgment	[subjective human judgments]	0.0	3	1	1	1
5415	130	clear evidence	[clear evidence]	0.0	2	1	1	1
5416	130	such as size perspective-taking in reference	[such as size perspective-taking in reference]	0.0	6	1	1	1
5417	130	Bayesian Network classifier	[Bayesian Network classifiers]	0.0	3	1	1	1
5418	130	node with great	[node with greater]	0.0	3	1	1	1
5419	130	new tagger use	[new tagger use]	0.0	3	1	1	1
5420	130	.24% accuracy a error reduction	[.24% accuracy an error reduction]	0.0	5	1	1	1
5421	130	such as robust communication procedure	[such as robust communication procedures]	0.0	5	1	1	1
5422	130	spatial constraint	[spatial constraints]	0.0	2	1	1	1
5423	130	domain independent acoustic model	[domain independent acoustic models]	0.0	4	1	1	1
5424	130	human communication need	[human communication needs]	0.0	3	1	1	1
5425	130	combine the two style	[combining the two styles]	0.0	4	1	1	1
5426	130	grammaticality 99% sentence	[grammaticality 99% sentences]	0.0	3	1	1	1
5427	130	subcategorization frame SCF	[subcategorization frame SCF]	0.0	3	1	1	1
5428	130	robust efficient parsing a verb language	[Robust efficient parsing a verb language]	0.0	6	1	1	1
5429	130	retrieval of phrase	[retrieval of phrases]	0.0	3	1	1	1
5430	130	method construct word vector method	[methods constructing word vectors methods]	0.0	5	1	1	1
5431	130	high-accuracy word-level alignment model	[high-accuracy word-level alignment models]	0.0	4	1	1	1
5432	130	Hierarchical Acyclic Graph	[Hierarchical Acyclic Graph]	0.0	3	1	1	1
5433	130	generalization corresponding in a chart	[generalization corresponding in a chart]	0.0	5	1	1	1
5434	130	segment method in term	[segment methods in terms]	0.0	4	1	1	1
5435	130	likelihood method	[likelihood method]	0.0	2	1	1	1
5436	130	qualitative constraint from the text	[qualitative constraints from the text]	0.0	5	1	1	1
5437	130	use corpus a commonly available resource	[using corpora a commonly available resource]	0.0	6	1	1	1
5438	130	cover that gap	[covering that gap]	0.0	3	1	1	1
5439	130	k-means algorithm	[k-means algorithm]	0.0	2	1	1	1
5440	130	semantic frame encode the meaning	[semantic frame encoding the meaning]	0.0	5	1	1	1
5441	130	predict linguistic concept	[predicting linguistic concepts]	0.0	3	1	1	1
5442	130	encode the meaning	[encoding the meaning]	0.0	3	1	1	1
5443	130	case study avoid failure	[case study avoiding failures]	0.0	4	1	1	1
5444	130	new corpus exploration tool	[new corpus exploration tool]	0.0	4	1	1	1
5445	130	robust parsing of Korean language	[Robust parsing of Korean language]	0.0	5	1	1	1
5446	130	manual word alignment	[manual word alignments]	0.0	3	1	1	1
5447	130	certain class of algorithm	[certain classes of algorithms]	0.0	4	1	1	1
5448	130	approximation of the analysis	[approximation of the analysis]	0.0	4	1	1	1
5449	130	evaluate a chinese-to-english SMT model	[evaluating a Chinese-to-English SMT model]	0.0	5	1	1	1
5450	130	especially in the muc-3 evaluation	[especially in the MUC-3 evaluation]	0.0	5	1	1	1
5451	130	parser for English	[parser for English]	0.0	3	1	1	1
5452	130	manual acquisition of constraint	[Manual acquisition of constraints]	0.0	4	1	1	1
5453	130	notion of valency	[notions of valency]	0.0	3	1	1	1
5454	130	distribution of knowledge	[distribution of knowledge]	0.0	3	1	1	1
5455	130	produce phrase	[producing phrases]	0.0	2	1	1	1
5456	130	new way	[new way]	0.0	2	1	1	1
5457	130	encode the full meaning	[encoding the full meaning]	0.0	4	1	1	1
5458	130	novel Model	[novel Model]	0.0	2	1	1	1
5459	130	translate in another language	[translates in another language]	0.0	4	1	1	1
5460	130	multiple parsing strategy	[multiple parsing strategies]	0.0	3	1	1	1
5461	130	collection of utterance from site	[collection of utterances from sites]	0.0	5	1	1	1
5462	130	pattern-matching	[pattern-matching]	0.0	1	1	1	1
5463	130	output sentence	[output sentence]	0.0	2	1	1	1
5464	130	corpus consist of 126 sentence	[corpus consisting of 126 sentences]	0.0	5	1	1	1
5465	130	use a bilingual corpus	[using an bilingual corpus]	0.0	4	1	1	1
5466	130	identification of paraphrase	[identification of paraphrases]	0.0	3	1	1	1
5467	130	framework of Grammar	[framework of Grammars]	0.0	3	1	1	1
5468	130	i parsing a verb	[i parsing a verb]	0.0	4	1	1	1
5469	130	subcategorization	[subcategorization]	0.0	1	1	1	1
5470	130	build that editor	[building that editor]	0.0	3	1	1	1
5471	130	WSD accuracy	[WSD accuracy]	0.0	2	1	1	1
5472	130	missing word	[missing words]	0.0	2	1	1	1
5473	130	system performance	[system performance]	0.0	2	1	1	1
5474	130	technique make analysis	[techniques making analysis]	0.0	3	1	1	1
5475	130	advantage from the point	[advantages from the point]	0.0	4	1	1	1
5476	130	35 .0% improvement	[35 .0% improvement]	0.0	3	1	1	1
5477	130	build upon the notion of valency	[building upon the notions of valency]	0.0	6	1	1	1
5478	130	only the topmost	[only the topmost]	0.0	3	1	1	1
5479	130	96% correct paraphrase	[96% correct paraphrases]	0.0	3	1	1	1
5480	130	use only 40 utterance	[Using only 40 utterances]	0.0	4	1	1	1
5481	130	pair of context-free grammar	[pairs of context-free grammars]	0.0	4	1	1	1
5482	130	navy message	[Navy messages]	0.0	2	1	1	1
5483	130	human-ranked sentence plan	[human-ranked sentence plan]	0.0	3	1	1	1
5484	130	language system domain	[language system domains]	0.0	3	1	1	1
5485	130	language generation community	[language generation community]	0.0	3	1	1	1
5486	130	failure of reference	[failures of reference]	0.0	3	1	1	1
5487	130	hardness result for the class NP	[hardness results for the class NP]	0.0	6	1	1	1
5488	130	word-sense of word	[word-senses of words]	0.0	3	1	1	1
5489	130	Multi-site	[Multi-site]	0.0	1	1	1	1
5490	130	follow speech recognition application	[following speech recognition applications]	0.0	4	1	1	1
5491	130	reference about the discourse	[reference about the discourse]	0.0	4	1	1	1
5492	130	dialogue duration for user	[dialogue duration for users]	0.0	4	1	1	1
5493	130	dimension skill level	[dimensions skill level]	0.0	3	1	1	1
5494	130	example similar	[examples similar]	0.0	2	1	1	1
5495	130	include a frequent word	[including a frequent word]	0.0	4	1	1	1
5496	130	key factor include limitation	[key factors including limitations]	0.0	4	1	1	1
5497	130	three technique make analysis	[three techniques making analysis]	0.0	4	1	1	1
5498	130	grammar work	[grammars working]	0.0	2	1	1	1
5499	130	scene	[scene]	0.0	1	1	1	1
5500	130	programming language	[programming language]	0.0	2	1	1	1
5501	130	head-orientation	[head-orientation]	0.0	1	1	1	1
5502	130	use of parsing strategy	[use of parsing strategies]	0.0	4	1	1	1
5503	130	detect speech repair	[detecting speech repairs]	0.0	3	1	1	1
5504	130	grounding act	[grounding acts]	0.0	2	1	1	1
5505	130	97 .24% accuracy	[97 .24% accuracy]	0.0	3	1	1	1
5506	130	training factor	[training factors]	0.0	2	1	1	1
5507	130	much simple set	[much simpler set]	0.0	3	1	1	1
5508	130	standard model	[standard models]	0.0	2	1	1	1
5509	130	nlp technique	[NLP techniques]	0.0	2	2	2	1
5510	130	13% decrease	[13% decrease]	0.0	2	1	1	1
5511	130	many classical formalism	[many classical formalisms]	0.0	3	1	1	1
5512	130	lexical cross-relation	[lexical cross-relations]	0.0	2	1	1	1
5513	130	access	[access]	0.0	1	2	2	1
5514	130	three word from high-accuracy alignment model	[three words from high-accuracy alignment models]	0.0	6	1	1	1
5515	130	bootstrapping approach	[bootstrapping approach]	0.0	2	1	1	1
5516	130	possible span	[possible span]	0.0	2	1	1	1
5517	130	information from machine-readable dictionary MRD's	[information from machine-readable dictionaries MRD's]	0.0	5	1	1	1
5518	130	dependency pointer	[dependency pointers]	0.0	2	1	1	1
5519	130	substitution for this element	[substitutions for these elements]	0.0	4	1	1	1
5520	130	ensemble method	[ensemble methods]	0.0	2	1	1	1
5521	130	effect use a out-of-domain corpus	[effect using an out-of-domain corpus]	0.0	5	1	1	1
5522	130	reflect the social relationship social rank	[reflecting the social relationship social ranks]	0.0	6	1	1	1
5523	130	stochastic output selection	[stochastic output selection]	0.0	3	1	1	1
5524	130	include conditioning	[including conditioning]	0.0	2	1	1	1
5525	130	appropriate pint	[appropriate pints]	0.0	2	1	1	1
5526	130	evaluation measure for machine translation	[evaluation measures for machine translation]	0.0	5	1	1	1
5527	130	length-based criterion	[length-based criterion]	0.0	2	1	1	1
5528	130	domain tutorial dialogue	[domains tutorial dialogue]	0.0	3	1	1	1
5529	130	two disambiguation system	[two disambiguation systems]	0.0	3	1	1	1
5530	130	COMPLEX a lexicon	[COMPLEX a lexicon]	0.0	3	1	1	1
5531	130	instantiation of the sense	[instantiation of the sense]	0.0	4	1	1	1
5532	130	Machine Translation SMT	[Machine Translation SMT]	0.0	3	1	1	1
5533	130	harvesting bitext from the web	[harvesting bitexts from the Web]	0.0	5	1	1	1
5534	130	explore English-Chinese bitext	[exploring English-Chinese bitexts]	0.0	3	1	1	1
5535	130	89 .75% f-measure a decrease	[89 .75% F-measure a decrease]	0.0	5	1	1	1
5536	130	engine machine	[engine machine]	0.0	2	1	1	1
5537	130	strength for each	[strength for each]	0.0	3	1	1	1
5538	130	for scoring set	[for scoring sets]	0.0	3	1	1	1
5539	130	generative model of language	[generative model of language]	0.0	4	1	1	1
5540	130	i efficient parsing final language	[i efficient parsing final language]	0.0	5	1	1	1
5541	130	valid interpretation	[valid interpretation]	0.0	2	1	1	1
5542	130	rule-reduction	[rule-reduction]	0.0	1	1	1	1
5543	130	degree	[degree]	0.0	1	3	3	1
5544	130	3 indicator for the prediction task	[3 indicators for the prediction task]	0.0	6	1	1	1
5545	130	progress from generation into the output	[progressing from generation into the output]	0.0	6	1	1	1
5546	130	most state-of-the-art evaluation measure	[Most state-of-the-art evaluation measures]	0.0	4	1	1	1
5547	130	operational semantics of language application	[operational semantics of language applications]	0.0	5	1	1	1
5548	130	question-answering	[question-answering]	0.0	1	1	1	1
5549	130	experience plan COMPLEX	[experience planning COMPLEX]	0.0	3	1	1	1
5550	130	problem combine several language model	[problem combining several language models]	0.0	5	1	1	1
5551	130	formalism for the combination	[formalism for the combination]	0.0	4	1	1	1
5552	130	FlexP a pattern-matching parser	[FlexP a pattern-matching parser]	0.0	4	1	1	1
5553	130	instead of pair	[instead of pairs]	0.0	3	1	1	1
5554	130	robust efficient parsing final language	[Robust efficient parsing final language]	0.0	5	1	1	1
5555	130	robust efficient parsing of Korean language	[Robust efficient parsing of Korean language]	0.0	6	1	1	1
5556	130	attribute of the entity	[attributes of the entities]	0.0	4	1	1	1
5557	130	learn morphology	[learning morphology]	0.0	2	1	1	1
5558	130	virtue	[virtue]	0.0	1	1	1	1
5559	130	miscommunication reference	[miscommunication reference]	0.0	2	1	1	1
5560	130	amount slightly superior	[amount slightly superior]	0.0	3	1	1	1
5561	130	Horn logic	[Horn logic]	0.0	2	1	1	1
5562	130	such discourse phenomenon	[such discourse phenomena]	0.0	3	1	1	1
5563	130	distinct dataset	[distinct datasets]	0.0	2	1	1	1
5564	130	hypothesis corresponding to the LM	[hypothesis corresponding to the LM]	0.0	5	1	1	1
5565	130	iii system development via acquisition	[iii system development via acquisition]	0.0	5	1	1	1
5566	130	class of word	[class of words]	0.0	3	1	1	1
5567	130	build upon the notion	[building upon the notions]	0.0	4	1	1	1
5568	130	amount of lexical variation	[amount of lexical variation]	0.0	4	1	1	1
5569	130	empirical result cast doubt	[empirical results casting doubt]	0.0	4	1	1	1
5570	130	overhead	[overheads]	0.0	1	1	1	1
5571	130	i parsing language	[i parsing language]	0.0	3	1	1	1
5572	130	segment boundary in multiparty dialogue	[segment boundaries in multiparty dialogue]	0.0	5	1	1	1
5573	130	parsing language with overt case marker	[parsing language with overt case markers]	0.0	6	1	1	1
5574	130	evaluation result involve extraction	[evaluation results involving extraction]	0.0	4	1	1	1
5575	130	cross-relation	[cross-relations]	0.0	1	1	1	1
5576	130	include limitation of current mt architecture	[including limitations of current MT architectures]	0.0	6	1	1	1
5577	130	novel datum structure	[novel data structure]	0.0	3	1	1	1
5578	130	tree adjoin Grammar	[Tree Adjoining Grammars]	0.0	3	1	1	1
5579	130	test of a common claim	[test of an common claim]	0.0	5	1	1	1
5580	130	instantiation of clause in a program	[instantiation of clauses in a program]	0.0	6	1	1	1
5581	130	analysis for a class	[analysis for a class]	0.0	4	1	1	1
5582	130	important issue in language processing	[important issue in Language Processing]	0.0	5	1	1	1
5583	130	language family	[language families]	0.0	2	1	1	1
5584	130	produce such phrase	[producing such phrases]	0.0	3	1	1	1
5585	130	approach to computer-assisted academic writing	[approach to computer-assisted academic writing]	0.0	5	1	1	1
5586	130	exploitation of limited domain semantics	[exploitation of limited domain semantics]	0.0	5	1	1	1
5587	130	n-gram model	[n-gram models]	0.0	2	1	1	1
5588	130	polysemic verb	[polysemic verbs]	0.0	2	1	1	1
5589	130	work harvesting bitext from the web	[work harvesting bitexts from the Web]	0.0	6	1	1	1
5590	130	site	[sites]	0.0	1	1	1	1
5591	130	implication automate certain aspect	[Implications automating certain aspects]	0.0	4	1	1	1
5592	130	use a statistical pos tagger	[using a statistical POS tagger]	0.0	5	1	1	1
5593	130	workability of the mechanism	[workability of the mechanism]	0.0	4	1	1	1
5594	130	traditional information retrieval	[Traditional information retrieval]	0.0	3	1	1	1
5595	130	old resource management corpus	[old Resource Management corpus]	0.0	4	1	1	1
5596	130	method method	[methods methods]	0.0	2	1	1	1
5597	130	level of chunk	[levels of chunks]	0.0	3	1	1	1
5598	130	output text	[outputting texts]	0.0	2	1	1	1
5599	130	novel bootstrapping approach to Entity	[novel bootstrapping approach to Entity]	0.0	5	1	1	1
5600	130	generalization of chart parsing instantiation	[generalization of chart Parsing instantiation]	0.0	5	1	1	1
5601	130	problem for reader such as word	[problems for readers such as words]	0.0	6	1	1	1
5602	130	traditional generative approach	[traditional generative approach]	0.0	3	1	1	1
5603	130	partial instantiation of clause	[partial instantiation of clauses]	0.0	4	1	1	1
5604	130	use parallel corpus a resource	[using parallel corpora a resource]	0.0	5	1	1	1
5605	130	heuristically-produced sentence	[heuristically-produced sentences]	0.0	2	1	1	1
5606	130	prediction task	[prediction task]	0.0	2	1	1	1
5607	130	Human-Machine Communication	[Human-Machine Communication]	0.0	2	1	1	1
5608	130	generative probabilistic model	[generative probabilistic model]	0.0	3	1	1	1
5609	130	describe the processing	[describing the processing]	0.0	3	1	1	1
5610	130	word-semantic information email	[word-semantic information e]	0.0	3	1	1	1
5611	130	integrate q/a application into real-world environment	[integrating Q/A applications into real-world environments]	0.0	6	1	1	1
5612	130	collect statistic on co-occurrence pattern	[collecting statistics on co-occurrence patterns]	0.0	5	1	1	1
5613	130	study of expression	[study of expressions]	0.0	3	1	1	1
5614	130	person	[PERSON]	0.0	1	1	1	1
5615	130	new theory	[new theory]	0.0	2	1	1	1
5616	130	great effectiveness	[great effectiveness]	0.0	2	1	1	1
5617	130	word-trigram recognition require transcription	[word-trigram recognition requiring transcription]	0.0	4	1	1	1
5618	130	spatial attribute	[spatial attributes]	0.0	2	1	1	1
5619	130	apply evaluation technique to the output	[applying evaluation techniques to the output]	0.0	6	1	1	1
5620	130	global behavior	[global behavior]	0.0	2	1	1	1
5621	130	robust recognition of input	[robust recognition of input]	0.0	4	1	1	1
5622	130	distance constraint	[distance constraints]	0.0	2	1	1	1
5623	130	method discover sentence in corpus	[method discovering sentences in corpora]	0.0	5	1	1	1
5624	130	identify argument structure	[identifying argument structures]	0.0	3	1	1	1
5625	130	correct error	[correcting errors]	0.0	2	1	1	1
5626	130	model analyze sentence	[model analyzing sentences]	0.0	3	1	1	1
5627	130	predictor	[predictors]	0.0	1	1	1	1
5628	130	part of a study of expression	[part of a study of expressions]	0.0	6	1	1	1
5629	130	state-of-the-art word sense	[state-of-the-art word sense]	0.0	3	1	1	1
5630	130	style perform inference	[styles performing inference]	0.0	3	1	1	1
5631	130	parsing capability	[parsing capability]	0.0	2	1	1	1
5632	130	learning phrase three word	[learning phrases three words]	0.0	4	1	1	1
5633	130	model for stochastic output selection	[model for stochastic output selection]	0.0	5	1	1	1
5634	130	more datum relevant	[more data relevant]	0.0	3	1	1	1
5635	130	orientation	[orientation]	0.0	1	1	1	1
5636	130	standard mt evaluation method (BLEU	[standard MT evaluation methods (BLEU]	0.0	5	1	1	1
5637	130	/soames	[/Soames]	0.0	1	1	1	1
5638	130	light of rhetorical function	[light of rhetorical functions]	0.0	4	1	1	1
5639	130	transcription error	[transcription errors]	0.0	2	1	1	1
5640	130	disjunctive form	[disjunctive form]	0.0	2	1	1	1
5641	130	interaction with knowledge about dialog	[interaction with knowledge about dialogs]	0.0	5	1	1	1
5642	130	class of miscommunication reference problem	[class of miscommunication reference problems]	0.0	5	1	1	1
5643	130	speed-up element in the design	[speed-up element in the design]	0.0	5	1	1	1
5644	130	acquire stem from a word	[acquiring stems from a word]	0.0	5	1	1	1
5645	130	corpus of .284 srh	[corpus of .284 SRHs]	0.0	4	1	1	1
5646	130	relatively simple means	[relatively simple means]	0.0	3	1	1	1
5647	130	approach to wsi	[approaches to WSI]	0.0	3	1	1	1
5648	130	model of machine translation mt	[model of machine translation MT]	0.0	5	1	1	1
5649	130	Grammar for Machine Translation	[Grammars for Machine Translation]	0.0	4	1	1	1
5650	130	system development	[system development]	0.0	2	1	1	1
5651	130	computational algorithm for unification	[computational algorithms for unification]	0.0	4	1	1	1
5652	130	build a non-English stemmer	[building a non-English stemmer]	0.0	4	1	1	1
5653	130	computational issue	[computational issues]	0.0	2	1	1	1
5654	130	pass some computational issue	[passing some computational issues]	0.0	4	1	1	1
5655	130	ongoing construction	[ongoing construction]	0.0	2	1	1	1
5656	130	robust efficient parsing	[Robust efficient parsing]	0.0	3	1	1	1
5657	130	little gain from information	[little gain from information]	0.0	4	1	1	1
5658	130	size	[size]	0.0	1	1	1	1
5659	130	i robust efficient parsing final language	[i Robust efficient parsing final language]	0.0	6	1	1	1
5660	130	several factor behind this finding	[several factors behind this finding]	0.0	5	1	1	1
5661	130	process example	[processing examples]	0.0	2	1	1	1
5662	130	wide coverage	[wide coverage]	0.0	2	1	1	1
5663	130	Hierarchical Graph HDAG	[Hierarchical Graph HDAG]	0.0	3	1	1	1
5664	130	method of traditional language interface	[methods of traditional language interfaces]	0.0	5	1	1	1
5665	130	distribution of behavior	[distribution of behaviors]	0.0	3	1	1	1
5666	130	generic formalism	[generic formalism]	0.0	2	1	1	1
5667	130	description in Japanese	[descriptions in Japanese]	0.0	3	1	1	1
5668	130	combinatorics	[combinatorics]	0.0	1	1	1	1
5669	130	real application	[real applications]	0.0	2	1	1	1
5670	130	New result on all four application	[New results on all four applications]	0.0	6	1	1	1
5671	130	corresponding to domain	[corresponding to domains]	0.0	3	1	1	1
5672	130	harvesting bitext	[harvesting bitexts]	0.0	2	1	1	1
5673	130	employ n-gram model for Thai prediction	[employing n-gram models for Thai prediction]	0.0	6	1	1	1
5674	130	definition of grammar	[definition of grammars]	0.0	3	1	1	1
5675	130	use of various feature	[use of various features]	0.0	4	1	1	1
5676	130	problem attribute such as size	[problems attributes such as size]	0.0	5	1	1	1
5677	130	.1% a reduction	[.1% a reduction]	0.0	3	1	1	1
5678	130	small number	[small number]	0.0	2	1	1	1
5679	130	explicit information	[explicit information]	0.0	2	1	1	1
5680	130	transition graph for a special type	[transition graphs for a special type]	0.0	6	1	1	1
5681	130	natural language application	[natural language applications]	0.0	3	1	1	1
5682	130	analysis of answer to question	[analysis of answers to questions]	0.0	5	1	1	1
5683	130	making	[making]	0.0	1	1	1	1
5684	130	score alternative speech recognition hypothese	[scoring alternative speech recognition hypotheses]	0.0	5	1	1	1
5685	130	question-answering system	[question-answering system]	0.0	2	1	1	1
5686	130	skill level to the system	[skill level to the system]	0.0	5	1	1	1
5687	130	informational goal	[informational goals]	0.0	2	1	1	1
5688	130	stochastic transducer	[stochastic transducer]	0.0	2	1	1	1
5689	130	computational theory of human reasoning	[computational theory of human reasoning]	0.0	5	1	1	1
5690	130	word-trigram recognition	[word-trigram recognition]	0.0	2	1	1	1
5691	130	lexical approach alone	[lexical approach alone]	0.0	3	1	1	1
5692	130	such as the dominance distribution	[such as the dominance distribution]	0.0	5	1	1	1
5693	130	corner parsing	[corner parsing]	0.0	2	1	1	1
5694	130	common speculative claim	[common speculative claim]	0.0	3	1	1	1
5695	130	use corpus	[using corpora]	0.0	2	1	1	1
5696	130	strategy one utilize knowledge-based mechanism	[strategies one utilizing knowledge-based mechanisms]	0.0	5	1	1	1
5697	130	exact match accuracy	[exact match accuracy]	0.0	3	1	1	1
5698	130	topic shift	[topic shifts]	0.0	2	1	1	1
5699	130	social relationship social rank	[social relationship social ranks]	0.0	4	1	1	1
5700	130	lexical distribution of grammatical knowledge	[lexical distribution of grammatical knowledge]	0.0	5	1	1	1
5701	130	subcategorization frame	[subcategorization frame]	0.0	2	1	1	1
5702	130	i robust parsing a verb	[i Robust parsing a verb]	0.0	5	1	1	1
5703	130	cluster subcategorization frame distribution	[clustering subcategorization frame distributions]	0.0	4	1	1	1
5704	130	simultaneously use memory	[simultaneously using memory]	0.0	3	1	1	1
5705	130	broad-coverage parser	[broad-coverage parser]	0.0	2	1	1	1
5706	130	inevitable inference	[inevitable inferences]	0.0	2	1	1	1
5707	130	ongoing construction as reliable basis	[ongoing construction as reliable basis]	0.0	5	1	1	1
5708	130	free text	[free text]	0.0	2	1	1	1
5709	130	language in natural setting	[language in natural settings]	0.0	4	1	1	1
5710	130	use the SI corpus	[using the SI corpus]	0.0	4	1	1	1
5711	130	quasi-destructive scheme' ability	[quasi-destructive scheme's ability]	0.0	3	1	1	1
5712	130	new word-sense system	[new word-sense systems]	0.0	3	1	1	1
5713	130	small-sized corpus	[small-sized corpus]	0.0	2	1	1	1
5714	130	grammar formalism Phonology structure Grammar	[grammar formalism Phonology Structure Grammar]	0.0	5	1	1	1
5715	130	formal analysis for a class	[formal analysis for a class]	0.0	5	1	1	1
5716	130	other area of language processing	[other areas of language processing]	0.0	5	1	1	1
5717	130	more frequently good predictor	[more frequently good predictors]	0.0	4	1	1	1
5718	130	discourse constraint	[discourse constraint]	0.0	2	1	1	1
5719	130	processing time	[processing time]	0.0	2	1	1	1
5720	130	personnel	[personnel]	0.0	1	1	1	1
5721	130	sequence of morpheme	[sequence of morphemes]	0.0	3	1	1	1
5722	130	particular format for meaning	[particular format for meaning]	0.0	4	1	1	1
5723	130	improvement of 22-38% in precision	[improvement of 22-38% in precision]	0.0	5	1	1	1
5724	130	coverage than account	[coverage than accounts]	0.0	3	1	1	1
5725	130	large inflow of datum	[large inflow of data]	0.0	4	1	1	1
5726	130	distribution of form	[distribution of forms]	0.0	3	1	1	1
5727	130	bilingual corpus a more available resource	[bilingual corpora a more available resource]	0.0	6	1	1	1
5728	130	treelet	[treelet]	0.0	1	1	1	1
5729	130	certain natural sublanguage	[certain natural sublanguages]	0.0	3	1	1	1
5730	130	/Gazdar	[/Gazdar]	0.0	1	1	1	1
5731	130	graphic display	[graphics displays]	0.0	2	1	1	1
5732	130	field exercise	[field exercises]	0.0	2	1	1	1
5733	130	statistical class grammar	[statistical class grammar]	0.0	3	1	1	1
5734	130	only the topmost of the edge	[only the topmost of the edges]	0.0	6	1	1	1
5735	130	two decade	[two decades]	0.0	2	1	1	1
5736	130	markov model	[Markov models]	0.0	2	1	1	1
5737	130	lexical distribution of knowledge	[lexical distribution of knowledge]	0.0	4	1	1	1
5738	130	complex mixture	[complex mixtures]	0.0	2	1	1	1
5739	130	information from dictionary MRD's	[information from dictionaries MRD's]	0.0	4	1	1	1
5740	130	estimation of continuous markov model	[estimation of continuous Markov models]	0.0	5	1	1	1
5741	130	mathematical formalism for the combination	[mathematical formalism for the combination]	0.0	5	1	1	1
5742	130	word with meaning ambiguity	[words with meanings ambiguity]	0.0	4	1	1	1
5743	130	little gain	[little gain]	0.0	2	1	1	1
5744	130	syntactic analysis robust---an scheduling parser	[syntactic analysis robust---an scheduling parser]	0.0	5	1	1	1
5745	130	efficient parsing of Korean language	[efficient parsing of Korean language]	0.0	5	1	1	1
5746	130	broad use	[broad use]	0.0	2	1	1	1
5747	130	use of feature include jointly conditioning	[use of features including jointly conditioning]	0.0	6	1	1	1
5748	130	inherently a conglomeration of lexical rule	[inherently a conglomeration of lexical rules]	0.0	6	1	1	1
5749	130	subject of recent debate	[subject of recent debate]	0.0	4	1	1	1
5750	130	U .s	[U .S]	0.0	2	1	1	1
5751	130	fundamentally different strategy one utilize mechanism	[fundamentally different strategies one utilizing mechanisms]	0.0	6	1	1	1
5752	130	non-deterministic parsing choice	[non-deterministic parsing choices]	0.0	3	1	1	1
5753	130	several advantage	[Several advantages]	0.0	2	1	1	1
5754	130	analysis of the coordinate sentence	[analysis of the coordinate sentences]	0.0	5	1	1	1
5755	130	knowledge concern translation equivalence	[knowledge concerning translation equivalence]	0.0	4	1	1	1
5756	130	linguistic database	[linguistic databases]	0.0	2	1	1	1
5757	130	tape-recorded description	[tape-recorded descriptions]	0.0	2	1	1	1
5758	130	use a model	[using a model]	0.0	3	1	1	1
5759	130	show the wsd accuracy low	[showing the WSD accuracy lower]	0.0	5	1	1	1
5760	130	decision use the reference	[decisions using the reference]	0.0	4	1	1	1
5761	130	pick the hypothesis	[picking the hypothesis]	0.0	3	1	1	1
5762	130	translation task	[translation task]	0.0	2	1	1	1
5763	130	correct Japanese homophone error	[correcting Japanese homophone errors]	0.0	4	1	1	1
5764	130	experiment with substitution cost	[experiments with substitution costs]	0.0	4	1	1	1
5765	130	decrease in f-measure error	[decrease in F-measure error]	0.0	4	1	1	1
5766	130	.5% word error rate	[.5% word error rate]	0.0	4	1	1	1
5767	130	acquire stem from a word corpus	[acquiring stems from a word corpus]	0.0	6	1	1	1
5768	130	depend on the type	[depending on the type]	0.0	4	1	1	1
5769	130	two word-sense system one	[two word-sense systems one]	0.0	4	1	1	1
5770	130	binary calculus notation	[binary calculus notation]	0.0	3	1	1	1
5771	130	well exist parsing model	[best existing parsing models]	0.0	4	1	1	1
5772	130	possibility	[possibility]	0.0	1	1	1	1
5773	130	elicit subjective judgment	[eliciting subjective judgments]	0.0	3	1	1	1
5774	130	context of dialog system	[context of dialog systems]	0.0	4	1	1	1
5775	130	robust probabilistic parsing model	[robust probabilistic parsing models]	0.0	4	1	1	1
5776	130	chart parsing partial instantiation	[chart Parsing partial instantiation]	0.0	4	1	1	1
5777	130	statistical mt architecture	[statistical MT architectures]	0.0	3	1	1	1
5778	130	term weighting assignment paradigm	[term weighting assignment paradigm]	0.0	4	1	1	1
5779	130	method of language interface	[methods of language interfaces]	0.0	4	1	1	1
5780	130	concern translation equivalence of expression	[concerning translation equivalence of expressions]	0.0	5	1	1	1
5781	130	event detection	[event detection]	0.0	2	2	2	1
5782	130	acquire datum	[acquiring data]	0.0	2	1	1	1
5783	130	learning of phrase translation from alignment	[learning of phrase translations from alignments]	0.0	6	1	1	1
5784	130	probabilistic grammar work	[probabilistic grammars working]	0.0	3	1	1	1
5785	130	parallel corpus a commonly available resource	[parallel corpora a commonly available resource]	0.0	6	1	1	1
5786	130	voting-	[voting-]	0.0	1	1	1	1
5787	130	simple explanation for the nature	[simple explanation for the nature]	0.0	5	1	1	1
5788	130	certain sublanguage	[certain sublanguages]	0.0	2	1	1	1
5789	130	identify subtopic boundary	[identifying subtopic boundaries]	0.0	3	1	1	1
5790	130	word sense disambiguation performance	[word sense disambiguation performance]	0.0	4	1	1	1
5791	130	construction as reliable basis	[construction as reliable basis]	0.0	4	1	1	1
5792	130	methodology train spot	[methodology training SPoT]	0.0	3	1	1	1
5793	130	symmetric pattern	[symmetric patterns]	0.0	2	1	1	1
5794	130	pass some issue	[passing some issues]	0.0	3	1	1	1
5795	130	problem of word sense	[problem of word sense]	0.0	4	1	1	1
5796	130	dialogue act	[dialogue act]	0.0	2	1	1	1
5797	130	mle reestimation	[MLE reestimation]	0.0	2	1	1	1
5798	130	validity of database	[validity of databases]	0.0	3	1	1	1
5799	130	explanation for the presuppositional nature	[explanation for the presuppositional nature]	0.0	5	1	1	1
5800	130	use in such case-based mt	[use in such case-based MT]	0.0	5	1	1	1
5801	130	strongly biase parsing	[strongly biasing parsing]	0.0	3	1	1	1
5802	130	surface representation	[surface representation]	0.0	2	1	1	1
5803	130	amount of variation	[amount of variation]	0.0	3	1	1	1
5804	130	Dynamic Phrasal Lexicon [Zernik88]	[Dynamic Phrasal Lexicon [Zernik88]]	0.0	4	1	1	1
5805	130	10 .4m word	[10 .4M words]	0.0	3	1	1	1
5806	130	smta check	[SmTA check]	0.0	2	1	1	1
5807	130	advantage like reproducability	[advantages like reproducability]	0.0	3	1	1	1
5808	130	factor include inherent limitation	[factors including inherent limitations]	0.0	4	1	1	1
5809	130	exploit a large corpus	[exploiting a large corpus]	0.0	4	1	1	1
5810	130	translation use	[translation use]	0.0	2	1	1	1
5811	130	training module	[training modules]	0.0	2	1	1	1
5812	130	complex linguistic database	[complex linguistic databases]	0.0	3	1	1	1
5813	130	test of a claim	[test of an claim]	0.0	4	1	1	1
5814	130	elementary speech act	[elementary speech acts]	0.0	3	1	1	1
5815	130	phrase structure parsing	[phrase structure parsing]	0.0	3	1	1	1
5816	130	independent mutli-document summarization approach	[independent mutli-document summarization approaches]	0.0	4	1	1	1
5817	130	theory of human plausible reasoning	[theory of human plausible reasoning]	0.0	5	1	1	1
5818	130	ACE corpus	[ACE corpora]	0.0	2	1	1	1
5819	130	necessary condition	[necessary conditions]	0.0	2	1	1	1
5820	130	addition the language model	[addition the language model]	0.0	4	1	1	1
5821	130	dependency insertion grammar	[dependency insertion grammar]	0.0	3	2	2	1
5822	130	open-domain question	[open-domain question]	0.0	2	1	1	1
5823	130	question-answering q/a	[question-answering Q/A]	0.0	2	1	1	1
5824	130	fashion	[fashion]	0.0	1	1	1	1
5825	130	corpus of srh	[corpus of SRHs]	0.0	3	1	1	1
5826	130	two learner	[two learners]	0.0	2	1	1	1
5827	130	sentence corpus as training resource	[sentences corpus as training resources]	0.0	5	1	1	1
5828	130	strong discourse effect	[strong discourse effect]	0.0	3	1	1	1
5829	130	co-occurrence pattern	[co-occurrence patterns]	0.0	2	1	1	1
5830	130	learning methodology	[learning methodology]	0.0	2	1	1	1
5831	130	approach alone	[approach alone]	0.0	2	1	1	1
5832	130	complicated knowledgebase	[complicated knowledgebases]	0.0	2	1	1	1
5833	130	separation	[separation]	0.0	1	1	1	1
5834	130	automatically gather a large number	[automatically gathering a large number]	0.0	5	1	1	1
5835	130	acquisition the construction	[acquisition the construction]	0.0	3	1	1	1
5836	130	simple technique learn morphology	[simple technique learning morphology]	0.0	4	1	1	1
5837	130	parsing accuracy	[parsing accuracy]	0.0	2	1	1	1
5838	130	minor extension	[minor extensions]	0.0	2	1	1	1
5839	130	top sentence plan	[top sentence plan]	0.0	3	1	1	1
5840	130	empirical test of a claim	[empirical test of an claim]	0.0	5	1	1	1
5841	130	word segmentation system	[word segmentation system]	0.0	3	1	1	1
5842	130	part-of-speech tagger use	[part-of-speech tagger use]	0.0	3	1	1	1
5843	130	large newspaper corpus	[large newspaper corpora]	0.0	3	1	1	1
5844	130	motivation	[motivation]	0.0	1	1	1	1
5845	130	independent extractive mutli-document summarization approach	[independent extractive mutli-document summarization approaches]	0.0	5	1	1	1
5846	130	primarily on explicit information	[primarily on explicit information]	0.0	4	1	1	1
5847	130	build translation system	[building translation systems]	0.0	3	1	1	1
5848	130	apply general-purpose language	[applying general-purpose language]	0.0	3	1	1	1
5849	130	computer text generation	[computer text generation]	0.0	3	1	1	1
5850	130	rhetorical function	[rhetorical functions]	0.0	2	1	1	1
5851	130	several language model	[several language models]	0.0	3	1	1	1
5852	130	f-measure a 13% relative decrease	[F-measure a 13% relative decrease]	0.0	5	1	1	1
5853	130	text through transformation	[text through transformation]	0.0	3	1	1	1
5854	130	evaluate a representative SMT model	[evaluating a representative SMT model]	0.0	5	1	1	1
5855	130	few equivalent reading	[fewer equivalent readings]	0.0	3	1	1	1
5856	130	corpus of English translation	[corpus of English translations]	0.0	4	1	1	1
5857	130	though in the context	[though in the context]	0.0	4	1	1	1
5858	130	resource logic	[resource logic]	0.0	2	1	1	1
5859	130	pair	[pairs]	0.0	1	1	1	1
5860	130	first test	[first test]	0.0	2	1	1	1
5861	130	sets( dop	[sets( DoPS]	0.0	2	1	1	1
5862	130	distinct task	[distinct tasks]	0.0	2	1	1	1
5863	130	analysis of move structure	[analysis of move structures]	0.0	4	1	1	1
5864	130	thereby reduce the amount	[thereby reducing the amount]	0.0	4	1	1	1
5865	130	bunsetsu	[bunsetsu]	0.0	1	2	2	1
5866	130	also a generalization roughly corresponding	[also a generalization roughly corresponding]	0.0	5	1	1	1
5867	130	right translation	[right translation]	0.0	2	1	1	1
5868	130	property of noun phrase	[properties of noun phrases]	0.0	4	1	1	1
5869	130	Network	[Network]	0.0	1	1	1	1
5870	130	factor behind this surprising finding	[factors behind this surprising finding]	0.0	5	1	1	1
5871	130	10k sentence corpus	[10K sentences corpus]	0.0	3	1	1	1
5872	130	translate a sentence	[translating a sentence]	0.0	3	1	1	1
5873	130	automatically training module	[automatically training modules]	0.0	3	1	1	1
5874	130	control the order	[controlling the order]	0.0	3	1	1	1
5875	130	multimedia articulation	[multimedia articulation]	0.0	2	1	1	1
5876	130	corpus a commonly available resource	[corpora a commonly available resource]	0.0	5	1	1	1
5877	130	time solution for any	[time solution for any]	0.0	4	1	1	1
5878	130	consist of ,610 sentence	[consisting of ,610 sentences]	0.0	4	1	1	1
5879	130	integrate automatic q/a application	[integrating automatic Q/A applications]	0.0	4	1	1	1
5880	130	adequate explanation	[adequate explanation]	0.0	2	1	1	1
5881	130	Phrasal Lexicon DHPL [Zernik88]	[Phrasal Lexicon DHPL [Zernik88]]	0.0	4	1	1	1
5882	130	conversational vehicle	[conversational vehicle]	0.0	2	1	1	1
5883	130	Bayesian Network	[Bayesian Network]	0.0	2	1	1	1
5884	130	automatically acquire stem	[automatically acquiring stems]	0.0	3	1	1	1
5885	130	use information retrieval	[using information retrieval]	0.0	3	1	1	1
5886	130	role system	[Role System]	0.0	2	1	1	1
5887	130	use a pos tagger in conjunction	[using a POS tagger in conjunction]	0.0	6	1	1	1
5888	130	experimental result on all four application	[experimental results on all four applications]	0.0	6	1	1	1
5889	130	three minute	[three minutes]	0.0	2	1	1	1
5890	130	two domain tutorial dialogue	[two domains tutorial dialogue]	0.0	4	1	1	1
5891	130	speed-up element	[speed-up element]	0.0	2	1	1	1
5892	130	understand system with mobile agent	[understanding system with mobile agents]	0.0	5	1	1	1
5893	130	include limitation of statistical mt architecture	[including limitations of statistical MT architectures]	0.0	6	1	1	1
5894	130	pint in mt	[pints in MT]	0.0	3	1	1	1
5895	130	cue such as cue phrase	[cues such as cue phrases]	0.0	5	1	1	1
5896	130	LHIP	[LHIP]	0.0	1	1	1	1
5897	130	new paradigm	[new paradigm]	0.0	2	2	2	1
5898	130	alternative—in term of simplicity	[alternative—in terms of simplicity]	0.0	4	1	1	1
5899	130	message-passing infrastructure	[message-passing infrastructure]	0.0	2	1	1	1
5900	130	construct a computational phonological system	[constructing a computational phonological system]	0.0	5	1	1	1
5901	130	linguistic parser/generator a transfer component	[linguistic parser/generator a transfer component]	0.0	5	1	1	1
5902	130	use a logical model	[using a logical model]	0.0	4	1	1	1
5903	130	theory for language presupposition	[theory for language presuppositions]	0.0	4	1	1	1
5904	130	add annotation to corpus	[adding annotation to corpora]	0.0	4	1	1	1
5905	130	model of the scene	[model of the scene]	0.0	4	1	1	1
5906	130	baseline result	[baseline results]	0.0	2	1	1	1
5907	130	collect statistic in a large corpus	[collecting statistics in a large corpus]	0.0	6	1	1	1
5908	130	evaluation of summarization quality	[evaluation of summarization quality]	0.0	4	1	1	1
5909	130	FERRET a question-answering q/a system	[FERRET an question-answering Q/A system]	0.0	5	1	1	1
5910	130	improvement of 22-38%	[improvement of 22-38%]	0.0	3	1	1	1
5911	130	argumentation system by Konolige	[argumentation system by Konolige]	0.0	4	1	1	1
5912	130	evaluation of machine translation	[evaluation of machine translation]	0.0	4	1	1	1
5913	130	name recognition	[name recognition]	0.0	2	1	1	1
5914	130	ordering model	[ordering model]	0.0	2	1	1	1
5915	130	information for use	[information for use]	0.0	3	1	1	1
5916	130	part of unification-based grammar parsing	[part of unification-based grammar parsing]	0.0	5	1	1	1
5917	130	speed-up over a SUN	[speed-up over a SUN]	0.0	4	1	1	1
5918	130	promising feature	[promising features]	0.0	2	1	1	1
5919	130	reflect the relationship rank	[reflecting the relationship ranks]	0.0	4	1	1	1
5920	130	new paradigm for interaction	[new paradigm for interaction]	0.0	4	1	1	1
5921	130	videodisc	[videodisc]	0.0	1	1	1	1
5922	130	correlation between the automatic evaluation	[correlation between the automatic evaluation]	0.0	5	1	1	1
5923	130	respectively on the October test set	[respectively on the October test set]	0.0	6	1	1	1
5924	130	long phrase	[long phrases]	0.0	2	1	1	1
5925	130	integral process	[integral process]	0.0	2	1	1	1
5926	130	protocol from answer	[protocols from answers]	0.0	3	1	1	1
5927	130	follow recent development in the evaluation	[Following recent developments in the evaluation]	0.0	6	1	1	1
5928	130	apartment	[apartment]	0.0	1	1	1	1
5929	130	edge as in treatment	[edges as in treatments]	0.0	4	1	1	1
5930	130	meaning through the use	[meaning through the use]	0.0	4	1	1	1
5931	130	set of high-precision rule	[sets of high-precision rules]	0.0	4	1	1	1
5932	130	require manual transcription	[requiring manual transcription]	0.0	3	1	1	1
5933	130	new concept	[new concepts]	0.0	2	1	1	1
5934	130	corner parsing algorithm	[corner parsing algorithm]	0.0	3	1	1	1
5935	130	word feature	[word features]	0.0	2	1	1	1
5936	130	improvement use the SI corpus	[improvement using the SI corpus]	0.0	5	1	1	1
5937	130	use a word sense	[Using a word sense]	0.0	4	1	1	1
5938	130	use a network	[using a network]	0.0	3	1	1	1
5939	130	99% accuracy	[99% accuracy]	0.0	2	1	1	1
5940	130	advantage from the point of view	[advantages from the point of view]	0.0	6	1	1	1
5941	130	robust method for speech understanding	[robust method for speech understanding]	0.0	5	1	1	1
5942	130	inference method	[inference method]	0.0	2	1	1	1
5943	130	good insight	[good insight]	0.0	2	1	1	1
5944	130	one speed-up element	[one speed-up element]	0.0	3	1	1	1
5945	130	Chinese word sense	[Chinese word sense]	0.0	3	1	1	1
5946	130	other contextual clue	[Other contextual clues]	0.0	3	1	1	1
5947	130	90 ,000 entry	[90 ,000 entries]	0.0	3	1	1	1
5948	130	question answer capability	[question answering capability]	0.0	3	1	1	1
5949	130	number of edge	[number of edges]	0.0	3	1	1	1
5950	130	i robust parsing of Korean language	[i Robust parsing of Korean language]	0.0	6	1	1	1
5951	130	only edge	[only edges]	0.0	2	1	1	1
5952	130	multi-site common evaluation	[multi-site common evaluation]	0.0	3	1	1	1
5953	130	various ambiguity	[various ambiguities]	0.0	2	1	1	1
5954	130	word token	[word tokens]	0.0	2	1	1	1
5955	130	system development to domain	[system development to domains]	0.0	4	1	1	1
5956	130	monitor the collection	[monitoring the collection]	0.0	3	1	1	1
5957	130	few year	[few years]	0.0	2	1	1	1
5958	130	two output from information extraction Ie	[two outputs from information extraction IE]	0.0	6	1	1	1
5959	130	model parameter than similar model	[model parameters than similar models]	0.0	5	1	1	1
5960	130	final language with case marker	[final language with case markers]	0.0	5	1	1	1
5961	130	involve	[involving]	0.0	1	1	1	1
5962	130	reflect the social relationship	[reflecting the social relationship]	0.0	4	1	1	1
5963	130	candidate for the error	[candidates for the errors]	0.0	4	1	1	1
5964	130	identification of 2-character Chinese name	[identification of 2-character Chinese names]	0.0	5	1	1	1
5965	130	Chinese name	[Chinese names]	0.0	2	1	1	1
5966	130	various feature from the context	[various features from the contexts]	0.0	5	1	1	1
5967	130	inflow of multilingual datum	[inflow of multilingual data]	0.0	4	1	1	1
5968	130	use knowledgebase	[using knowledgebases]	0.0	2	1	1	1
5969	130	rule-reduction algorithm apply information	[rule-reduction algorithm applying information]	0.0	4	1	1	1
5970	130	Coalition Language System	[Coalition Language System]	0.0	3	1	1	1
5971	130	clue	[clues]	0.0	1	1	1	1
5972	130	extend activity in a manner	[extending activity in a manner]	0.0	5	1	1	1
5973	130	good hypothesis corresponding	[best hypothesis corresponding]	0.0	3	1	1	1
5974	130	large-scale acquisition the construction	[large-scale acquisition the construction]	0.0	4	1	1	1
5975	130	transition probability	[transition probabilities]	0.0	2	1	1	1
5976	130	list of the feature set	[lists of the feature sets]	0.0	5	1	1	1
5977	130	multiple meaning	[multiple meanings]	0.0	2	1	1	1
5978	130	task-based evaluation use information retrieval	[Task-based evaluation using information retrieval]	0.0	5	1	1	1
5979	130	dependency parser	[dependency parser]	0.0	2	1	1	1
5980	130	satisfy human communication need	[satisfying human communication needs]	0.0	4	1	1	1
5981	130	sentence hypothese	[sentence hypotheses]	0.0	2	1	1	1
5982	130	indicator for the top-level prediction task	[indicators for the top-level prediction task]	0.0	6	1	1	1
5983	130	environment TGE	[environment TGE]	0.0	2	1	1	1
5984	130	rule-reduction algorithm	[rule-reduction algorithm]	0.0	2	1	1	1
5985	130	novel evaluation method	[novel evaluation method]	0.0	3	1	1	1
5986	130	35 .0% improvement over system	[35 .0% improvement over system]	0.0	5	1	1	1
5987	130	much work	[much work]	0.0	2	1	1	1
5988	130	miss	[missing]	0.0	1	1	1	1
5989	130	method lsa-based method	[methods LSA-based methods]	0.0	3	1	1	1
5990	130	lack of sense-tagged datum	[lack of sense-tagged data]	0.0	4	1	1	1
5991	130	heuristic learning	[heuristic learning]	0.0	2	1	1	1
5992	130	qa system performance	[QA system performance]	0.0	3	1	1	1
5993	130	large list of sentence plan	[large list of sentence plans]	0.0	5	1	1	1
5994	130	involve automatic extraction	[involving automatic extraction]	0.0	3	1	1	1
5995	130	likelihood criterion	[likelihood criterion]	0.0	2	1	1	1
5996	130	distinct domain	[distinct domains]	0.0	2	1	1	1
5997	130	strategy one utilize mechanism	[strategies one utilizing mechanisms]	0.0	4	1	1	1
5998	130	i parsing	[i parsing]	0.0	2	1	1	1
5999	130	source dependency	[source dependency]	0.0	2	1	1	1
6000	130	efficient broad-coverage parser	[efficient broad-coverage parser]	0.0	3	1	1	1
6001	130	.1% a reduction in error	[.1% a reduction in error]	0.0	5	1	1	1
6002	130	language understanding work	[Language understanding work]	0.0	3	1	1	1
6003	130	identify pattern in translation datum	[identifying patterns in translation data]	0.0	5	1	1	1
6004	130	use of prior in conditional model	[use of priors in conditional models]	0.0	6	1	1	1
6005	130	improvement on machine translation test	[improvement on machine translation tests]	0.0	5	1	1	1
6006	130	improvement over the WI system	[improvements over the WI systems]	0.0	5	1	1	1
6007	130	two system	[two systems]	0.0	2	1	1	1
6008	130	state-of-the-art machine translation system	[state-of-the-art machine translation system]	0.0	4	1	1	1
6009	130	abduction of temporal relation	[abduction of temporal relations]	0.0	4	1	1	1
6010	130	main parser	[main parser]	0.0	2	1	1	1
6011	130	individual rule	[individual rules]	0.0	2	1	1	1
6012	130	method add annotation	[method adding annotation]	0.0	3	1	1	1
6013	130	compare the expressiveness	[comparing the expressiveness]	0.0	3	1	1	1
6014	130	parsing time complexity	[parsing time complexity]	0.0	3	1	1	1
6015	130	amount of internal variation	[amount of internal variation]	0.0	4	1	1	1
6016	130	inflective language	[inflective languages]	0.0	2	1	1	1
6017	130	learning of phrase translation	[learning of phrase translations]	0.0	4	1	1	1
6018	130	strategy utilize primarily knowledge-based mechanism	[strategies utilizing primarily knowledge-based mechanisms]	0.0	5	1	1	1
6019	130	language with overt case marker	[language with overt case markers]	0.0	5	1	1	1
6020	130	take advantage	[taking advantage]	0.0	2	1	1	1
6021	130	major problem include vocabulary	[major problems including vocabulary]	0.0	4	1	1	1
6022	130	man	[men]	0.0	1	2	2	1
6023	130	more traditional language interface	[more traditional language interfaces]	0.0	4	1	1	1
6024	130	planning algorithm	[planning algorithm]	0.0	2	1	1	1
6025	130	attribute such as size perspective-taking	[attributes such as size perspective-taking]	0.0	5	1	1	1
6026	130	simple explanation	[simple explanation]	0.0	2	1	1	1
6027	130	reasoning with disposition	[reasoning with dispositions]	0.0	3	1	1	1
6028	130	examination of the effect	[Examination of the effect]	0.0	4	1	1	1
6029	130	New experimental result	[New experimental results]	0.0	3	1	1	1
6030	130	use processor aps	[using processors APs]	0.0	3	1	1	1
6031	130	domain independent acoustic model with classifier	[domain independent acoustic models with classifiers]	0.0	6	1	1	1
6032	130	copy of text	[copy of texts]	0.0	3	1	1	1
6033	130	i parsing language with case marker	[i parsing language with case markers]	0.0	6	1	1	1
6034	130	Chinese name without title	[Chinese names without title]	0.0	4	1	1	1
6035	130	generalization corresponding to arc	[generalization corresponding to arcs]	0.0	4	1	1	1
6036	130	evaluation of word sense	[evaluation of word sense]	0.0	4	1	1	1
6037	130	example discourse	[example discourses]	0.0	2	1	1	1
6038	130	pronominal anaphora resolution module	[pronominal anaphora resolution module]	0.0	4	1	1	1
6039	130	more remote event	[more remote events]	0.0	3	1	1	1
6040	130	sentence-level equivalence classification	[sentence-level equivalence classification]	0.0	3	1	1	1
6041	130	imitation of performance	[imitation of performance]	0.0	3	1	1	1
6042	130	broad-coverage parser for English	[broad-coverage parser for English]	0.0	4	1	1	1
6043	130	ideal	[ideal]	0.0	1	1	1	1
6044	130	tree adjoin	[tree adjoining]	0.0	2	1	1	1
6045	130	abduction	[abduction]	0.0	1	1	1	1
6046	130	man like mostly woman	[men like mostly women]	0.0	4	1	1	1
6047	130	analysis more robust---an agenda-based scheduling parser	[analysis more robust---an agenda-based scheduling parser]	0.0	6	1	1	1
6048	130	cross-relation between English	[cross-relations between English]	0.0	3	1	1	1
6049	130	indexation algorithm	[indexation algorithm]	0.0	2	1	1	1
6050	130	limitation in comparison	[limitations in comparison]	0.0	3	1	1	1
6051	130	LCS	[LCS]	0.0	1	1	1	1
6052	130	use conventional word-trigram recognition	[using conventional word-trigram recognition]	0.0	4	1	1	1
6053	130	pool of wsd system	[pool of WSD systems]	0.0	4	1	1	1
6054	130	elicit human judgment	[eliciting human judgments]	0.0	3	1	1	1
6055	130	i parsing of Korean language	[i parsing of Korean language]	0.0	5	1	1	1
6056	130	approach of grammar tailoring	[approach of grammar tailoring]	0.0	4	1	1	1
6057	130	probabilistic mapping	[probabilistic mapping]	0.0	2	1	1	1
6058	130	relationship email .g	[relationship e .g]	0.0	3	1	1	1
6059	130	NE type	[NE types]	0.0	2	1	1	1
6060	130	target domain	[target domain]	0.0	2	1	1	1
6061	130	deal of language text	[deal of language texts]	0.0	4	1	1	1
6062	130	parser/generator a transfer component	[parser/generator a transfer component]	0.0	4	1	1	1
6063	130	tutorial dialogue	[tutorial dialogue]	0.0	2	1	1	1
6064	130	definition question	[definition questions]	0.0	2	1	1	1
6065	130	utterance from site	[utterances from sites]	0.0	3	1	1	1
6066	130	korean-to-english machine translation system	[Korean-to-English machine translation system]	0.0	4	1	1	1
6067	130	109 training speaker	[109 training speakers]	0.0	3	1	1	1
6068	130	train spot on the basis	[training SPoT on the basis]	0.0	5	1	1	1
6069	130	i parsing a verb final language	[i parsing a verb final language]	0.0	6	1	1	1
6070	130	word-sense disambiguation system	[word-sense disambiguation systems]	0.0	3	1	1	1
6071	130	content understanding	[content understanding]	0.0	2	1	1	1
6072	130	style and/or	[style and/or]	0.0	2	1	1	1
6073	130	aspect of the predictive performance	[aspects of the predictive performance]	0.0	5	1	1	1
6074	130	nlp problem	[NLP problems]	0.0	2	1	1	1
6075	130	development time	[development time]	0.0	2	1	1	1
6076	130	dictionary MRD's	[dictionaries MRD's]	0.0	2	1	1	1
6077	130	insightful counterexample	[insightful counterexamples]	0.0	2	1	1	1
6078	130	pair of probabilistic grammar	[pairs of probabilistic grammars]	0.0	4	1	1	1
6079	130	machine translation module	[machine translation modules]	0.0	3	1	1	1
6080	130	corpus analysis accessible to researcher	[corpus analysis accessible to researchers]	0.0	5	1	1	1
6081	130	machine translation quality	[machine translation quality]	0.0	3	1	1	1
6082	130	progress from generation	[progressing from generation]	0.0	3	1	1	1
6083	130	strategy one	[strategies one]	0.0	2	1	1	1
6084	130	utterance of spontaneous speech	[utterances of spontaneous speech]	0.0	4	1	1	1
6085	130	learning approach build a stemmer	[learning approach building a stemmer]	0.0	5	1	1	1
6086	130	middle	[middle]	0.0	1	1	1	1
6087	130	kind include reference failure	[kinds including reference failures]	0.0	4	1	1	1
6088	130	design for agent	[design for agents]	0.0	3	1	1	1
6089	130	kind of misunderstanding	[kinds of misunderstandings]	0.0	3	1	1	1
6090	130	collaboration	[collaborations]	0.0	1	1	1	1
6091	130	technology of personal computer	[technology of personal computers]	0.0	4	1	1	1
6092	130	use bilingual parallel corpus a resource	[using bilingual parallel corpora a resource]	0.0	6	1	1	1
6093	130	different lm	[different LM]	0.0	2	1	1	1
6094	130	datum structure	[data structure]	0.0	2	1	1	1
6095	130	12,000 utterance from five site	[12,000 utterances from five sites]	0.0	5	1	1	1
6096	130	segment contiguity model	[segment contiguity models]	0.0	3	1	1	1
6097	130	i parsing final language	[i parsing final language]	0.0	4	1	1	1
6098	130	analysis robust---an agenda-based scheduling parser	[analysis robust---an agenda-based scheduling parser]	0.0	5	1	1	1
6099	130	condition for the use	[conditions for the use]	0.0	4	1	1	1
6100	130	improve the performance	[improving the performance]	0.0	3	1	1	1
6101	130	free word	[free word]	0.0	2	1	1	1
6102	130	scoring system output	[scoring system output]	0.0	3	1	1	1
6103	130	strategy for language understanding	[strategies for language understanding]	0.0	4	1	1	1
6104	130	industry watch	[industry watch]	0.0	2	1	1	1
6105	130	small sentence parallel corpus	[small sentences parallel corpus]	0.0	4	1	1	1
6106	130	information from a large inflow	[information from a large inflow]	0.0	5	1	1	1
6107	130	language processing nlp system	[Language Processing NLP systems]	0.0	4	1	1	1
6108	130	acquisition of word-semantic information	[acquisition of word-semantic information]	0.0	4	1	1	1
6109	130	155 million word	[155 million word]	0.0	3	1	1	1
6110	130	adjacency graph's laplacian	[adjacency graph's Laplacian]	0.0	3	1	1	1
6111	130	correct homophone error in compound noun	[correcting homophone errors in compound nouns]	0.0	6	1	1	1
6112	130	modelling	[modelling]	0.0	1	1	1	1
6113	130	deal of natural language text	[deal of natural language texts]	0.0	5	1	1	1
6114	130	use in a evaluation	[use in a evaluation]	0.0	4	1	1	1
6115	130	more frequently predictor	[more frequently predictors]	0.0	3	1	1	1
6116	130	temporal adverb	[temporal adverbs]	0.0	2	1	1	1
6117	130	clustering process	[clustering process]	0.0	2	1	1	1
6118	130	inherently a conglomeration of individual rule	[inherently a conglomeration of individual rules]	0.0	6	1	1	1
6119	130	annealing	[annealing]	0.0	1	1	1	1
6120	130	common claim	[common claim]	0.0	2	1	1	1
6121	130	use this idea	[Using these ideas]	0.0	3	1	1	1
6122	130	path of arc	[path of arcs]	0.0	3	1	1	1
6123	130	geometric model	[geometric model]	0.0	2	1	1	1
6124	130	weather forecast	[weather forecasts]	0.0	2	1	1	1
6125	130	associative processor aps	[associative processors APs]	0.0	3	1	1	1
6126	130	parsing model	[parsing models]	0.0	2	1	1	1
6127	130	automatic scheme	[automatic scheme]	0.0	2	1	1	1
6128	130	last year	[last years]	0.0	2	1	1	1
6129	130	use in such mt	[use in such MT]	0.0	4	1	1	1
6130	130	hardness result	[hardness results]	0.0	2	1	1	1
6131	130	unification algorithm	[unification algorithms]	0.0	2	1	1	1
6132	130	two algorithm	[two algorithms]	0.0	2	1	1	1
6133	130	proper substitution	[proper substitutions]	0.0	2	1	1	1
6134	130	meaning structure	[meaning structure]	0.0	2	1	1	1
6135	130	employ a generative approach	[employing a generative approach]	0.0	4	1	1	1
6136	130	English-Chinese corpus	[English-Chinese corpora]	0.0	2	1	1	1
6137	130	forward-backward algorithm	[forward-backward algorithm]	0.0	2	1	1	1
6138	130	more redundancy	[more redundancy]	0.0	2	1	1	1
6139	130	scope information of the structure	[scope information of the structure]	0.0	5	1	1	1
6140	130	interactive environment	[interactive environment]	0.0	2	1	1	1
6141	130	robust efficient parsing a verb	[Robust efficient parsing a verb]	0.0	5	1	1	1
6142	130	problem in user modeling	[problem in user modeling]	0.0	4	1	1	1
6143	130	two dataset	[two datasets]	0.0	2	1	1	1
6144	130	importance to Machine Translation	[importance to Machine Translation]	0.0	4	1	1	1
6145	130	lexical-functional Grammar	[Lexical-Functional Grammars]	0.0	2	1	1	1
6146	130	additional 500 feature	[additional 500 features]	0.0	3	1	1	1
6147	130	such linguistic database	[such linguistic databases]	0.0	3	1	1	1
6148	130	automatically acquire new stem	[automatically acquiring new stems]	0.0	4	1	1	1
6149	130	predict topic shift to the problem	[predicting topic shifts to the problem]	0.0	6	1	1	1
6150	130	use a phrase	[using a phrase]	0.0	3	1	1	1
6151	130	number of abstract	[number of abstracts]	0.0	3	1	1	1
6152	130	complex language processing	[complex language processing]	0.0	3	1	1	1
6153	130	segment boundary	[segment boundaries]	0.0	2	1	1	1
6154	130	hold candidate for understanding result	[holding candidates for understanding results]	0.0	5	2	2	1
6155	130	enter through message passing	[entering through message passing]	0.0	4	1	1	1
6156	130	writing	[writing]	0.0	1	1	1	1
6157	130	89 .75% f-measure a 13% decrease	[89 .75% F-measure a 13% decrease]	0.0	6	1	1	1
6158	130	potential increase in mrr	[potential increase in MRR]	0.0	4	1	1	1
6159	130	investigate the combinatorics	[investigating the combinatorics]	0.0	3	1	1	1
6160	130	language understand system	[language understanding system]	0.0	3	1	1	1
6161	130	triplet of word	[triplets of words]	0.0	3	1	1	1
6162	130	word sense induction	[word sense induction]	0.0	3	1	1	1
6163	130	use word-trigram recognition	[using word-trigram recognition]	0.0	3	1	1	1
6164	130	linguistic analysis component	[linguistic analysis component]	0.0	3	1	1	1
6165	130	simple sentence-plan-generator SPG	[simple sentence-plan-generator SPG]	0.0	3	1	1	1
6166	130	SimTran	[SimTran]	0.0	1	1	1	1
6167	130	three dimension of user model	[three dimensions of user models]	0.0	5	1	1	1
6168	130	appealing term	[appealing terms]	0.0	2	1	1	1
6169	130	precise way	[precise way]	0.0	2	1	1	1
6170	130	less memory	[less memory]	0.0	2	1	1	1
6171	130	iii system development via knowledge-based acquisition	[iii system development via knowledge-based acquisition]	0.0	6	1	1	1
6172	130	relation between entity	[relations between entities]	0.0	3	1	1	1
6173	130	factor behind this finding include limitation	[factors behind this finding including limitations]	0.0	6	1	1	1
6174	130	explanation of such discourse phenomenon	[explanation of such discourse phenomena]	0.0	5	1	1	1
6175	130	information from the parse tree	[information from the parse tree]	0.0	5	1	1	1
6176	130	DARPA speech	[DARPA speech]	0.0	2	1	1	1
6177	130	understand technology to language understanding	[understanding technology to language understanding]	0.0	5	1	1	1
6178	130	histogram	[histogram]	0.0	1	1	1	1
6179	130	estimation	[estimation]	0.0	1	1	1	1
6180	130	system development via acquisition	[system development via acquisition]	0.0	4	1	1	1
6181	130	follow development in the automatic evaluation	[Following developments in the automatic evaluation]	0.0	6	1	1	1
6182	130	translation of dependency structure	[translation of dependency structures]	0.0	4	1	1	1
6183	130	many different form	[many different forms]	0.0	3	1	1	1
6184	130	many structure	[many structures]	0.0	2	1	1	1
6185	130	class of algorithm	[classes of algorithms]	0.0	3	1	1	1
6186	130	i efficient parsing of Korean	[i efficient parsing of Korean]	0.0	5	1	1	1
6187	130	result on addressee identification meeting	[results on addressee identification meetings]	0.0	5	1	1	1
6188	130	work harvesting bitext of the law	[work harvesting bitexts of the laws]	0.0	6	1	1	1
6189	130	extend the CSR corpus	[extending the CSR corpus]	0.0	4	1	1	1
6190	130	string with the performance (typically	[string with the performance (typically]	0.0	5	1	1	1
6191	130	database of tv show	[database of TV shows]	0.0	4	1	1	1
6192	130	two word-sense disambiguation system	[two word-sense disambiguation systems]	0.0	4	1	1	1
6193	130	Chinese language processing	[Chinese Language Processing]	0.0	3	1	1	1
6194	130	effective use of prior	[effective use of priors]	0.0	4	1	1	1
6195	130	chain of syllable	[chain of syllables]	0.0	3	1	1	1
6196	130	command	[commands]	0.0	1	1	1	1
6197	130	progress into the output	[progressing into the output]	0.0	4	1	1	1
6198	130	parallel sentence in non-parallel corpus	[parallel sentences in non-parallel corpora]	0.0	5	1	1	1
6199	130	language understand technology to language understanding	[language understanding technology to language understanding]	0.0	6	1	1	1
6200	130	nature of the fragment	[nature of the fragments]	0.0	4	1	1	1
6201	130	speech act	[speech acts]	0.0	2	1	1	1
6202	130	development model	[development model]	0.0	2	1	1	1
6203	130	probabilistic parsing model	[probabilistic parsing models]	0.0	3	1	1	1
6204	130	Street	[Street]	0.0	1	1	1	1
6205	130	retrieval accuracy superior	[retrieval accuracy superior]	0.0	3	1	1	1
6206	130	traditional language interface	[traditional language interfaces]	0.0	3	1	1	1
6207	130	Dynamic Phrasal Lexicon DHPL [Zernik88]	[Dynamic Phrasal Lexicon DHPL [Zernik88]]	0.0	5	1	1	1
6208	130	recent claim	[recent claims]	0.0	2	1	1	1
6209	130	require development time	[requiring development time]	0.0	3	1	1	1
6210	130	homophone error in compound noun	[homophone errors in compound nouns]	0.0	5	2	2	1
6211	130	Rapid system development to new domain	[Rapid system development to new domains]	0.0	6	1	1	1
6212	130	framework restrictive	[framework restrictive]	0.0	2	1	1	1
6213	130	Hierarchical Phrasal Lexicon	[Hierarchical Phrasal Lexicon]	0.0	3	1	1	1
6214	130	selection	[selection]	0.0	1	2	2	1
6215	130	boolean condition	[Boolean conditions]	0.0	2	1	1	1
6216	130	word from high-accuracy alignment model	[words from high-accuracy alignment models]	0.0	5	1	1	1
6217	130	corpus consist of ,610 sentence	[corpus consisting of ,610 sentences]	0.0	5	1	1	1
6218	130	multi-site evaluation of speech	[multi-site evaluation of speech]	0.0	4	1	1	1
6219	130	forest	[forest]	0.0	1	1	1	1
6220	130	equivalence between the use	[equivalence between the use]	0.0	4	1	1	1
6221	130	language model combination	[language model combination]	0.0	3	1	1	1
6222	130	generation of true text	[generation of true text]	0.0	4	1	1	1
6223	130	use tag sequence	[using tag sequences]	0.0	3	1	1	1
6224	130	increase of correlation	[increase of correlation]	0.0	3	1	1	1
6225	130	use triplet instead of pair	[using triplets instead of pairs]	0.0	5	1	1	1
6226	130	collection of utterance of speech	[collection of utterances of speech]	0.0	5	1	1	1
6227	130	transform a disposition	[transforming a disposition]	0.0	3	1	1	1
6228	130	most benefit	[most benefit]	0.0	2	1	1	1
6229	130	extract information	[extracting information]	0.0	2	1	1	1
6230	130	text processing software	[text processing software]	0.0	3	1	1	1
6231	130	improvement in recognition accuracy	[improvement in recognition accuracy]	0.0	4	1	1	1
6232	130	comparison with wsd model	[comparison with WSD models]	0.0	4	1	1	1
6233	130	already in place	[already in place]	0.0	3	1	1	1
6234	130	man like woman	[men like women]	0.0	3	2	2	1
6235	130	analysis of answer	[analysis of answers]	0.0	3	1	1	1
6236	130	learning Model SLM	[Learning Model SLM]	0.0	3	1	1	1
6237	130	average the statistic	[averaging the statistics]	0.0	3	1	1	1
6238	130	different answer agent search	[different answering agents searching]	0.0	4	1	1	1
6239	130	switchboard dialogue	[Switchboard dialogues]	0.0	2	1	1	1
6240	130	precede tag context	[preceding tag contexts]	0.0	3	1	1	1
6241	130	achieve similar performance	[achieving similar performance]	0.0	3	1	1	1
6242	130	novel solution	[novel solution]	0.0	2	1	1	1
6243	130	much debate	[much debate]	0.0	2	1	1	1
6244	130	effect of the sense	[effect of the sense]	0.0	4	1	1	1
6245	130	verbal means	[verbal means]	0.0	2	1	1	1
6246	130	accurate word	[accurate word]	0.0	2	1	1	1
6247	130	average time	[average times]	0.0	2	1	1	1
6248	130	99% correct sentence	[99% correct sentences]	0.0	3	1	1	1
6249	130	information piece	[information piece]	0.0	2	1	1	1
6250	130	.4m English word	[.4M English words]	0.0	3	1	1	1
6251	130	validity of such linguistic database	[validity of such linguistic databases]	0.0	5	1	1	1
6252	130	classification method	[classification method]	0.0	2	1	1	1
6253	130	state dynamic	[state dynamic]	0.0	2	1	1	1
6254	130	volume	[volume]	0.0	1	1	1	1
6255	130	10k sentence parallel corpus	[10K sentences parallel corpus]	0.0	4	1	1	1
6256	130	principle-based parser for English	[principle-based parser for English]	0.0	4	1	1	1
6257	130	approach for evaluation	[approach for evaluation]	0.0	3	1	1	1
6258	130	means heuristic learning	[means heuristic learning]	0.0	3	1	1	1
6259	130	automatic translation	[automatic translation]	0.0	2	1	1	1
6260	130	broad use include jointly conditioning	[broad use including jointly conditioning]	0.0	5	1	1	1
6261	130	effective use of prior in model	[effective use of priors in models]	0.0	6	1	1	1
6262	130	paraphrasing of output	[paraphrasing of output]	0.0	3	1	1	1
6263	130	breakdown	[breakdown]	0.0	1	1	1	1
6264	130	user evaluation	[user evaluation]	0.0	2	1	1	1
6265	130	model parameter than phrase-based model	[model parameters than phrase-based models]	0.0	5	1	1	1
6266	130	use a non-terminal symbol	[using a non-terminal symbol]	0.0	4	1	1	1
6267	130	abduction of relation	[abduction of relations]	0.0	3	1	1	1
6268	130	language processing nlp	[Language Processing NLP]	0.0	3	1	1	1
6269	130	object-oriented language parsing	[object-oriented language parsing]	0.0	3	1	1	1
6270	130	utilize example	[utilizing examples]	0.0	2	1	1	1
6271	130	NTHU's system	[NTHU's system]	0.0	2	1	1	1
6272	130	coverage lexicon	[coverage lexicon]	0.0	2	1	1	1
6273	130	involve extraction of translation lexicon	[involving extraction of translation lexicons]	0.0	5	1	1	1
6274	130	generative approach	[generative approach]	0.0	2	1	1	1
6275	130	two disambiguation system one	[two disambiguation systems one]	0.0	4	1	1	1
6276	130	ii high quality translation	[ii High quality translation]	0.0	4	1	1	1
6277	130	automate certain aspect	[automating certain aspects]	0.0	3	1	1	1
6278	130	also in the area	[also in the areas]	0.0	4	1	1	1
6279	130	arc between node	[arcs between nodes]	0.0	3	1	1	1
6280	130	evaluation result involve extraction from text	[evaluation results involving extraction from text]	0.0	6	1	1	1
6281	130	category transition	[category transitions]	0.0	2	1	1	1
6282	130	Arabic' morphology	[Arabic's morphology]	0.0	2	1	1	1
6283	130	corpus of the language	[corpus of the language]	0.0	4	1	1	1
6284	130	exist parsing model	[existing parsing models]	0.0	3	1	1	1
6285	130	unsupervised technique	[unsupervised technique]	0.0	2	1	1	1
6286	130	elicit judgment	[eliciting judgments]	0.0	2	1	1	1
6287	130	event-based summarization approach	[event-based summarization approaches]	0.0	3	1	1	1
6288	130	expression in English	[expressions in English]	0.0	3	1	1	1
6289	130	disambiguation the interpretation	[disambiguation the interpretation]	0.0	3	1	1	1
6290	130	.4m word	[.4M words]	0.0	2	1	1	1
6291	130	formulas with value	[formulas with values]	0.0	3	1	1	1
6292	130	judge	[judges]	0.0	1	1	1	1
6293	130	application of path-based inference rule	[Applications of path-based inference rules]	0.0	5	1	1	1
6294	130	build polarity-tagged corpus	[building polarity-tagged corpus]	0.0	3	1	1	1
6295	130	45% reduction	[45% reduction]	0.0	2	1	1	1
6296	130	155 million word corpus	[155 million word corpus]	0.0	4	1	1	1
6297	130	result word segmentation system	[resulting word segmentation system]	0.0	4	1	1	1
6298	130	relevance	[relevance]	0.0	1	1	1	1
6299	130	important bear on reasoning	[important bearing on reasoning]	0.0	4	1	1	1
6300	130	concern translation equivalence	[concerning translation equivalence]	0.0	3	1	1	1
6301	130	scheme collect statistic	[scheme collecting statistics]	0.0	3	1	1	1
6302	130	real-valued feature	[real-valued features]	0.0	2	1	1	1
6303	130	subcategorization frame distribution	[subcategorization frame distributions]	0.0	3	1	1	1
6304	130	result cast doubt	[results casting doubt]	0.0	3	1	1	1
6305	130	acquisition of information email	[acquisition of information e]	0.0	4	1	1	1
6306	130	two formalism tree Grammar	[two formalisms Tree Grammars]	0.0	4	1	1	1
6307	130	key factor	[key factors]	0.0	2	1	1	1
6308	130	light of function	[light of functions]	0.0	3	1	1	1
6309	130	wsd evaluation	[WSD evaluation]	0.0	2	1	1	1
6310	130	aspect include the influence	[aspects including the influence]	0.0	4	1	1	1
6311	130	understand system reasoning	[understanding system reasoning]	0.0	3	1	1	1
6312	130	acquisition of constraint in broad domain	[acquisition of constraints in broad domains]	0.0	6	1	1	1
6313	130	most text-understanding system	[Most text-understanding systems]	0.0	3	1	1	1
6314	130	translation lexicon	[translation lexicons]	0.0	2	1	1	1
6315	130	phrase boundary	[phrase boundary]	0.0	2	1	1	1
6316	130	constraint on the translation relation	[constraints on the translation relation]	0.0	5	1	1	1
6317	130	datum relevant to task	[data relevant to task]	0.0	4	1	1	1
6318	130	induce verb class from corpus datum	[inducing verb classes from corpus data]	0.0	6	1	1	1
6319	130	wide coverage grammar	[wide coverage grammar]	0.0	3	1	1	1
6320	130	use dialogue datum	[using dialogue data]	0.0	3	1	1	1
6321	130	use a out-of-domain corpus	[using an out-of-domain corpus]	0.0	4	1	1	1
6322	130	75% accuracy	[75% accuracy]	0.0	2	1	1	1
6323	130	reestimation	[reestimation]	0.0	1	1	1	1
6324	130	iv modeling of word feature	[iv modeling of word features]	0.0	5	1	1	1
6325	130	pre-exist component	[pre-existing components]	0.0	2	1	1	1
6326	130	important role in cognition	[important role in cognition]	0.0	4	1	1	1
6327	130	proposition with implicit quantifier	[proposition with implicit quantifiers]	0.0	4	1	1	1
6328	130	semantic role	[semantic roles]	0.0	2	1	1	1
6329	130	coordinate sentence	[coordinate sentences]	0.0	2	1	1	1
6330	130	planning-based architecture	[planning-based architecture]	0.0	2	1	1	1
6331	130	apply automated evaluation technique	[applying automated evaluation techniques]	0.0	4	1	1	1
6332	130	explanation for the nature	[explanation for the nature]	0.0	4	1	1	1
6333	130	three method method	[three methods methods]	0.0	3	1	1	1
6334	130	application of Unification Grammar UCG	[application of Unification Grammar UCG]	0.0	5	1	1	1
6335	130	Wall Street Journal treebank	[Wall Street Journal treebank]	0.0	4	1	1	1
6336	130	result cast doubt on this assumption	[results casting doubt on this assumption]	0.0	6	1	1	1
6337	130	addition to a component	[addition to an component]	0.0	4	1	1	1
6338	130	numbering	[numbering]	0.0	1	1	1	1
6339	130	domain of HK law	[domain of HK laws]	0.0	4	1	1	1
6340	130	feature of a representation	[features of a representation]	0.0	4	1	1	1
6341	130	cooperative Model for language understanding	[Cooperative Model for language understanding]	0.0	5	1	1	1
6342	130	more datum	[more data]	0.0	2	1	1	1
6343	130	corpus resource	[corpus resource]	0.0	2	1	1	1
6344	130	representation the Phrasal Lexicon	[representation the Phrasal Lexicon]	0.0	4	1	1	1
6345	130	property of grammar	[properties of grammars]	0.0	3	1	1	1
6346	130	direct the process	[directing the process]	0.0	3	1	1	1
6347	130	corresponding with the confidence	[corresponding with the confidence]	0.0	4	1	1	1
6348	130	such database	[such databases]	0.0	2	1	1	1
6349	130	method construct word vector	[methods constructing word vectors]	0.0	4	1	1	1
6350	130	several advantage from the point	[Several advantages from the point]	0.0	5	1	1	1
6351	130	basis for the large-scale acquisition	[basis for the large-scale acquisition]	0.0	5	1	1	1
6352	130	grammar parsing	[grammar parsing]	0.0	2	1	1	1
6353	130	neat relation	[neat relation]	0.0	2	1	1	1
6354	130	output of a parser	[output of an parser]	0.0	4	1	1	1
6355	130	fulfillment of well-formedness condition	[fulfillment of well-formedness conditions]	0.0	4	1	1	1
6356	130	precision over unstemmed text	[precision over unstemmed text]	0.0	4	1	1	1
6357	130	technique for analysis	[techniques for analysis]	0.0	3	1	1	1
6358	130	parsing time	[parsing time]	0.0	2	1	1	1
6359	130	generative character recognition ocr model	[generative character recognition OCR model]	0.0	5	1	1	1
6360	130	general preference	[general preference]	0.0	2	1	1	1
6361	130	possible word-sense	[possible word-senses]	0.0	2	1	1	1
6362	130	formalization	[formalization]	0.0	1	1	1	1
6363	130	SMT system	[SMT system]	0.0	2	1	1	1
6364	130	domain independent model with off-the-shelf classifier	[domain independent models with off-the-shelf classifiers]	0.0	6	1	1	1
6365	130	achieve baseline result in a time	[achieving baseline results in a time]	0.0	6	1	1	1
6366	130	understand system with intelligent agent	[understanding system with intelligent agents]	0.0	5	1	1	1
6367	130	language modeling of speech	[language modeling of speech]	0.0	4	1	1	1
6368	130	meeting context	[meeting context]	0.0	2	2	2	1
6369	130	bilingual resource	[bilingual resource]	0.0	2	1	1	1
6370	130	machine translation system alone	[machine translation system alone]	0.0	4	1	1	1
6371	130	fully-connected class grammar	[fully-connected class grammar]	0.0	3	1	1	1
6372	130	combination of structure	[combination of structures]	0.0	3	1	1	1
6373	130	domain entity	[domain entities]	0.0	2	1	1	1
6374	130	domain strategy for the articulation	[domain strategy for the articulation]	0.0	5	1	1	1
6375	130	namely the algorithm	[namely the algorithm]	0.0	3	1	1	1
6376	130	grammar generation system	[grammar generation system]	0.0	3	1	1	1
6377	130	Konolige	[Konolige]	0.0	1	1	1	1
6378	130	various kind include reference failure	[various kinds including reference failures]	0.0	5	1	1	1
6379	130	event type	[event type]	0.0	2	1	1	1
6380	130	category in tree	[categories in trees]	0.0	3	1	1	1
6381	130	manual evaluation	[manual evaluation]	0.0	2	1	1	1
6382	130	other strategy	[other strategies]	0.0	2	1	1	1
6383	130	employ n-gram model for prediction	[employing n-gram models for prediction]	0.0	5	1	1	1
6384	130	only 14 .0%	[only 14 .0%]	0.0	3	1	1	1
6385	130	word-sense disambiguation system one	[word-sense disambiguation systems one]	0.0	4	1	1	1
6386	130	evaluation of string	[evaluation of strings]	0.0	3	1	1	1
6387	130	three domain-independent capability	[three domain-independent capabilities]	0.0	3	1	1	1
6388	130	vagueness of the concept	[vagueness of the concepts]	0.0	4	1	1	1
6389	130	SI recognition	[SI recognition]	0.0	2	1	1	1
6390	130	two language pair	[two language pairs]	0.0	3	1	1	1
6391	130	advantage of the pattern	[advantage of the patterns]	0.0	4	1	1	1
6392	130	statistical learning Model	[Statistical Learning Model]	0.0	3	1	1	1
6393	130	direct imitation	[direct imitation]	0.0	2	1	1	1
6394	130	correct this error in bunsetsu	[correcting these errors in bunsetsu]	0.0	5	1	1	1
6395	130	computational algorithm	[computational algorithms]	0.0	2	1	1	1
6396	130	word co-occurrence	[word co-occurrences]	0.0	2	1	1	1
6397	130	different strategy one	[different strategies one]	0.0	3	1	1	1
6398	130	computation of the form	[computation of the form]	0.0	4	1	1	1
6399	130	document classification	[document classification]	0.0	2	2	2	1
6400	130	ambiguity packing technique	[ambiguity packing techniques]	0.0	3	1	1	1
6401	130	cull information	[culling information]	0.0	2	1	1	1
6402	130	information email	[information e]	0.0	2	1	1	1
6403	130	recent work harvesting bitext	[recent work harvesting bitexts]	0.0	4	1	1	1
6404	130	translingual reach	[translingual reach]	0.0	2	1	1	1
6405	130	improvement for speaker adaptation	[improvement for speaker adaptation]	0.0	4	1	1	1
6406	130	design with the Senseval series	[designing with the Senseval series]	0.0	5	1	1	1
6407	130	automatically gather a number	[automatically gathering a number]	0.0	4	1	1	1
6408	130	automatic evaluation method	[automatic evaluation method]	0.0	3	1	1	1
6409	130	knowledge about dialog	[knowledge about dialogs]	0.0	3	1	1	1
6410	130	pick the good hypothesis	[picking the best hypothesis]	0.0	4	1	1	1
6411	130	pair of context-free grammar work	[pairs of context-free grammars working]	0.0	5	1	1	1
6412	130	system with agent	[system with agents]	0.0	3	1	1	1
6413	130	individually good predictor	[individually good predictors]	0.0	3	1	1	1
6414	130	problem combine language model	[problem combining language models]	0.0	4	1	1	1
6415	130	such language	[such languages]	0.0	2	1	1	1
6416	130	gather a number	[gathering a number]	0.0	3	1	1	1
6417	130	simple interpolation method like interpolation	[simple interpolation methods like interpolation]	0.0	5	1	1	1
6418	130	various property	[Various properties]	0.0	2	1	1	1
6419	130	provide a interpretation	[providing an interpretation]	0.0	3	1	1	1
6420	130	decrease	[decrease]	0.0	1	1	1	1
6421	130	third of the sentence	[third of the sentences]	0.0	4	1	1	1
6422	130	parsing of Korean	[parsing of Korean]	0.0	3	1	1	1
6423	130	category of the word	[categories of the words]	0.0	4	1	1	1
6424	130	new linguistic representation	[new linguistic representation]	0.0	3	1	1	1
6425	130	suppress the fuzzy quantifier	[suppressing the fuzzy quantifier]	0.0	4	1	1	1
6426	130	make syntactic analysis robust---an scheduling parser	[making syntactic analysis robust---an scheduling parser]	0.0	6	1	1	1
6427	130	argumentation system	[argumentation system]	0.0	2	1	1	1
6428	130	non	[Non]	0.0	1	1	1	1
6429	130	scalability	[scalability]	0.0	1	1	1	1
6430	130	combiner with hard decision	[combiner with hard decisions]	0.0	4	1	1	1
6431	130	sophisticated representation for document classification	[sophisticated representations for document classification]	0.0	5	1	1	1
6432	130	large amount	[large amount]	0.0	2	2	2	1
6433	130	new paradigm for SI training	[new paradigm for SI training]	0.0	5	1	1	1
6434	130	reduction in the search space	[reduction in the search space]	0.0	5	1	1	1
6435	130	interpolation method	[interpolation methods]	0.0	2	1	1	1
6436	130	two new disambiguation system	[two new disambiguation systems]	0.0	4	1	1	1
6437	130	very small corpus	[very small corpus]	0.0	3	1	1	1
6438	130	paraphrase mean equivalence	[paraphrases meaning equivalence]	0.0	3	1	1	1
6439	130	such as system	[such as systems]	0.0	3	1	1	1
6440	130	parsing a verb	[parsing a verb]	0.0	3	1	1	1
6441	130	appealing term of simplicity	[appealing terms of simplicity]	0.0	4	1	1	1
6442	130	compare the expressiveness of the formalism	[comparing the expressiveness of the formalisms]	0.0	6	1	1	1
6443	130	antecedence recovery	[antecedence recovery]	0.0	2	1	1	1
6444	130	subset 14 .0%	[subset 14 .0%]	0.0	3	1	1	1
6445	130	other kernel function	[other kernel functions]	0.0	3	1	1	1
6446	130	word order generation	[word order generation]	0.0	3	1	1	1
6447	130	San Diego	[San Diego]	0.0	2	1	1	1
6448	130	set of phrase	[set of phrases]	0.0	3	1	1	1
6449	130	.24%	[.24%]	0.0	1	1	1	1
6450	130	generation of text	[generation of text]	0.0	3	1	1	1
6451	130	probabilistic optical character recognition model	[probabilistic optical character recognition model]	0.0	5	1	1	1
6452	130	more available resource	[more available resource]	0.0	3	1	1	1
6453	130	machine learning experiment	[machine learning experiments]	0.0	3	1	1	1
6454	130	term of accuracy	[terms of accuracy]	0.0	3	1	1	1
6455	130	Lucy a understanding system	[Lucy a understanding system]	0.0	4	1	1	1
6456	130	include a word	[including a word]	0.0	3	1	1	1
6457	130	Hong Kong	[Hong Kong]	0.0	2	1	1	1
6458	130	use only 40 utterance for adaptation	[Using only 40 utterances for adaptation]	0.0	6	1	1	1
6459	130	type of finite automaton	[type of finite automaton]	0.0	4	1	1	1
6460	130	stage of 1	[stages of 1]	0.0	3	1	1	1
6461	130	task of semantic interpretation	[task of semantic interpretation]	0.0	4	1	1	1
6462	130	research on major problem include vocabulary	[research on major problems including vocabulary]	0.0	6	1	1	1
6463	130	many grammar formalism	[many grammar formalisms]	0.0	3	1	1	1
6464	130	IBM statistical mt system	[IBM statistical MT system]	0.0	4	1	1	1
6465	130	text hierarchy	[text hierarchy]	0.0	2	1	1	1
6466	130	therefore a choice for the parser	[therefore a choice for the parser]	0.0	6	1	1	1
6467	130	harvesting bitext of the law	[harvesting bitexts of the laws]	0.0	5	1	1	1
6468	130	extractive mutli-document summarization approach	[extractive mutli-document summarization approaches]	0.0	4	1	1	1
6469	130	definition of typicality	[definition of typicality]	0.0	3	1	1	1
6470	130	parsing language	[parsing language]	0.0	2	1	1	1
6471	130	state-of-the-art question	[State-of-the-art Question]	0.0	2	1	1	1
6472	130	resource management rm corpus	[Resource Management RM corpus]	0.0	4	1	1	1
6473	130	width	[width]	0.0	1	1	1	1
6474	130	hypertext link	[hypertext links]	0.0	2	1	1	1
6475	130	cull information from a inflow	[culling information from a inflow]	0.0	5	1	1	1
6476	130	algorithm automatically acquire stem	[algorithm automatically acquiring stems]	0.0	4	1	1	1
6477	130	indicator	[indicators]	0.0	1	1	1	1
6478	130	message passing between actor	[message passing between actors]	0.0	4	1	1	1
6479	130	broad-coverage principle-based parser for English	[broad-coverage principle-based parser for English]	0.0	5	1	1	1
6480	130	role from the discourse	[role from the discourse]	0.0	4	1	1	1
6481	130	technique make syntactic analysis	[techniques making syntactic analysis]	0.0	4	1	1	1
6482	130	corpus of 110 word	[corpus of 110 words]	0.0	4	1	1	1
6483	130	use a Chinese word sense	[Using a Chinese word sense]	0.0	5	1	1	1
6484	130	large number of abstract	[large number of abstracts]	0.0	4	1	1	1
6485	130	theory of plausible reasoning	[theory of plausible reasoning]	0.0	4	1	1	1
6486	130	principle advantage	[principle advantage]	0.0	2	1	1	1
6487	130	head-to-head test	[head-to-head tests]	0.0	2	1	1	1
6488	130	employ a traditional phonological approach	[employing a traditional phonological approach]	0.0	5	1	1	1
6489	130	message passing	[message passing]	0.0	2	1	1	1
6490	130	link to translation equivalent	[linking to translation equivalents]	0.0	4	1	1	1
6491	130	N-gram	[N-gram]	0.0	1	1	1	1
6492	130	state of research	[state of research]	0.0	3	1	1	1
6493	130	marker	[markers]	0.0	1	1	1	1
6494	130	extractive summarization approach	[extractive summarization approaches]	0.0	3	1	1	1
6495	130	yield way	[yielding ways]	0.0	2	1	1	1
6496	130	word alignment component	[word alignment component]	0.0	3	1	1	1
6497	130	speech dialogue interaction	[speech dialogue interactions]	0.0	3	1	1	1
6498	130	field exercise with the Marine	[field exercises with the Marines]	0.0	5	1	1	1
6499	130	language model of move	[language model of moves]	0.0	4	1	1	1
6500	130	outstanding problem	[outstanding problems]	0.0	2	1	1	1
6501	130	subject of much debate	[subject of much debate]	0.0	4	1	1	1
6502	130	parallel	[parallel]	0.0	1	1	1	1
6503	130	focus on the change	[focusing on the changes]	0.0	4	1	1	1
6504	130	literal meaning of input	[literal meaning of input]	0.0	4	1	1	1
6505	130	such as size b perspective-taking	[such as size b perspective-taking]	0.0	5	1	1	1
6506	130	demonstration of a language system	[demonstration of an Language System]	0.0	5	1	1	1
6507	130	constraint on the attribute	[constraints on the attributes]	0.0	4	1	1	1
6508	130	shallow feature of question	[shallow features of questions]	0.0	4	1	1	1
6509	130	7 .5% word	[7 .5% word]	0.0	3	1	1	1
6510	130	i parsing of Korean	[i parsing of Korean]	0.0	4	1	1	1
6511	130	ellipsis	[ellipsis]	0.0	1	1	1	1
6512	130	Transfer-Driven Machine Translation	[Transfer-Driven Machine Translation]	0.0	3	1	1	1
6513	130	structural disambiguation	[structural disambiguation]	0.0	2	1	1	1
6514	130	robust recognition	[robust recognition]	0.0	2	1	1	1
6515	130	induce a grammar from corpus	[inducing a grammar from corpora]	0.0	5	1	1	1
6516	130	a attribute such as size perspective-taking	[a attributes such as size perspective-taking]	0.0	6	1	1	1
6517	130	use of prior in loglinear model	[use of priors in loglinear models]	0.0	6	1	1	1
6518	130	iv fine-grained modeling of word feature	[iv fine-grained modeling of word features]	0.0	6	1	1	1
6519	130	text collection cover the domain	[text collection covering the domain]	0.0	5	1	1	1
6520	130	Entity	[Entity]	0.0	1	1	1	1
6521	130	test against one	[tests against one]	0.0	3	1	1	1
6522	130	spelling-checker for English	[spelling-checkers for English]	0.0	3	1	1	1
6523	130	phonological knowledge base system	[phonological knowledge base system]	0.0	4	1	1	1
6524	130	different language system domain	[different language system domains]	0.0	4	1	1	1
6525	130	warfare	[warfare]	0.0	1	1	1	1
6526	130	empirical test	[empirical test]	0.0	2	1	1	1
6527	130	advantage of information	[advantage of information]	0.0	3	1	1	1
6528	130	incorporate a approximation	[incorporating an approximation]	0.0	3	1	1	1
6529	130	traditional generative phonological approach	[traditional generative phonological approach]	0.0	4	1	1	1
6530	130	class NP	[class NP]	0.0	2	1	1	1
6531	130	paradigm for human interaction	[paradigm for human interaction]	0.0	4	1	1	1
6532	130	frame semantics	[frame semantics]	0.0	2	1	1	1
6533	130	corpus a more available resource	[corpora a more available resource]	0.0	5	1	1	1
6534	130	non-English Arabic stemmer	[non-English Arabic stemmer]	0.0	3	1	1	1
6535	130	integrate q/a application	[integrating Q/A applications]	0.0	3	1	1	1
6536	130	part of the processing time	[part of the processing time]	0.0	5	1	1	1
6537	130	qa system	[QA system]	0.0	2	1	1	1
6538	130	novel automatic evaluation method	[novel automatic evaluation method]	0.0	4	1	1	1
6539	130	start with a corpus word	[starting with a corpus words]	0.0	5	1	1	1
6540	130	Schutze idea	[Schutze's idea]	0.0	2	1	1	1
6541	130	parsing method	[parsing method]	0.0	2	1	1	1
6542	130	unsupervised algorithm acquire new stem	[unsupervised algorithm acquiring new stems]	0.0	5	1	1	1
6543	130	reflect the relationship email .g rank	[reflecting the relationship e .g ranks]	0.0	6	1	1	1
6544	130	traditional phonological approach	[traditional phonological approach]	0.0	3	1	1	1
6545	130	explore bitext in a volume	[exploring bitexts in a volume]	0.0	5	1	1	1
6546	130	area during IDUS development	[areas during IDUS development]	0.0	4	1	1	1
6547	130	grammaticality at least 99% correct sentence	[grammaticality at least 99% correct sentences]	0.0	6	1	1	1
6548	130	interpolation of n-gram	[interpolation of N-grams]	0.0	3	1	1	1
6549	130	45% reduction in error	[45% reduction in error]	0.0	4	1	1	1
6550	130	efficient parsing language	[efficient parsing language]	0.0	3	1	1	1
6551	130	importance to statistical Machine Translation	[importance to Statistical Machine Translation]	0.0	5	1	1	1
6552	130	use the SemCor datum set	[using the SemCor data sets]	0.0	5	1	1	1
6553	130	natural language search engine	[natural language search engines]	0.0	4	1	1	1
6554	130	logical well-formedness condition	[logical well-formedness conditions]	0.0	3	1	1	1
6555	130	muc-3	[MUC-3]	0.0	1	1	1	1
6556	130	principle-based parser	[principle-based parser]	0.0	2	1	1	1
6557	130	also graph	[also graphs]	0.0	2	1	1	1
6558	130	use of representation	[use of representations]	0.0	3	1	1	1
6559	130	domain independent model with classifier	[domain independent models with classifiers]	0.0	5	1	1	1
6560	130	mle reestimation algorithm the algorithm	[MLE reestimation algorithms the algorithm]	0.0	5	1	1	1
6561	130	different aspect	[different aspects]	0.0	2	1	1	1
6562	130	study of anaphoric expression	[study of anaphoric expressions]	0.0	4	1	1	1
6563	130	adopt a simple paraphrasing technique	[adopting a simple paraphrasing technique]	0.0	5	1	1	1
6564	130	role in cognition	[role in cognition]	0.0	3	1	1	1
6565	130	Paramax language	[Paramax language]	0.0	2	1	1	1
6566	130	efficiency—to	[efficiency—to]	0.0	1	1	1	1
6567	130	interactive system	[interactive systems]	0.0	2	1	1	1
6568	130	antecedence of each element	[antecedence of each element]	0.0	4	1	1	1
6569	130	natural setting	[natural settings]	0.0	2	1	1	1
6570	130	induce verb class	[inducing verb classes]	0.0	3	1	1	1
6571	130	only .0%	[only .0%]	0.0	2	1	1	1
6572	130	grammar formalism such as rewrite system	[grammar formalisms such as rewriting systems]	0.0	6	1	1	1
6573	130	sentence in comparable corpus	[sentences in comparable corpora]	0.0	4	1	1	1
6574	130	research article	[research articles]	0.0	2	1	1	1
6575	130	language model score	[language model score]	0.0	3	1	1	1
6576	130	shallow feature	[shallow features]	0.0	2	1	1	1
6577	130	three-tiered approach	[three-tiered approach]	0.0	2	1	1	1
6578	130	decision tree building	[decision tree building]	0.0	3	1	1	1
6579	130	information extraction Ie	[information extraction IE]	0.0	3	1	1	1
6580	130	previous paper	[previous papers]	0.0	2	1	1	1
6581	130	automatic generation	[automatic generation]	0.0	2	1	1	1
6582	130	optical character recognition ocr	[optical character recognition OCR]	0.0	4	1	1	1
6583	130	poor construction	[poor construction]	0.0	2	1	1	1
6584	130	other index	[other indices]	0.0	2	1	1	1
6585	130	alignment model	[alignment models]	0.0	2	1	1	1
6586	130	efficient broad-coverage parser for English	[efficient broad-coverage parser for English]	0.0	5	1	1	1
6587	130	dynamic combiner with decision	[dynamic combiner with decisions]	0.0	4	1	1	1
6588	130	management	[management]	0.0	1	1	1	1
6589	130	intelligent document Understanding System	[Intelligent Document Understanding System]	0.0	4	1	1	1
6590	130	way identify predicate argument structure	[way identifying predicate argument structures]	0.0	5	1	1	1
6591	130	exploit sophisticated representation	[exploiting sophisticated representations]	0.0	3	1	1	1
6592	130	generation of hypertext link	[generation of hypertext links]	0.0	4	1	1	1
6593	130	speech datum	[speech data]	0.0	2	1	1	1
6594	130	good choice for the parser	[good choice for the parser]	0.0	5	1	1	1
6595	130	many grammatical formalism	[many grammatical formalisms]	0.0	3	1	1	1
6596	130	entity-oriented approach to parsing	[entity-oriented approach to parsing]	0.0	4	1	1	1
6597	130	fundamentally different strategy one	[fundamentally different strategies one]	0.0	4	1	1	1
6598	130	include conditioning on consecutive word	[including conditioning on consecutive words]	0.0	5	1	1	1
6599	130	Chomsky	[Chomsky's]	0.0	1	1	1	1
6600	130	paper [Zernik87]	[papers [Zernik87]]	0.0	2	1	1	1
6601	130	ambiguity packing stochastic disambiguation technique	[ambiguity packing stochastic disambiguation techniques]	0.0	5	1	1	1
6602	130	robust parsing final language	[Robust parsing final language]	0.0	4	1	1	1
6603	130	n-b speech/language integration architecture	[n-best speech/language integration architecture]	0.0	4	1	1	1
6604	130	scope information of the syntactic structure	[scope information of the syntactic structure]	0.0	6	1	1	1
6605	130	means for grounding	[means for grounding]	0.0	3	1	1	1
6606	130	acquisition of semantic constraint	[acquisition of semantic constraints]	0.0	4	1	1	1
6607	130	such redundancy	[such redundancy]	0.0	2	1	1	1
6608	130	edge with a interpretation	[edges with a interpretation]	0.0	4	1	1	1
6609	130	speaker prior to training	[speakers prior to training]	0.0	4	1	1	1
6610	130	translation datum	[translation data]	0.0	2	1	1	1
6611	130	phrase-based machine translation method	[phrase-based machine translation method]	0.0	4	1	1	1
6612	130	node structure	[node structures]	0.0	2	1	1	1
6613	130	increase in mrr	[increase in MRR]	0.0	3	1	1	1
6614	130	bilingual material	[bilingual material]	0.0	2	1	1	1
6615	130	use a speech from speaker	[using a speech from speakers]	0.0	5	1	1	1
6616	130	algorithm for phrase structure parsing	[algorithm for phrase structure parsing]	0.0	5	1	1	1
6617	130	document preference	[document preference]	0.0	2	1	1	1
6618	130	most young man like woman	[most young men like women]	0.0	5	1	1	1
6619	130	respect to the question	[respect to the question]	0.0	4	1	1	1
6620	130	internal variation in a set	[internal variation in a set]	0.0	5	1	1	1
6621	130	finite State Model	[Finite State Model]	0.0	3	1	1	1
6622	130	introspection in the hope	[introspection in the hope]	0.0	4	1	1	1
6623	130	set of model parameter	[set of model parameters]	0.0	4	1	1	1
6624	130	efficient principle-based parser	[efficient principle-based parser]	0.0	3	1	1	1
6625	130	acquisition of semantic constraint in domain	[acquisition of semantic constraints in domains]	0.0	6	1	1	1
6626	130	Listen-Communicate-Show	[Listen-Communicate-Show]	0.0	1	1	1	1
6627	130	i efficient parsing a verb	[i efficient parsing a verb]	0.0	5	1	1	1
6628	130	weighting	[weighting]	0.0	1	1	1	1
6629	130	14 .0%	[14 .0%]	0.0	2	1	1	1
6630	130	use training speaker	[using training speakers]	0.0	3	1	1	1
6631	130	Coalition	[Coalition]	0.0	1	1	1	1
6632	130	i robust parsing final language	[i Robust parsing final language]	0.0	5	1	1	1
6633	130	word sense induction wsi	[word sense induction WSI]	0.0	4	1	1	1
6634	130	consist of sentence	[consisting of sentences]	0.0	3	1	1	1
6635	130	addressee identification	[addressee identification]	0.0	2	1	1	1
6636	130	interactive Machine Translation	[Interactive Machine Translation]	0.0	3	1	1	1
6637	130	evaluation algorithm	[evaluation algorithms]	0.0	2	1	1	1
6638	130	use semantic syntactic category	[using semantic syntactic categories]	0.0	4	1	1	1
6639	130	make corpus analysis	[making corpus analysis]	0.0	3	1	1	1
6640	130	model in place	[model in place]	0.0	3	1	1	1
6641	130	mt evaluation method (BLEU	[MT evaluation methods (BLEU]	0.0	4	1	1	1
6642	130	design of the discourse understanding process	[design of the discourse understanding process]	0.0	6	1	1	1
6643	130	sentence-plan-ranker	[sentence-plan-ranker]	0.0	1	1	1	1
6644	130	formalism tree Grammar	[formalisms Tree Grammars]	0.0	3	1	1	1
6645	130	major new problem	[major new problems]	0.0	3	1	1	1
6646	130	utilize mechanism	[utilizing mechanisms]	0.0	2	1	1	1
6647	130	mutual information	[mutual information]	0.0	2	1	1	1
6648	130	story link	[story link]	0.0	2	1	1	1
6649	130	unsupervised algorithm automatically acquire stem	[unsupervised algorithm automatically acquiring stems]	0.0	5	1	1	1
6650	130	two problem	[two problems]	0.0	2	1	1	1
6651	130	various kind	[various kinds]	0.0	2	1	1	1
6652	130	token classification	[token classification]	0.0	2	1	1	1
6653	130	discover sentence in comparable corpus	[discovering sentences in comparable corpora]	0.0	5	1	1	1
6654	130	reasoning in machine translation	[reasoning in machine translation]	0.0	4	1	1	1
6655	130	parallel corpus 100 word	[parallel corpus 100 words]	0.0	4	1	1	1
6656	130	new part-of-speech tagger explicit use	[new part-of-speech tagger explicit use]	0.0	5	1	1	1
6657	130	start with a corpus	[starting with a corpus]	0.0	4	1	1	1
6658	130	such as cue phrase	[such as cue phrases]	0.0	4	1	1	1
6659	130	various training factor on performance	[various training factors on performance]	0.0	5	1	1	1
6660	130	two new word-sense system	[two new word-sense systems]	0.0	4	1	1	1
6661	130	term weighting	[term weighting]	0.0	2	1	1	1
6662	130	good tagging result	[best tagging result]	0.0	3	1	1	1
6663	130	proposal	[proposals]	0.0	1	1	1	1
6664	130	sentence-aligned corpus	[sentence-aligned corpus]	0.0	2	1	1	1
6665	130	various function	[various functions]	0.0	2	1	1	1
6666	130	least 96% paraphrase	[least 96% paraphrases]	0.0	3	1	1	1
6667	130	achieve baseline result	[achieving baseline results]	0.0	3	1	1	1
6668	130	meeting use Bayesian Network classifier	[meetings using Bayesian Network classifiers]	0.0	5	1	1	1
6669	130	probable morpheme sequence	[probable morpheme sequence]	0.0	3	1	1	1
6670	130	relationship email .g social rank	[relationship e .g social ranks]	0.0	5	1	1	1
6671	130	manual acquisition of constraint in domain	[Manual acquisition of constraints in domains]	0.0	6	1	1	1
6672	130	analyze metaphor	[analyzing metaphors]	0.0	2	1	1	1
6673	130	light of aspect	[light of aspects]	0.0	3	1	1	1
6674	130	independent extractive summarization approach	[independent extractive summarization approaches]	0.0	4	1	1	1
6675	130	generalization roughly corresponding to arc	[generalization roughly corresponding to arcs]	0.0	5	1	1	1
6676	130	Japanese sentence analysis	[Japanese sentence analyses]	0.0	3	1	1	1
6677	130	predict topic shift	[predicting topic shifts]	0.0	3	1	1	1
6678	130	apply information	[applying information]	0.0	2	1	1	1
6679	130	include conditioning on multiple word	[including conditioning on multiple words]	0.0	5	1	1	1
6680	130	probabilistic model of natural language	[probabilistic model of natural language]	0.0	5	1	1	1
6681	130	hand-crafted generation component	[hand-crafted generation component]	0.0	3	1	1	1
6682	130	adopt technique	[adopting techniques]	0.0	2	1	1	1
6683	130	posteriori estimation of markov model	[posteriori estimation of Markov models]	0.0	5	1	1	1
6684	130	character recognition ocr	[character recognition OCR]	0.0	3	1	1	1
6685	130	various feature	[various features]	0.0	2	1	1	1
6686	130	spatial description	[spatial descriptions]	0.0	2	1	1	1
6687	130	thai-english language	[Thai-English language]	0.0	2	1	1	1
6688	130	several factor include limitation	[several factors including limitations]	0.0	4	1	1	1
6689	130	thus the number	[thus the number]	0.0	3	1	1	1
6690	130	performance versus	[performance vs]	0.0	2	1	1	1
6691	130	common assumption	[common assumption]	0.0	2	1	1	1
6692	130	also graph indispensable	[also graphs indispensable]	0.0	3	1	1	1
6693	130	algorithm for the analysis	[algorithms for the analysis]	0.0	4	1	1	1
6694	130	qualitative spatial constraint	[qualitative spatial constraints]	0.0	3	1	1	1
6695	130	good previous tagging result	[best previous tagging result]	0.0	4	1	1	1
6696	130	noisy output	[noisy output]	0.0	2	1	1	1
6697	130	correlation between evaluation measure	[correlation between evaluation measures]	0.0	4	1	1	1
6698	130	use a little speech	[using a little speech]	0.0	4	1	1	1
6699	130	mission success	[mission success]	0.0	2	1	1	1
6700	130	several kind	[several kinds]	0.0	2	1	1	1
6701	130	manual acquisition of semantic constraint	[Manual acquisition of semantic constraints]	0.0	5	1	1	1
6702	130	classify SCF datum	[classifying SCF data]	0.0	3	1	1	1
6703	130	several level	[several levels]	0.0	2	1	1	1
6704	130	mathematical formalism	[mathematical formalism]	0.0	2	1	1	1
6705	130	feature of the system	[features of the system]	0.0	4	1	1	1
6706	130	graph indispensable	[graphs indispensable]	0.0	2	1	1	1
6707	130	method construct word vector .e method	[methods constructing word vectors .e methods]	0.0	6	1	1	1
6708	130	formalization of reasoning	[formalization of reasoning]	0.0	3	1	1	1
6709	130	reformulation of the sentence	[reformulation of the sentence]	0.0	4	1	1	1
6710	130	meaning of a utterance	[meaning of an utterance]	0.0	4	1	1	1
6711	130	syntax formalism	[syntax formalisms]	0.0	2	1	1	1
6712	130	support	[support]	0.0	1	1	1	1
6713	130	automatic parameter optimization	[automatic parameter optimization]	0.0	3	1	1	1
6714	130	use the reference	[using the reference]	0.0	3	1	1	1
6715	130	Dependency diagram CDD	[Dependency Diagram CDD]	0.0	3	1	1	1
6716	130	discourse phenomenon as cue phrase	[discourse phenomena as cue phrases]	0.0	5	1	1	1
6717	130	mixture of technique	[mixtures of techniques]	0.0	3	1	1	1
6718	130	rank of this parse	[ranking of these parses]	0.0	4	1	1	1
6719	130	form in the input document	[forms in the input documents]	0.0	5	1	1	1
6720	130	two output from information extraction	[two outputs from information extraction]	0.0	5	1	1	1
6721	130	sentence corpus	[sentences corpus]	0.0	2	1	1	1
6722	130	use associative processor	[using associative processors]	0.0	3	1	1	1
6723	130	type of automaton	[type of automaton]	0.0	3	1	1	1
6724	130	framework less restrictive than one	[framework less restrictive than ones]	0.0	5	1	1	1
6725	130	96% paraphrase	[96% paraphrases]	0.0	2	1	1	1
6726	130	parameter optimization	[parameter optimization]	0.0	2	1	1	1
6727	130	word-semantic information .g	[word-semantic information .g]	0.0	3	1	1	1
6728	130	nature of this sentence	[nature of these sentences]	0.0	4	1	1	1
6729	130	heuristic learning of phrase translation	[heuristic learning of phrase translations]	0.0	5	1	1	1
6730	130	expression of language	[expressions of languages]	0.0	3	1	1	1
6731	130	understanding process	[understanding process]	0.0	2	1	1	1
6732	130	method .e method	[methods .e methods]	0.0	3	1	1	1
6733	130	traditional natural language interface	[traditional natural language interfaces]	0.0	4	1	1	1
6734	130	example-retrieval	[example-retrieval]	0.0	1	1	1	1
6735	130	around 97% match accuracy	[around 97% match accuracy]	0.0	4	1	1	1
6736	130	resource management rm	[Resource Management RM]	0.0	3	1	1	1
6737	130	use feature	[using features]	0.0	2	1	1	1
6738	130	connection	[connection]	0.0	1	1	1	1
6739	130	drastic generalization of chart parsing	[drastic generalization of chart Parsing]	0.0	5	1	1	1
6740	130	reflect the relationship	[reflecting the relationship]	0.0	3	1	1	1
6741	130	researcher	[researchers]	0.0	1	1	1	1
6742	130	FERRET a interactive question-answering system	[FERRET an interactive question-answering system]	0.0	5	1	1	1
6743	130	most man like mostly woman	[most men like mostly women]	0.0	5	1	1	1
6744	130	part-of-speech tagger explicit use	[part-of-speech tagger explicit use]	0.0	4	1	1	1
6745	130	combine the result	[combining the results]	0.0	3	1	1	1
6746	130	use of expression	[use of expressions]	0.0	3	1	1	1
6747	130	parsing of Korean a verb	[parsing of Korean a verb]	0.0	5	1	1	1
6748	130	beam-search decoder	[beam-search decoder]	0.0	2	1	1	1
6749	130	efficiency of RCL parser	[efficiency of RCL parsers]	0.0	4	1	1	1
6750	130	use machine translation output	[using machine translation output]	0.0	4	1	1	1
6751	130	term of coherence	[terms of coherence]	0.0	3	1	1	1
6752	130	user's goal	[user's goals]	0.0	2	1	1	1
6753	130	research on problem include unlimited vocabulary	[research on problems including unlimited vocabulary]	0.0	6	1	1	1
6754	130	contain set	[containing sets]	0.0	2	1	1	1
6755	130	transfer system	[transfer system]	0.0	2	1	1	1
6756	130	performance threshold	[performance thresholds]	0.0	2	1	1	1
6757	130	new representation the Phrasal Lexicon	[new representation the Phrasal Lexicon]	0.0	5	1	1	1
6758	130	LHIP a system	[LHIP a system]	0.0	3	1	1	1
6759	130	system with intelligent agent	[system with intelligent agents]	0.0	4	1	1	1
6760	130	referential property	[referential properties]	0.0	2	1	1	1
6761	130	subgraph	[subgraphs]	0.0	1	1	1	1
6762	130	customizable Ie paradigm	[customizable IE paradigm]	0.0	3	1	1	1
6763	130	owe	[Owing]	0.0	1	1	1	1
6764	130	determine the vehicle	[determining the vehicle]	0.0	3	1	1	1
6765	130	data	[Data]	0.0	1	1	1	1
6766	130	performance in real application	[performance in real applications]	0.0	4	1	1	1
6767	130	a numeric-valued attribute such as size	[a numeric-valued attributes such as size]	0.0	6	1	1	1
6768	130	behavior of a grammar	[behavior of a grammar]	0.0	4	1	1	1
6769	130	translation probability	[translation probabilities]	0.0	2	1	1	1
6770	130	determine the literal meaning	[determining the literal meaning]	0.0	4	1	1	1
6771	130	small sentence corpus	[small sentences corpus]	0.0	3	1	1	1
6772	130	central problem	[central problem]	0.0	2	1	1	1
6773	130	exploit representation	[exploiting representations]	0.0	2	1	1	1
6774	130	predict new concept	[predicting new concepts]	0.0	3	1	1	1
6775	130	parameter optimization of the wsi algorithm	[parameter optimization of the WSI algorithm]	0.0	6	1	1	1
6776	130	significant improvement over word alignment technique	[Significant improvement over word alignment techniques]	0.0	6	1	1	1
6777	130	strength of antecedence recovery	[strength of antecedence recovery]	0.0	4	1	1	1
6778	130	mixture of trigram model	[mixture of trigram models]	0.0	4	1	1	1
6779	130	variation in a set	[variation in a set]	0.0	4	1	1	1
6780	130	use criterion	[using criterion]	0.0	2	1	1	1
6781	130	recent speculative claim	[recent speculative claims]	0.0	3	1	1	1
6782	130	ground	[ground]	0.0	1	1	1	1
6783	130	phonological system	[phonological system]	0.0	2	1	1	1
6784	130	use triplet	[using triplets]	0.0	2	1	1	1
6785	130	2-character name	[2-character names]	0.0	2	1	1	1
6786	130	word block	[word blocks]	0.0	2	1	1	1
6787	130	corpus consist of sentence	[corpus consisting of sentences]	0.0	4	1	1	1
6788	130	accomplishment of MADCOW	[accomplishments of MADCOW]	0.0	3	1	1	1
6789	130	linguistic expressiveness of the formalism	[linguistic expressiveness of the formalisms]	0.0	5	1	1	1
6790	130	m-th order	[m-th order]	0.0	2	1	1	1
6791	130	numerous non-literal aspect of communication	[numerous non-literal aspects of communication]	0.0	5	1	1	1
6792	130	maximally plausible interpretation	[maximally plausible interpretation]	0.0	3	1	1	1
6793	130	metaphor mapping	[metaphor mappings]	0.0	2	1	1	1
6794	130	grammaticality correct sentence	[grammaticality correct sentences]	0.0	3	1	1	1
6795	130	frequently predictor	[frequently predictors]	0.0	2	1	1	1
6796	130	experimental result on all application	[experimental results on all applications]	0.0	5	1	1	1
6797	130	discourse task	[discourse task]	0.0	2	1	1	1
6798	130	associative processor	[associative processors]	0.0	2	1	1	1
6799	130	naturally sound dialog system	[naturally sounding dialog systems]	0.0	4	1	1	1
6800	130	pair of probabilistic grammar work	[pairs of probabilistic grammars working]	0.0	5	1	1	1
6801	130	island-based parsing method	[island-based parsing method]	0.0	3	1	1	1
6802	130	cognition	[cognition]	0.0	1	1	1	1
6803	130	subset only .0%	[subset only .0%]	0.0	3	1	1	1
6804	130	number of experiment	[number of experiments]	0.0	3	1	1	1
6805	130	small amount	[small amount]	0.0	2	1	1	1
6806	130	boolean	[Boolean]	0.0	1	1	1	1
6807	130	offer advantage	[Offering advantages]	0.0	2	1	1	1
6808	130	SUN 4	[SUN 4]	0.0	2	1	1	1
6809	130	m-component mixture	[m-component mixture]	0.0	2	1	1	1
6810	130	likely answer candidate	[likely answer candidates]	0.0	3	1	1	1
6811	130	user modeling component	[user modeling component]	0.0	3	1	1	1
6812	130	corpus ,000 word	[corpus ,000 words]	0.0	3	1	1	1
6813	130	conventional algorithm	[conventional algorithms]	0.0	2	1	1	1
6814	130	six extract	[six extracts]	0.0	2	1	1	1
6815	130	reference of the pronoun	[references of the pronoun]	0.0	4	1	1	1
6816	130	layout structure	[layout structures]	0.0	2	1	1	1
6817	130	TDMT Transfer-Driven Machine Translation	[TDMT Transfer-Driven Machine Translation]	0.0	4	1	1	1
6818	130	new SI corpus	[new SI corpus]	0.0	3	1	1	1
6819	130	automatically gather a number of abstract	[automatically gathering a number of abstracts]	0.0	6	1	1	1
6820	130	property of tree-adjoining grammar TAG	[properties of tree-adjoining grammars TAG]	0.0	5	1	1	1
6821	130	constraint in domain	[constraints in domains]	0.0	3	1	1	1
6822	130	different form	[different forms]	0.0	2	1	1	1
6823	130	speaker-independent SI training	[speaker-independent SI training]	0.0	3	1	1	1
6824	130	word-sense system one	[word-sense systems one]	0.0	3	1	1	1
6825	130	rewrite	[rewriting]	0.0	1	1	1	1
6826	130	perspective-taking in reference	[perspective-taking in reference]	0.0	3	1	1	1
6827	130	understand accuracy	[understanding accuracy]	0.0	2	1	1	1
6828	130	morpheme sequence	[morpheme sequence]	0.0	2	1	1	1
6829	130	iii Rapid system development to domain	[iii Rapid system development to domains]	0.0	6	1	1	1
6830	130	corresponding to the LM	[corresponding to the LM]	0.0	4	1	1	1
6831	130	semantic representation	[semantic representation]	0.0	2	1	1	1
6832	130	develop the grammar	[developing the grammars]	0.0	3	1	1	1
6833	130	precision over text	[precision over text]	0.0	3	1	1	1
6834	130	state-of-the-art evaluation measure	[state-of-the-art evaluation measures]	0.0	3	1	1	1
6835	130	different strategy utilize primarily knowledge-based mechanism	[different strategies utilizing primarily knowledge-based mechanisms]	0.0	6	1	1	1
6836	130	take runtime	[taking runtime]	0.0	2	1	1	1
6837	130	several key factor behind this finding	[several key factors behind this finding]	0.0	6	1	1	1
6838	130	Natural language processing system	[Natural Language Processing systems]	0.0	4	1	1	1
6839	130	consecutive word	[consecutive words]	0.0	2	1	1	1
6840	130	broad use of lexical feature	[broad use of lexical features]	0.0	5	1	1	1
6841	130	vocabulary size	[vocabulary size]	0.0	2	1	1	1
6842	130	combiner	[combiner]	0.0	1	1	1	1
6843	130	application of inference rule	[Applications of inference rules]	0.0	4	1	1	1
6844	130	use the language model	[using the language model]	0.0	4	1	1	1
6845	130	easy exploitation of domain semantics	[easy exploitation of domain semantics]	0.0	5	1	1	1
6846	130	use 40 utterance	[Using 40 utterances]	0.0	3	1	1	1
6847	130	automatically evaluate answer	[automatically evaluating answers]	0.0	3	1	1	1
6848	130	patent claim	[patent claim]	0.0	2	1	1	1
6849	130	enter through message passing between actor	[entering through message passing between actors]	0.0	6	1	1	1
6850	130	7% improvement	[7% improvement]	0.0	2	1	1	1
6851	130	subcategorization frame SCF distribution	[subcategorization frame SCF distributions]	0.0	4	1	1	1
6852	130	undisambiguated SCF datum	[undisambiguated SCF data]	0.0	3	1	1	1
6853	130	different answer agent	[different answering agents]	0.0	3	1	1	1
6854	130	user modeling system	[user modeling systems]	0.0	3	1	1	1
6855	130	structure finding method	[structure finding methods]	0.0	3	1	1	1
6856	130	time decoding algorithm	[time decoding algorithm]	0.0	3	1	1	1
6857	130	inflow of multimedia datum	[inflow of multimedia data]	0.0	4	1	1	1
6858	130	system reasoning	[system reasoning]	0.0	2	1	1	1
6859	130	advantage of the symmetric pattern	[advantage of the symmetric patterns]	0.0	5	1	1	1
6860	130	English coordinate sentence	[English coordinate sentences]	0.0	3	1	1	1
6861	130	grammaticality 99% sentence equivalence	[grammaticality 99% sentences equivalence]	0.0	4	1	1	1
6862	130	pronominal anaphora resolution	[pronominal anaphora resolution]	0.0	3	1	1	1
6863	130	careful examination of the complexity	[careful examination of the complexity]	0.0	5	1	1	1
6864	130	implication automate aspect of language learning	[Implications automating aspects of language learning]	0.0	6	1	1	1
6865	130	current language interface	[Current language interfaces]	0.0	3	1	1	1
6866	130	one superiority	[one superiority]	0.0	2	1	1	1
6867	130	aspect such as communication procedure	[aspects such as communication procedures]	0.0	5	1	1	1
6868	130	use a argumentation system	[using an argumentation system]	0.0	4	1	1	1
6869	130	natural language understanding	[natural language understanding]	0.0	3	1	1	1
6870	130	contain set of feature	[containing sets of features]	0.0	4	1	1	1
6871	130	ONTOSCORE	[ONTOSCORE]	0.0	1	1	1	1
6872	130	unmodified subgraph	[unmodified subgraphs]	0.0	2	1	1	1
6873	130	pick the hypothesis corresponding	[picking the hypothesis corresponding]	0.0	4	1	1	1
6874	130	great deal	[great deal]	0.0	2	1	1	1
6875	130	cooperative response adaptive	[cooperative responses adaptive]	0.0	3	1	1	1
6876	130	real-time response	[real-time response]	0.0	2	1	1	1
6877	130	contribute to the word	[contributing to the word]	0.0	4	1	1	1
6878	130	New result on all application	[New results on all applications]	0.0	5	1	1	1
6879	130	efficient parsing a verb final language	[efficient parsing a verb final language]	0.0	6	1	1	1
6880	130	wsd system	[WSD systems]	0.0	2	1	1	1
6881	130	approach to web-based computer-assisted writing	[approach to Web-based computer-assisted writing]	0.0	5	1	1	1
6882	130	information in memory	[information in memory]	0.0	3	1	1	1
6883	130	edge with a semantic interpretation	[edges with a semantic interpretation]	0.0	5	1	1	1
6884	130	ocr accuracy	[OCR accuracy]	0.0	2	1	1	1
6885	130	Marine	[Marines]	0.0	1	1	1	1
6886	130	contrast to the approach	[contrast to the approach]	0.0	4	1	1	1
6887	130	argument structure	[argument structures]	0.0	2	1	1	1
6888	130	use memory	[using memory]	0.0	2	1	1	1
6889	130	Hierarchical Acyclic Graph HDAG Kernel	[Hierarchical Acyclic Graph HDAG Kernel]	0.0	5	1	1	1
6890	130	expensive part	[expensive part]	0.0	2	1	1	1
6891	130	find datum relevant	[finding data relevant]	0.0	3	1	1	1
6892	130	period	[periods]	0.0	1	1	1	1
6893	130	combine language model	[combining language models]	0.0	3	1	1	1
6894	130	unique property of tree-adjoining grammar	[unique properties of tree-adjoining grammars]	0.0	5	1	1	1
6895	130	restriction	[restrictions]	0.0	1	1	1	1
6896	130	intelligibility	[intelligibility]	0.0	1	1	1	1
6897	130	mixed-initiative speech dialogue interaction	[mixed-initiative speech dialogue interactions]	0.0	4	1	1	1
6898	130	perform inference in semantic network	[performing inference in semantic networks]	0.0	5	1	1	1
6899	130	novel Model for language understanding	[novel Model for language understanding]	0.0	5	1	1	1
6900	130	exist metric	[existing metrics]	0.0	2	1	1	1
6901	130	empirical comparison	[empirical comparison]	0.0	2	1	1	1
6902	130	feature set	[feature sets]	0.0	2	1	1	1
6903	130	simple interpolation method	[simple interpolation methods]	0.0	3	1	1	1
6904	130	search for answer in corpus	[searching for answers in corpora]	0.0	5	1	1	1
6905	130	phrase structure parsing of language	[phrase structure parsing of language]	0.0	5	1	1	1
6906	130	system development to domain via acquisition	[system development to domains via acquisition]	0.0	6	1	1	1
6907	130	biase	[biasing]	0.0	1	1	1	1
6908	130	determine the literal meaning of input	[determining the literal meaning of input]	0.0	6	1	1	1
6909	130	type of transfer system	[type of transfer system]	0.0	4	1	1	1
6910	130	three word from alignment model	[three words from alignment models]	0.0	5	1	1	1
6911	130	basis for a tool	[basis for a tool]	0.0	4	1	1	1
6912	130	machine-readable dictionary	[machine-readable dictionaries]	0.0	2	1	1	1
6913	130	possible word-sense of word	[possible word-senses of words]	0.0	4	1	1	1
6914	130	500 feature	[500 features]	0.0	2	1	1	1
6915	130	simple set of model parameter	[simpler set of model parameters]	0.0	5	1	1	1
6916	130	new performance enhancing technique	[new performance enhancing techniques]	0.0	4	1	1	1
6917	130	theory for natural language presupposition	[theory for natural language presuppositions]	0.0	5	1	1	1
6918	130	i robust efficient parsing	[i Robust efficient parsing]	0.0	4	1	1	1
6919	130	formalism tree adjoin Grammar	[formalisms Tree Adjoining Grammars]	0.0	4	1	1	1
6920	130	judgment	[judgment]	0.0	1	3	3	1
6921	130	aspect of the interaction requirement	[aspects of the interaction requirements]	0.0	5	1	1	1
6922	130	system response	[system response]	0.0	2	1	1	1
6923	130	suppress the quantifier most	[suppressing the quantifier most]	0.0	4	1	1	1
6924	130	rarely in the context	[rarely in the context]	0.0	4	1	1	1
6925	130	syntax-based statistical machine translation	[Syntax-based statistical machine translation]	0.0	4	2	2	1
6926	130	use evaluate the summarization quality	[use evaluating the summarization quality]	0.0	5	1	1	1
6927	130	discover sentence in comparable non-parallel corpus	[discovering sentences in comparable non-parallel corpora]	0.0	6	1	1	1
6928	130	speed-up over a SUN 4	[speed-up over a SUN 4]	0.0	5	1	1	1
6929	130	identification of name without title	[identification of names without title]	0.0	5	1	1	1
6930	130	robust parsing a verb	[Robust parsing a verb]	0.0	4	1	1	1
6931	130	substring	[substring]	0.0	1	1	1	1
6932	130	interesting combination of hand-built resource	[interesting combination of hand-built resources]	0.0	5	1	1	1
6933	130	semantic verb class	[semantic verb classes]	0.0	3	1	1	1
6934	130	quality translation via word sense	[quality translation via word sense]	0.0	5	1	1	1
6935	130	accomplishment monitor the collection	[accomplishments monitoring the collection]	0.0	4	1	1	1
6936	130	bootstrapping approach to Entity NE	[bootstrapping approach to Entity NE]	0.0	5	1	1	1
6937	130	validity of complex linguistic database	[validity of complex linguistic databases]	0.0	5	1	1	1
6938	130	new tagger explicit use	[new tagger explicit use]	0.0	4	1	1	1
6939	130	two contribution to vocabulary	[two contributions to vocabulary]	0.0	4	1	1	1
6940	130	model for output selection	[model for output selection]	0.0	4	1	1	1
6941	130	evaluation scheme	[evaluation scheme]	0.0	2	1	1	1
6942	130	mobile agent	[mobile agent]	0.0	2	2	2	1
6943	130	advance in automatic speech Recognition technology	[advances in Automatic Speech Recognition technology]	0.0	6	1	1	1
6944	130	impact on model	[impact on models]	0.0	3	1	1	1
6945	130	major problem	[major problems]	0.0	2	1	1	1
6946	130	apply evaluation technique	[applying evaluation techniques]	0.0	3	1	1	1
6947	130	make analysis robust---an scheduling parser	[making analysis robust---an scheduling parser]	0.0	5	1	1	1
6948	130	generalization roughly corresponding	[generalization roughly corresponding]	0.0	3	1	1	1
6949	130	transfer phase	[transfer phase]	0.0	2	1	1	1
6950	130	probabilistic model of language	[probabilistic model of language]	0.0	4	1	1	1
6951	130	major problem include unlimited vocabulary	[major problems including unlimited vocabulary]	0.0	5	1	1	1
6952	130	96% correct paraphrase mean equivalence	[96% correct paraphrases meaning equivalence]	0.0	5	1	1	1
6953	130	lexica	[lexica]	0.0	1	1	1	1
6954	130	cover the specific domain	[covering the specific domain]	0.0	4	1	1	1
6955	130	small parallel corpus	[small parallel corpus]	0.0	3	1	1	1
6956	130	result of a evaluation	[results of a evaluation]	0.0	4	1	1	1
6957	130	processor	[processors]	0.0	1	1	1	1
6958	130	factor include limitation	[factors including limitations]	0.0	3	1	1	1
6959	130	use this model in combination	[using these models in combination]	0.0	5	1	1	1
6960	130	disjoint phrase	[disjoint phrases]	0.0	2	1	1	1
6961	130	mutli-document summarization approach	[mutli-document summarization approaches]	0.0	3	1	1	1
6962	130	paraphrasing technique	[paraphrasing technique]	0.0	2	1	1	1
6963	130	word order	[word order]	0.0	2	1	1	1
6964	130	style perform inference in semantic network	[styles performing inference in semantic networks]	0.0	6	1	1	1
6965	130	collection of utterance	[collection of utterances]	0.0	3	1	1	1
6966	130	phrasal SMT	[phrasal SMT]	0.0	2	1	1	1
6967	130	novel method discover sentence	[novel method discovering sentences]	0.0	4	1	1	1
6968	130	careful examination	[careful examination]	0.0	2	1	1	1
6969	130	correspondence	[correspondences]	0.0	1	1	1	1
6970	130	comparable corpus	[comparable corpora]	0.0	2	1	1	1
6971	130	dimensionality space	[dimensionality space]	0.0	2	1	1	1
6972	130	three method lsa-based method	[three methods LSA-based methods]	0.0	4	1	1	1
6973	130	starting point	[starting point]	0.0	2	1	1	1
6974	130	sense of much	[sense of much]	0.0	3	1	1	1
6975	130	basis for a disambiguation tool	[basis for a disambiguation tool]	0.0	5	1	1	1
6976	130	so many structure	[so many structures]	0.0	3	1	1	1
6977	130	context language	[context languages]	0.0	2	1	1	1
6978	130	use processor for real-time language translation	[using processors for real-time language translation]	0.0	6	1	1	1
6979	130	expert system	[expert systems]	0.0	2	1	1	1
6980	130	bilingual parallel corpus a resource	[bilingual parallel corpora a resource]	0.0	5	1	1	1
6981	130	validity of linguistic database	[validity of linguistic databases]	0.0	4	1	1	1
6982	130	such case-based mt	[such case-based MT]	0.0	3	1	1	1
6983	130	domain acoustic model	[domain acoustic models]	0.0	3	1	1	1
6984	130	dimension skill level to the system	[dimensions skill level to the system]	0.0	6	1	1	1
6985	130	SENSEVAL-2 sample task	[SENSEVAL-2 sample task]	0.0	3	1	1	1
6986	130	exact match	[exact match]	0.0	2	1	1	1
6987	130	problem with the approach	[problems with the approach]	0.0	4	1	1	1
6988	130	bootstrapping approach to Entity	[bootstrapping approach to Entity]	0.0	4	1	1	1
6989	130	rank of the referent	[ranks of the referents]	0.0	4	1	1	1
6990	130	result along with a analysis	[results along with a analysis]	0.0	5	1	1	1
6991	130	entropy classifier	[entropy classifier]	0.0	2	1	1	1
6992	130	multi engine machine	[multi engine machine]	0.0	3	1	1	1
6993	130	parsing final language with case marker	[parsing final language with case markers]	0.0	6	1	1	1
6994	130	English phrase	[English phrases]	0.0	2	1	1	1
6995	130	.1%	[.1%]	0.0	1	1	1	1
6996	130	correlation between the evaluation	[correlation between the evaluation]	0.0	4	1	1	1
6997	130	face-to-face meeting	[face-to-face meetings]	0.0	2	1	1	1
6998	130	set of task	[set of tasks]	0.0	3	1	1	1
6999	130	draw a space	[drawing a space]	0.0	3	1	1	1
7000	130	biase parsing	[biasing parsing]	0.0	2	1	1	1
7001	130	integral process in human understanding	[integral process in human understanding]	0.0	5	1	1	1
7002	130	parsing flexibility	[parsing flexibilities]	0.0	2	1	1	1
7003	130	overview	[overview]	0.0	1	2	2	1
7004	130	language environment	[language environment]	0.0	2	1	1	1
7005	130	over a third	[over a third]	0.0	3	1	1	1
7006	130	relational database	[relational database]	0.0	2	1	1	1
7007	130	evaluation of this method	[evaluation of this method]	0.0	4	1	1	1
7008	130	corresponding in a chart	[corresponding in a chart]	0.0	4	1	1	1
7009	130	accuracy superior to any	[accuracy superior to any]	0.0	4	1	1	1
7010	130	span edge	[spanning edges]	0.0	2	1	1	1
7011	130	domain-independent capability	[domain-independent capabilities]	0.0	2	1	1	1
7012	130	parse forest	[parse forests]	0.0	2	1	1	1
7013	130	-speak analyst	[-speaking analyst]	0.0	2	1	1	1
7014	130	.3%	[.3%]	0.0	1	1	1	1
7015	130	evaluate answer	[evaluating answers]	0.0	2	1	1	1
7016	130	common evaluation of speech	[common evaluation of speech]	0.0	4	1	1	1
7017	130	probabilistic parser	[probabilistic parser]	0.0	2	1	1	1
7018	130	indexing accord to character bigram	[indexing according to character bigrams]	0.0	5	1	1	1
7019	130	direct application of metric	[direct application of metrics]	0.0	4	1	1	1
7020	130	verbal interaction of dialog	[verbal interactions of dialogs]	0.0	4	1	1	1
7021	130	valuable method of language interface	[valuable methods of language interfaces]	0.0	5	1	1	1
7022	130	hypothesis corresponding with the confidence	[hypothesis corresponding with the confidence]	0.0	5	1	1	1
7023	130	instead of the practice	[instead of the practice]	0.0	4	1	1	1
7024	130	parse a set	[parsing a set]	0.0	3	1	1	1
7025	130	automatically training	[automatically training]	0.0	2	1	1	1
7026	130	start with a corpus ,000 word	[starting with a corpus ,000 words]	0.0	6	1	1	1
7027	130	high performance in application	[high performance in applications]	0.0	4	1	1	1
7028	130	personal name	[personal names]	0.0	2	1	1	1
7029	130	two new word-sense system one	[two new word-sense systems one]	0.0	5	1	1	1
7030	130	perform inference in network	[performing inference in networks]	0.0	4	1	1	1
7031	130	improve system performance	[improving system performance]	0.0	3	1	1	1
7032	130	phonological approach	[phonological approach]	0.0	2	1	1	1
7033	130	.3m	[.3M]	0.0	1	1	1	1
7034	130	novel cooperative Model	[novel Cooperative Model]	0.0	3	1	1	1
7035	130	good indicator	[better indicators]	0.0	2	1	1	1
7036	130	complicated large knowledgebase	[complicated large knowledgebases]	0.0	3	1	1	1
7037	130	bad coverage	[worse coverage]	0.0	2	1	1	1
7038	130	knowledge about dialog schemata	[knowledge about dialog schemata]	0.0	4	1	1	1
7039	130	syntactic analysis more robust---an scheduling parser	[syntactic analysis more robust---an scheduling parser]	0.0	6	1	1	1
7040	130	fact transition graph	[fact transition graphs]	0.0	3	1	1	1
7041	130	approach via the introspection	[approach via the introspection]	0.0	4	1	1	1
7042	130	discourse effect	[discourse effect]	0.0	2	1	1	1
7043	130	set of word alignment	[set of word alignments]	0.0	4	1	1	1
7044	130	lexical cohesion-based approach	[lexical cohesion-based approach]	0.0	3	1	1	1
7045	130	several extension of this idea	[Several extensions of this idea]	0.0	5	1	1	1
7046	130	.4m	[.4M]	0.0	1	1	1	1
7047	130	behavioral specification	[behavioral specification]	0.0	2	1	1	1
7048	130	other clue such as editing term	[Other clues such as editing terms]	0.0	6	1	1	1
7049	130	recall in the task	[recall in the task]	0.0	4	1	1	1
7050	130	new tagger	[new tagger]	0.0	2	1	1	1
7051	130	use a word alignment	[using an word alignment]	0.0	4	1	1	1
7052	130	large inflow	[large inflow]	0.0	2	1	1	1
7053	130	usual introspection	[usual introspection]	0.0	2	1	1	1
7054	130	answering	[Answering]	0.0	1	1	1	1
7055	130	combinatorics of indexation	[combinatorics of indexation]	0.0	3	1	1	1
7056	130	phrase re-ordering	[phrase re-ordering]	0.0	2	1	1	1
7057	130	learning methodology applicable	[learning methodology applicable]	0.0	3	1	1	1
7058	130	method add annotation to corpus	[method adding annotation to corpora]	0.0	5	1	1	1
7059	130	statistical learning	[Statistical Learning]	0.0	2	1	1	1
7060	130	example similar to a input expression	[examples similar to an input expression]	0.0	6	1	1	1
7061	130	direct application	[direct application]	0.0	2	1	1	1
7062	130	theory for language presupposition 1988/	[theory for language presuppositions 1988/]	0.0	5	1	1	1
7063	130	format for meaning	[format for meaning]	0.0	3	1	1	1
7064	130	segment of actual description	[segments of actual descriptions]	0.0	4	1	1	1
7065	130	typical kind	[typical kinds]	0.0	2	1	1	1
7066	130	logical expression	[logical expression]	0.0	2	1	1	1
7067	130	include inherent limitation	[including inherent limitations]	0.0	3	1	1	1
7068	130	combination of structure string	[combination of structures strings]	0.0	4	1	1	1
7069	130	exist robust parsing model	[existing robust parsing models]	0.0	4	1	1	1
7070	130	miscommunication reference problem	[miscommunication reference problems]	0.0	3	1	1	1
7071	130	output from information extraction Ie	[outputs from information extraction IE]	0.0	5	1	1	1
7072	130	reflect the social relationship rank	[reflecting the social relationship ranks]	0.0	5	1	1	1
7073	130	speech processing	[speech processing]	0.0	2	1	1	1
7074	130	grammaticality correct sentence equivalence in meaning	[grammaticality correct sentences equivalence in meaning]	0.0	6	1	1	1
7075	130	feature email .g	[features e .g]	0.0	3	1	1	1
7076	130	general-purpose nlp component	[general-purpose NLP components]	0.0	3	1	1	1
7077	130	application to the representation	[Applications to the representation]	0.0	4	1	1	1
7078	130	5K vocabulary	[5K vocabulary]	0.0	2	1	1	1
7079	130	assumption in form	[assumption in form]	0.0	3	1	1	1
7080	130	solution to word sense induction	[solution to word sense induction]	0.0	5	1	1	1
7081	130	Senseval-3 Chinese sample task	[Senseval-3 Chinese sample task]	0.0	4	1	1	1
7082	130	relation to grammar	[relation to grammar]	0.0	3	1	1	1
7083	130	language for a need	[language for a need]	0.0	4	1	1	1
7084	130	evaluation of speech	[evaluation of speech]	0.0	3	1	1	1
7085	130	case-based mt cbmt	[case-based MT CBMT]	0.0	3	1	1	1
7086	130	parsing of Korean language	[parsing of Korean language]	0.0	4	1	1	1
7087	130	employ n-gram model for key prediction	[employing n-gram models for key prediction]	0.0	6	1	1	1
7088	130	gather a large number	[gathering a large number]	0.0	4	1	1	1
7089	130	roughly corresponding	[roughly corresponding]	0.0	2	1	1	1
7090	130	translation of the word	[translation of the words]	0.0	4	1	1	1
7091	130	recognition of extra-grammatical input	[recognition of extra-grammatical input]	0.0	4	1	1	1
7092	130	aspect of a parse tree	[aspects of a parse tree]	0.0	5	1	1	1
7093	130	parsing-as-deduction in a resource	[parsing-as-deduction in a resource]	0.0	4	1	1	1
7094	130	kind of phrase	[kinds of phrases]	0.0	3	1	1	1
7095	130	a attribute such as size	[a attributes such as size]	0.0	5	1	1	1
7096	130	aspect of the performance	[aspects of the performance]	0.0	4	1	1	1
7097	130	computational theory	[computational theory]	0.0	2	1	1	1
7098	130	iv fine-grained modeling	[iv fine-grained modeling]	0.0	3	1	1	1
7099	130	effect of feature	[effect of features]	0.0	3	1	1	1
7100	130	tagger use of tag context	[tagger use of tag contexts]	0.0	5	1	1	1
7101	130	small corpus ,000 word	[small corpus ,000 words]	0.0	4	1	1	1
7102	130	110 ,000 word	[110 ,000 words]	0.0	3	1	1	1
7103	130	corpus of translation	[corpus of translations]	0.0	3	1	1	1
7104	130	model of machine translation	[model of machine translation]	0.0	4	1	1	1
7105	130	string comparison method	[string comparison methods]	0.0	3	1	1	1
7106	130	automatic translation of language	[automatic translation of language]	0.0	4	1	1	1
7107	130	clause in a program	[clauses in a program]	0.0	4	1	1	1
7108	130	use in word n-gram	[use in word n-grams]	0.0	4	1	1	1
7109	130	efficiency gain for the new algorithm	[efficiency gains for the new algorithm]	0.0	6	1	1	1
7110	130	basic of UCG	[basics of UCG]	0.0	3	1	1	1
7111	130	interpolation method like interpolation	[interpolation methods like interpolation]	0.0	4	1	1	1
7112	130	determine the literal meaning from user	[determining the literal meaning from users]	0.0	6	1	1	1
7113	130	numerous aspect such as communication procedure	[numerous aspects such as communication procedures]	0.0	6	1	1	1
7114	130	numeric-valued attribute such as size perspective-taking	[numeric-valued attributes such as size perspective-taking]	0.0	6	1	1	1
7115	130	system for grammar development	[system for grammar development]	0.0	4	1	1	1
7116	130	fine-grained modeling of word feature	[fine-grained modeling of word features]	0.0	5	1	1	1
7117	130	accord to need	[according to needs]	0.0	3	1	1	1
7118	130	word alignment technique	[word alignment techniques]	0.0	3	1	1	1
7119	130	generic mathematical formalism	[generic mathematical formalism]	0.0	3	1	1	1
7120	130	.284 srh	[.284 SRHs]	0.0	2	1	1	1
7121	130	performance for representation	[performance for representations]	0.0	3	1	1	1
7122	130	standard translation task	[standard translation task]	0.0	3	1	1	1
7123	130	enumerate all indexing	[enumerating all indexings]	0.0	3	1	1	1
7124	130	verbal interaction with knowledge	[verbal interaction with knowledge]	0.0	4	1	1	1
7125	130	Markov model	[Markov models]	0.0	2	1	1	1
7126	130	collocation observation	[collocation observation]	0.0	2	2	2	1
7127	130	new problem include unlimited vocabulary	[new problems including unlimited vocabulary]	0.0	5	1	1	1
7128	130	optimal indexation algorithm	[optimal indexation algorithm]	0.0	3	1	1	1
7129	130	use of tag context	[use of tag contexts]	0.0	4	1	1	1
7130	130	candidate for understanding result	[candidates for understanding results]	0.0	4	1	1	1
7131	130	symbolic resource	[symbolic resources]	0.0	2	1	1	1
7132	130	compete	[competing]	0.0	1	1	1	1
7133	130	free indexation algorithm	[free indexation algorithm]	0.0	3	1	1	1
7134	130	extensional equivalence of concept	[extensional equivalence of concepts]	0.0	4	1	1	1
7135	130	use a state-of-the-art Chinese word sense	[Using a state-of-the-art Chinese word sense]	0.0	6	1	1	1
7136	130	ne .g he/she/man/woman	[NE .g he/she/man/woman]	0.0	3	1	1	1
7137	130	prior	[priors]	0.0	1	1	1	1
7138	130	contain ,449 word token	[containing ,449 word tokens]	0.0	4	1	1	1
7139	130	identification of Chinese name	[identification of Chinese names]	0.0	4	1	1	1
7140	130	new representation	[new representation]	0.0	2	1	1	1
7141	130	different strategy one utilize mechanism	[different strategies one utilizing mechanisms]	0.0	5	1	1	1
7142	130	Role	[Role]	0.0	1	1	1	1
7143	130	new system	[new systems]	0.0	2	1	1	1
7144	130	final language	[final language]	0.0	2	1	1	1
7145	130	first-order class grammar	[first-order class grammar]	0.0	3	1	1	1
7146	130	approach to academic writing	[approach to academic writing]	0.0	4	1	1	1
7147	130	context feature	[context features]	0.0	2	1	1	1
7148	130	model parameter than model	[model parameters than models]	0.0	4	1	1	1
7149	130	corpus word	[corpus words]	0.0	2	1	1	1
7150	130	Educational Research method	[Educational Research Methods]	0.0	3	1	1	1
7151	130	validity of complex database	[validity of complex databases]	0.0	4	1	1	1
7152	130	grammar development	[grammar development]	0.0	2	1	1	1
7153	130	meta-inference	[meta-inference]	0.0	1	1	1	1
7154	130	spectral mapping	[spectral mapping]	0.0	2	1	1	1
7155	130	respects	[respects]	0.0	1	1	1	1
7156	130	straight c code	[straight C code]	0.0	3	1	1	1
7157	130	apply a set to entry	[applying a set to entries]	0.0	5	1	1	1
7158	130	analysis for a large class	[analysis for a large class]	0.0	5	1	1	1
7159	130	post-process the output	[post-processing the output]	0.0	3	1	1	1
7160	130	individual theory	[individual theories]	0.0	2	1	1	1
7161	130	tag context	[tag contexts]	0.0	2	1	1	1
7162	130	more frequently predictor of document descriptor	[more frequently predictors of document descriptors]	0.0	6	1	1	1
7163	130	algorithm automatically acquire new stem	[algorithm automatically acquiring new stems]	0.0	5	1	1	1
7164	130	systematic pattern in translation datum	[systematic patterns in translation data]	0.0	5	1	1	1
7165	130	block bigram model	[block bigram model]	0.0	3	1	1	1
7166	130	char t	[char t]	0.0	2	1	1	1
7167	130	improvement over traditional word alignment technique	[improvement over traditional word alignment techniques]	0.0	6	1	1	1
7168	130	wsd evaluation methodology	[WSD evaluation methodology]	0.0	3	1	1	1
7169	130	word corpus	[word corpus]	0.0	2	1	1	1
7170	130	spatial constraint from the text	[spatial constraints from the text]	0.0	5	1	1	1
7171	130	area of natural language processing	[areas of natural language processing]	0.0	5	1	1	1
7172	130	many nlp problem	[many NLP problems]	0.0	3	1	1	1
7173	130	separate component	[separate components]	0.0	2	1	1	1
7174	130	search engine	[search engine]	0.0	2	1	1	1
7175	130	inductive decision tree learning	[inductive decision tree learning]	0.0	4	1	1	1
7176	130	syllogistic reasoning	[Syllogistic reasoning]	0.0	2	1	1	1
7177	130	99% accuracy in language identification	[99% accuracy in language identification]	0.0	5	1	1	1
7178	130	attribute sequence	[attribute sequences]	0.0	2	1	1	1
7179	130	ATIS data COllection	[ATIS Data COllection]	0.0	3	1	1	1
7180	130	Lucy a English understanding system	[Lucy a English understanding system]	0.0	5	1	1	1
7181	130	also a drastic generalization corresponding	[also a drastic generalization corresponding]	0.0	5	1	1	1
7182	130	efficiency gain for the algorithm	[efficiency gains for the algorithm]	0.0	5	1	1	1
7183	130	speech/language integration architecture	[speech/language integration architecture]	0.0	3	1	1	1
7184	130	recognition of input	[recognition of input]	0.0	3	1	1	1
7185	130	individually good predictor of document descriptor	[individually good predictors of document descriptors]	0.0	6	1	1	1
7186	130	gather a large number of abstract	[gathering a large number of abstracts]	0.0	6	1	1	1
7187	130	prototype concordancer	[prototype concordancer]	0.0	2	1	1	1
7188	130	WI system	[WI systems]	0.0	2	1	1	1
7189	130	bear on reasoning	[bearing on reasoning]	0.0	3	1	1	1
7190	130	protocol analysis	[protocol analysis]	0.0	2	1	1	1
7191	130	tdmt the most dominant part	[TDMT the most dominant part]	0.0	5	1	1	1
7192	130	f-measure a decrease in f-measure error	[F-measure a decrease in F-measure error]	0.0	6	1	1	1
7193	130	such as communication procedure	[such as communication procedures]	0.0	4	1	1	1
7194	130	temporary belief	[temporary belief]	0.0	2	1	1	1
7195	130	Dependency diagram	[Dependency Diagram]	0.0	2	1	1	1
7196	130	Schutze idea of evaluation	[Schutze's idea of evaluation]	0.0	4	1	1	1
7197	130	crucial issue	[crucial issue]	0.0	2	1	1	1
7198	130	methodology automatically train spot	[methodology automatically training SPoT]	0.0	4	1	1	1
7199	130	make analysis more robust---an scheduling parser	[making analysis more robust---an scheduling parser]	0.0	6	1	1	1
7200	130	extent entailment	[extent entailment]	0.0	2	1	1	1
7201	130	several language model LMs	[several language models LMs]	0.0	4	1	1	1
7202	130	more complex mixture	[more complex mixtures]	0.0	3	1	1	1
7203	130	use the grammar formalism	[using the grammar formalism]	0.0	4	1	1	1
7204	130	social relationship rank	[social relationship ranks]	0.0	3	1	1	1
7205	130	framework restrictive than one	[framework restrictive than ones]	0.0	4	1	1	1
7206	130	phrase in another language	[phrase in another language]	0.0	4	1	1	1
7207	130	complexity of unification	[complexity of unification]	0.0	3	1	1	1
7208	130	mt research	[MT research]	0.0	2	1	1	1
7209	130	involve automatic extraction from text	[involving automatic extraction from text]	0.0	5	1	1	1
7210	130	well-formedness condition	[well-formedness conditions]	0.0	2	1	1	1
7211	130	problem of vagueness	[problems of vagueness]	0.0	3	1	1	1
7212	130	determine the meaning	[determining the meaning]	0.0	3	1	1	1
7213	130	execution	[execution]	0.0	1	1	1	1
7214	130	coherence rule	[coherence rules]	0.0	2	1	1	1
7215	130	Chinese word sense disambiguation model	[Chinese word sense disambiguation model]	0.0	5	1	1	1
7216	130	use likelihood method	[using likelihood method]	0.0	3	1	1	1
7217	130	speech/language	[speech/language]	0.0	1	1	1	1
7218	130	superiority over exist method	[superiority over existing methods]	0.0	4	1	1	1
7219	130	unannotated text	[unannotated text]	0.0	2	1	1	1
7220	130	disambiguation WSD model	[disambiguation WSD models]	0.0	3	1	1	1
7221	130	two problem a numeric-valued attribute	[two problems a numeric-valued attributes]	0.0	5	1	1	1
7222	130	node in a graph	[node in a graph]	0.0	4	1	1	1
7223	130	several machine translation test	[several machine translation tests]	0.0	4	1	1	1
7224	130	three dimension skill level	[three dimensions skill level]	0.0	4	1	1	1
7225	130	word segmenter	[word segmenter]	0.0	2	1	1	1
7226	130	test suite	[test suite]	0.0	2	1	1	1
7227	130	string with the good performance	[string with the best performance]	0.0	5	1	1	1
7228	130	many possible structure	[many possible structures]	0.0	3	1	1	1
7229	130	training factor on performance	[training factors on performance]	0.0	4	1	1	1
7230	130	contribution	[contributions]	0.0	1	1	1	1
7231	130	place for the analysis component	[place for the analysis component]	0.0	5	1	1	1
7232	130	tag LMs	[tagging LMs]	0.0	2	1	1	1
7233	130	foundation explore bitext	[foundation exploring bitexts]	0.0	3	1	1	1
7234	130	3 indicator	[3 indicators]	0.0	2	1	1	1
7235	130	automatic speech Recognition technology	[Automatic Speech Recognition technology]	0.0	4	1	1	1
7236	130	advance in phrasal translation	[advances in phrasal translation]	0.0	4	1	1	1
7237	130	unsupervised learning algorithm	[unsupervised learning algorithm]	0.0	3	1	1	1
7238	130	process in understanding of language	[process in understanding of language]	0.0	5	1	1	1
7239	130	application of exist metric	[application of existing metrics]	0.0	4	1	1	1
7240	130	utilize primarily knowledge-based mechanism	[utilizing primarily knowledge-based mechanisms]	0.0	4	1	1	1
7241	130	instantiation of clause	[instantiation of clauses]	0.0	3	1	1	1
7242	130	domain strategy	[domain strategy]	0.0	2	1	1	1
7243	130	event type network	[event type networks]	0.0	3	1	1	1
7244	130	include jointly conditioning on consecutive word	[including jointly conditioning on consecutive words]	0.0	6	1	1	1
7245	130	topmost	[topmost]	0.0	1	1	1	1
7246	130	superiority	[superiority]	0.0	1	1	1	1
7247	130	suitable superset	[suitable superset]	0.0	2	1	1	1
7248	130	annotator	[annotators]	0.0	1	1	1	1
7249	130	extract information from unrestricted text	[extracting information from unrestricted texts]	0.0	5	1	1	1
7250	130	dynamic combiner	[dynamic combiner]	0.0	2	1	1	1
7251	130	evidence from a 500 feature	[evidence from an 500 features]	0.0	5	1	1	1
7252	130	key factor behind this finding	[key factors behind this finding]	0.0	5	1	1	1
7253	130	parsing choice	[parsing choices]	0.0	2	1	1	1
7254	130	use rule	[using rules]	0.0	2	1	1	1
7255	130	common ground in interaction	[common ground in interaction]	0.0	4	1	1	1
7256	130	approach to web-based writing	[approach to Web-based writing]	0.0	4	1	1	1
7257	130	inflection such as English	[inflection such as English]	0.0	4	1	1	1
7258	130	automated evaluation technique	[automated evaluation techniques]	0.0	3	1	1	1
7259	130	CCLINC	[CCLINC]	0.0	1	1	1	1
7260	130	meteorological event	[meteorological events]	0.0	2	1	1	1
7261	130	model of statistical machine translation	[model of statistical machine translation]	0.0	5	1	1	1
7262	130	three dimension	[three dimensions]	0.0	2	1	1	1
7263	130	include reference failure	[including reference failures]	0.0	3	1	1	1
7264	130	develop the grammar in parallel	[developing the grammars in parallel]	0.0	5	1	1	1
7265	130	research on problem	[research on problems]	0.0	3	1	1	1
7266	130	level of performance	[levels of performance]	0.0	3	1	1	1
7267	130	open-domain question answer capability	[open-domain question answering capability]	0.0	4	1	1	1
7268	130	utterance classification	[utterance classification]	0.0	2	1	1	1
7269	130	relatively domain-independent capability	[relatively domain-independent capabilities]	0.0	3	1	1	1
7270	130	text-understanding	[text-understanding]	0.0	1	1	1	1
7271	130	robust speech dialogue interaction	[robust speech dialogue interactions]	0.0	4	1	1	1
7272	130	collection	[collection]	0.0	1	1	1	1
7273	130	small amount of speech	[small amount of speech]	0.0	4	1	1	1
7274	130	target recognition	[target recognition]	0.0	2	1	1	1
7275	130	feature of question	[features of questions]	0.0	3	1	1	1
7276	130	pattern-matching parser	[pattern-matching parser]	0.0	2	1	1	1
7277	130	instance mt system	[instance MT systems]	0.0	3	1	1	1
7278	130	scale of spelling-checker	[scale of spelling-checkers]	0.0	3	1	1	1
7279	130	template-based generation component	[template-based generation component]	0.0	3	1	1	1
7280	130	conglomeration of individual lexical rule	[conglomeration of individual lexical rules]	0.0	5	1	1	1
7281	130	complete distribution of grammatical knowledge	[Complete distribution of grammatical knowledge]	0.0	5	1	1	1
7282	130	considerable importance to Machine Translation	[considerable importance to Machine Translation]	0.0	5	1	1	1
7283	130	technology of powerful computer	[technology of powerful computers]	0.0	4	1	1	1
7284	130	possible candidate the structure	[possible candidates the structure]	0.0	4	1	1	1
7285	130	loglinear model	[loglinear models]	0.0	2	1	1	1
7286	130	explicit use	[explicit use]	0.0	2	1	1	1
7287	130	include a word on aps	[including a word on APs]	0.0	5	1	1	1
7288	130	default reasoning	[default reasoning]	0.0	2	1	1	1
7289	130	identify word	[identifying words]	0.0	2	1	1	1
7290	130	zero occurrence	[zero occurrences]	0.0	2	1	1	1
7291	130	contribute in the sentence	[contributing in the sentence]	0.0	4	1	1	1
7292	130	speed of the program	[speed of the program]	0.0	4	1	1	1
7293	130	novel approach to machine translation	[novel approach to machine translation]	0.0	5	1	1	1
7294	130	phonological system speech recognition	[phonological system speech recognition]	0.0	4	1	1	1
7295	130	ne email he/she/man/woman	[NE e he/she/man/woman]	0.0	3	1	1	1
7296	130	syntactic category	[syntactic categories]	0.0	2	1	1	1
7297	130	phrase structure	[phrase structure]	0.0	2	1	1	1
7298	130	combine several language model	[combining several language models]	0.0	4	1	1	1
7299	130	application of some	[applications of some]	0.0	3	1	1	1
7300	130	generative phonological approach	[generative phonological approach]	0.0	3	1	1	1
7301	130	neural network	[neural network]	0.0	2	1	1	1
7302	130	challenge extend the CSR corpus	[challenge extending the CSR corpus]	0.0	5	1	1	1
7303	130	improvement in alignment quality	[improvement in alignment quality]	0.0	4	1	1	1
7304	130	consist	[consisting]	0.0	1	1	1	1
7305	130	sublanguage with text structure	[sublanguages with text structure]	0.0	4	1	1	1
7306	130	answer in multiple corpus	[answers in multiple corpora]	0.0	4	1	1	1
7307	130	resultant corpus	[resultant corpus]	0.0	2	1	1	1
7308	130	paraphrase classification accuracy	[paraphrase classification accuracy]	0.0	3	1	1	1
7309	130	novel bootstrapping approach tag	[novel bootstrapping approach tagging]	0.0	4	1	1	1
7310	130	sentence in comparable non-parallel corpus	[sentences in comparable non-parallel corpora]	0.0	5	1	1	1
7311	130	f-measure a decrease	[F-measure a decrease]	0.0	3	1	1	1
7312	130	statistical machine translation mt	[statistical machine translation MT]	0.0	4	2	2	1
7313	130	broad range	[broad range]	0.0	2	1	1	1
7314	130	certain aspect	[certain aspects]	0.0	2	1	1	1
7315	130	segment of description	[segments of descriptions]	0.0	3	1	1	1
7316	130	Senseval-3 sample task	[Senseval-3 sample task]	0.0	3	1	1	1
7317	130	identify word in Chinese newspaper	[identifying words in Chinese newspapers]	0.0	5	1	1	1
7318	130	important task	[important task]	0.0	2	1	1	1
7319	130	character size	[character size]	0.0	2	1	1	1
7320	130	parsing-as-deduction	[parsing-as-deduction]	0.0	1	1	1	1
7321	130	equivalence of the two formalism	[equivalence of the two formalisms]	0.0	5	1	1	1
7322	130	corresponding to semantic domain	[corresponding to semantic domains]	0.0	4	1	1	1
7323	130	asynchronous message passing	[asynchronous message passing]	0.0	3	1	1	1
7324	130	problem a numeric-valued attribute	[problems a numeric-valued attributes]	0.0	4	1	1	1
7325	130	maximum a estimation of markov model	[maximum a estimation of Markov models]	0.0	6	1	1	1
7326	130	accuracy superior	[accuracy superior]	0.0	2	1	1	1
7327	130	probabilistic grammar	[probabilistic grammars]	0.0	2	1	1	1
7328	130	method build corpus	[method building corpus]	0.0	3	1	1	1
7329	130	induce semantic verb class	[inducing semantic verb classes]	0.0	4	1	1	1
7330	130	approach to web-based academic writing	[approach to Web-based academic writing]	0.0	5	1	1	1
7331	130	linguistic expressiveness	[linguistic expressiveness]	0.0	2	1	1	1
7332	130	natural sublanguage	[natural sublanguages]	0.0	2	1	1	1
7333	130	handle word	[handling words]	0.0	2	1	1	1
7334	130	learning phrase long than three word	[learning phrases longer than three words]	0.0	6	1	1	1
7335	130	approach for interactive Machine Translation	[approach for Interactive Machine Translation]	0.0	5	1	1	1
7336	130	minute per extract	[minutes per extract]	0.0	3	1	1	1
7337	130	SI training of Markov model	[SI training of Markov models]	0.0	5	1	1	1
7338	130	use bilingual corpus a resource	[using bilingual corpora a resource]	0.0	5	1	1	1
7339	130	i parsing a verb language	[i parsing a verb language]	0.0	5	1	1	1
7340	130	most importantly text	[Most importantly texts]	0.0	3	1	1	1
7341	130	32 improvement	[32 improvement]	0.0	2	1	1	1
7342	130	heuristic Parsing Model	[Heuristic Parsing Model]	0.0	3	1	1	1
7343	130	indicator for the prediction task	[indicators for the prediction task]	0.0	5	1	1	1
7344	130	probabilistic character recognition model	[probabilistic character recognition model]	0.0	4	1	1	1
7345	130	system for scoring set	[system for scoring sets]	0.0	4	1	1	1
7346	130	concurrent computation model	[concurrent computation model]	0.0	3	1	1	1
7347	130	also a generalization of chart parsing	[also a generalization of chart Parsing]	0.0	6	1	1	1
7348	130	project stage	[project stage]	0.0	2	1	1	1
7349	130	evaluation use Arabic information retrieval	[evaluation using Arabic information retrieval]	0.0	5	1	1	1
7350	130	use bilingual corpus a available resource	[using bilingual corpora a available resource]	0.0	6	1	1	1
7351	130	multilingual input	[multilingual input]	0.0	2	1	1	1
7352	130	role convey the meaning	[role conveying the meaning]	0.0	4	1	1	1
7353	130	aggregate	[aggregate]	0.0	1	1	1	1
7354	130	several translation model	[several translation models]	0.0	3	1	1	1
7355	130	state proprietary Arabic stemmer	[state proprietary Arabic stemmer]	0.0	4	1	1	1
7356	130	hold candidate	[holding candidates]	0.0	2	2	2	1
7357	130	cluster method	[clustering methods]	0.0	2	1	1	1
7358	130	binary relational calculus notation	[binary relational calculus notation]	0.0	4	1	1	1
7359	130	use a model in place	[using a model in place]	0.0	5	1	1	1
7360	130	argumentation	[argumentation]	0.0	1	1	1	1
7361	130	.3m character	[.3M characters]	0.0	2	1	1	1
7362	130	system development to new domain	[system development to new domains]	0.0	5	1	1	1
7363	130	property of discourse	[properties of discourse]	0.0	3	1	1	1
7364	130	intelligent mobile agent	[intelligent mobile agents]	0.0	3	1	1	1
7365	130	document descriptor keyword	[document descriptors keywords]	0.0	3	1	1	1
7366	130	lexical feature from the context	[lexical features from the contexts]	0.0	5	1	1	1
7367	130	use corpus a resource	[using corpora a resource]	0.0	4	1	1	1
7368	130	use interpolation	[using interpolation]	0.0	2	1	1	1
7369	130	analysis of people' answer	[analysis of people's answers]	0.0	4	1	1	1
7370	130	new problem include vocabulary	[new problems including vocabulary]	0.0	4	1	1	1
7371	130	exploit a non-parallel corpus	[exploiting a non-parallel corpus]	0.0	4	1	1	1
7372	130	concept of the theory	[concepts of the theory]	0.0	4	1	1	1
7373	130	setting	[settings]	0.0	1	1	1	1
7374	130	role in human cognition	[role in human cognition]	0.0	4	1	1	1
7375	130	new approach for Machine Translation	[new approach for Machine Translation]	0.0	5	1	1	1
7376	130	drastic generalization roughly corresponding	[drastic generalization roughly corresponding]	0.0	4	1	1	1
7377	130	segment order	[segment order]	0.0	2	1	1	1
7378	130	representation the Phrasal Lexicon DHPL	[representation the Phrasal Lexicon DHPL]	0.0	5	1	1	1
7379	130	predict segment boundary in multiparty dialogue	[predicting segment boundaries in multiparty dialogue]	0.0	6	1	1	1
7380	130	scheme' ability	[scheme's ability]	0.0	2	1	1	1
7381	130	large non-parallel corpus	[large non-parallel corpus]	0.0	3	1	1	1
7382	130	use bilingual corpus	[using bilingual corpora]	0.0	3	1	1	1
7383	130	well-formedness condition on tree	[well-formedness conditions on trees]	0.0	4	1	1	1
7384	130	feature over parse tree	[features over parse trees]	0.0	4	1	1	1
7385	130	improvement in average precision	[improvement in average precision]	0.0	4	1	1	1
7386	130	train spot	[training SPoT]	0.0	2	1	1	1
7387	130	additional increase	[additional increase]	0.0	2	1	1	1
7388	130	build corpus	[building corpus]	0.0	2	1	1	1
7389	130	placement	[placement]	0.0	1	1	1	1
7390	130	improvement in the BLEU score	[improvements in the BLEU scores]	0.0	5	1	1	1
7391	130	89 .75% f-measure a relative decrease	[89 .75% F-measure a relative decrease]	0.0	6	1	1	1
7392	130	Phrasal Lexicon DHPL	[Phrasal Lexicon DHPL]	0.0	3	1	1	1
7393	130	constraint in broad domain	[constraints in broad domains]	0.0	4	1	1	1
7394	130	rightward manner	[rightward manner]	0.0	2	1	1	1
7395	130	use include conditioning	[use including conditioning]	0.0	3	1	1	1
7396	130	describe the processing of utterance	[describing the processing of utterances]	0.0	5	1	1	1
7397	130	three separate component	[three separate components]	0.0	3	1	1	1
7398	130	robust parsing a verb final language	[Robust parsing a verb final language]	0.0	6	1	1	1
7399	130	question answering qa system	[Question Answering QA systems]	0.0	4	1	1	1
7400	130	usual pooling	[usual pooling]	0.0	2	1	1	1
7401	130	need develop time approximation	[need developing time approximations]	0.0	4	1	1	1
7402	130	IDUS development	[IDUS development]	0.0	2	1	1	1
7403	130	development model construct a system	[development model constructing a system]	0.0	5	1	1	1
7404	130	new methodology	[new methodology]	0.0	2	1	1	1
7405	130	effect use a bilingual corpus	[effect using an bilingual corpus]	0.0	5	1	1	1
7406	130	single frame	[single frame]	0.0	2	1	1	1
7407	130	set of extraction	[set of extraction]	0.0	3	1	1	1
7408	130	two domain	[two domains]	0.0	2	1	1	1
7409	130	valid dictionary	[valid dictionary]	0.0	2	1	1	1
7410	130	present language search engine	[present language search engines]	0.0	4	1	1	1
7411	130	Dynamic Hierarchical Phrasal Lexicon [Zernik88]	[Dynamic Hierarchical Phrasal Lexicon [Zernik88]]	0.0	5	1	1	1
7412	130	technique learn morphology	[technique learning morphology]	0.0	3	1	1	1
7413	130	various relation	[various relations]	0.0	2	1	1	1
7414	130	problem enumerate all indexing	[problem enumerating all indexings]	0.0	4	1	1	1
7415	130	recent improvement	[recent improvements]	0.0	2	1	1	1
7416	130	Acyclic Graph Kernel	[Acyclic Graph Kernel]	0.0	3	1	1	1
7417	130	knowledge-based mechanism	[knowledge-based mechanisms]	0.0	2	1	1	1
7418	130	fair copy	[fair copy]	0.0	2	1	1	1
7419	130	information retrieval technique	[information retrieval techniques]	0.0	3	1	1	1
7420	130	ambiguity packing technique for Grammar	[ambiguity packing techniques for Grammars]	0.0	5	1	1	1
7421	130	method .e	[methods .e]	0.0	2	1	1	1
7422	130	interaction of task-oriented dialog	[interactions of task-oriented dialogs]	0.0	4	1	1	1
7423	130	role in the determination	[role in the determination]	0.0	4	1	1	1
7424	130	use utterance	[Using utterances]	0.0	2	1	1	1
7425	130	use in case-based mt cbmt	[use in case-based MT CBMT]	0.0	5	1	1	1
7426	130	frame encode the full meaning	[frame encoding the full meaning]	0.0	5	1	1	1
7427	130	conglomeration	[conglomeration]	0.0	1	1	1	1
7428	130	most approach	[most approaches]	0.0	2	1	1	1
7429	130	paraphrase in language	[paraphrases in language]	0.0	3	1	1	1
7430	130	three technique make syntactic analysis	[three techniques making syntactic analysis]	0.0	5	1	1	1
7431	130	interaction with knowledge	[interaction with knowledge]	0.0	3	1	1	1
7432	130	standard output system entity annotation	[standard outputs systems entity annotations]	0.0	5	1	1	1
7433	130	use additional feature of the tree	[using additional features of the tree]	0.0	6	1	1	1
7434	130	large non-parallel newspaper corpus	[large non-parallel newspaper corpora]	0.0	4	1	1	1
7435	130	theory of discourse structure	[theory of discourse structure]	0.0	4	1	1	1
7436	130	Semantic role system	[Semantic Role System]	0.0	3	1	1	1
7437	130	real dialogue datum	[real dialogue data]	0.0	3	1	1	1
7438	130	additional 500 ,000 feature	[additional 500 ,000 features]	0.0	4	1	1	1
7439	130	use of feature include conditioning	[use of features including conditioning]	0.0	5	1	1	1
7440	130	evaluate the effectiveness	[evaluating the effectiveness]	0.0	3	1	1	1
7441	130	wsd program	[WSD programs]	0.0	2	1	1	1
7442	130	perform inference	[performing inference]	0.0	2	1	1	1
7443	130	IDUS	[IDUS]	0.0	1	1	1	1
7444	130	common noun	[common noun]	0.0	2	1	1	1
7445	130	mixture language model	[mixture language model]	0.0	3	1	1	1
7446	130	word from high-accuracy word-level alignment model	[words from high-accuracy word-level alignment models]	0.0	6	1	1	1
7447	130	make analysis	[making analysis]	0.0	2	1	1	1
7448	130	monitor the collection of utterance	[monitoring the collection of utterances]	0.0	5	1	1	1
7449	130	new user	[new users]	0.0	2	1	1	1
7450	130	entity-oriented language definition	[entity-oriented language definition]	0.0	3	1	1	1
7451	130	three language system domain	[three language system domains]	0.0	4	1	1	1
7452	130	addressee identification meeting	[addressee identification meetings]	0.0	3	1	1	1
7453	130	multiparty dialogue	[multiparty dialogue]	0.0	2	1	1	1
7454	130	advantage of linguistic information	[advantage of linguistic information]	0.0	4	1	1	1
7455	130	robust efficient parsing of Korean	[Robust efficient parsing of Korean]	0.0	5	1	1	1
7456	130	bilingual material the Hansards	[bilingual material the Hansards]	0.0	4	1	1	1
7457	130	candidate parse	[candidate parses]	0.0	2	1	1	1
7458	130	produce phrase from a corpus	[producing phrases from a corpora]	0.0	5	1	1	1
7459	130	purpose in discourse	[purpose in discourse]	0.0	3	1	1	1
7460	130	transfer mapping	[transfer mappings]	0.0	2	1	1	1
7461	130	use semantic category on the terminal	[using semantic categories on the terminal]	0.0	6	1	1	1
7462	130	approach for Machine Translation	[approach for Machine Translation]	0.0	4	1	1	1
7463	130	novel method discover parallel sentence	[novel method discovering parallel sentences]	0.0	5	1	1	1
7464	130	area of language processing	[areas of language processing]	0.0	4	1	1	1
7465	130	Rapid system development via knowledge-based acquisition	[Rapid system development via knowledge-based acquisition]	0.0	6	1	1	1
7466	130	alternative speech recognition hypothese	[alternative speech recognition hypotheses]	0.0	4	1	1	1
7467	130	machine-readable dictionary MRD's	[machine-readable dictionaries MRD's]	0.0	3	1	1	1
7468	130	segment method	[segment methods]	0.0	2	1	1	1
7469	130	defeat	[defeat]	0.0	1	1	1	1
7470	130	parse-based evaluation	[parse-based evaluation]	0.0	2	1	1	1
7471	130	examination of the complexity	[examination of the complexity]	0.0	4	1	1	1
7472	130	speak	[speaking]	0.0	1	1	1	1
7473	130	different language pair	[different language pairs]	0.0	3	1	1	1
7474	130	advance in speech Recognition technology	[advances in Speech Recognition technology]	0.0	5	1	1	1
7475	130	CFG filter technique	[CFG filtering techniques]	0.0	3	1	1	1
7476	130	combination of hand-built resource	[combination of hand-built resources]	0.0	4	1	1	1
7477	130	two contribution speech recognition	[two contributions speech recognition]	0.0	4	1	1	1
7478	130	many theory	[many theories]	0.0	2	1	1	1
7479	130	Wall Street	[Wall Street]	0.0	2	1	1	1
7480	130	89 .75% f-measure	[89 .75% F-measure]	0.0	3	1	1	1
7481	130	dynamic word	[dynamic word]	0.0	2	1	1	1
7482	130	use corpus a more available resource	[using corpora a more available resource]	0.0	6	1	1	1
7483	130	rating	[rating]	0.0	1	1	1	1
7484	130	current state	[current state]	0.0	2	1	1	1
7485	130	speech recognition of a natural language	[speech recognition of a natural language]	0.0	6	1	1	1
7486	130	include limitation of mt architecture	[including limitations of MT architectures]	0.0	5	1	1	1
7487	130	three method .e method	[three methods .e methods]	0.0	4	1	1	1
7488	130	dialogue duration	[dialogue duration]	0.0	2	1	1	1
7489	130	graph indispensable to paper	[graphs indispensable to papers]	0.0	4	1	1	1
7490	130	simple mt-based paraphrasing technique	[simple MT-based paraphrasing technique]	0.0	4	1	1	1
7491	130	performance enhancing	[performance enhancing]	0.0	2	1	1	1
7492	130	reflect the social relationship email .g	[reflecting the social relationship e .g]	0.0	6	1	1	1
7493	130	role in the frame semantics paradigm	[roles in the frame semantics paradigm]	0.0	6	1	1	1
7494	130	evidence from the input	[evidence from the input]	0.0	4	1	1	1
7495	130	four speech recognition application	[four speech recognition applications]	0.0	4	1	1	1
7496	130	amount of theoretical work	[amount of theoretical work]	0.0	4	1	1	1
7497	130	conventional word-trigram recognition require transcription	[conventional word-trigram recognition requiring transcription]	0.0	5	1	1	1
7498	130	system combination	[system combination]	0.0	2	1	1	1
7499	130	maintain the scheme' ability	[maintaining the scheme's ability]	0.0	4	1	1	1
7500	130	use alignment technique	[Using alignment techniques]	0.0	3	1	1	1
7501	130	expression of language to semantics	[expressions of languages to semantics]	0.0	5	1	1	1
7502	130	less precise adverb	[less precise adverbs]	0.0	3	1	1	1
7503	130	important issue	[important issue]	0.0	2	1	1	1
7504	130	make syntactic analysis	[making syntactic analysis]	0.0	3	1	1	1
7505	130	scheduling parser	[scheduling parser]	0.0	2	1	1	1
7506	130	Unification Grammar UCG	[Unification Grammar UCG]	0.0	3	1	1	1
7507	130	grammaticality at least 99% sentence equivalence	[grammaticality at least 99% sentences equivalence]	0.0	6	1	1	1
7508	130	syntactic analysis robust---an agenda-based scheduling parser	[syntactic analysis robust---an agenda-based scheduling parser]	0.0	6	1	1	1
7509	130	Bayes	[Bayes]	0.0	1	1	1	1
7510	130	domain model with off-the-shelf classifier	[domain models with off-the-shelf classifiers]	0.0	5	1	1	1
7511	130	utilize knowledge-based mechanism	[utilizing knowledge-based mechanisms]	0.0	3	1	1	1
7512	130	balance matching	[balance matching]	0.0	2	1	1	1
7513	130	principled technique	[principled techniques]	0.0	2	1	1	1
7514	130	integrate automatic q/a application into environment	[integrating automatic Q/A applications into environments]	0.0	6	1	1	1
7515	130	combine knowledge	[combining knowledge]	0.0	2	1	1	1
7516	130	training factor on predictive performance	[training factors on predictive performance]	0.0	5	1	1	1
7517	130	beginning of the sentence	[beginning of the sentence]	0.0	4	1	1	1
7518	130	identification of name	[identification of names]	0.0	3	1	1	1
7519	130	complex mixture of technique	[complex mixtures of techniques]	0.0	4	1	1	1
7520	130	ne email	[NE e]	0.0	2	1	1	1
7521	130	problem of the detection	[problem of the detection]	0.0	4	1	1	1
7522	130	use a relational calculus notation	[using a relational calculus notation]	0.0	5	1	1	1
7523	130	marine weather	[marine weather]	0.0	2	1	1	1
7524	130	constraint as level of representation	[constraints as levels of representation]	0.0	5	1	1	1
7525	130	correct sentence	[correct sentences]	0.0	2	1	1	1
7526	130	enrichment	[enrichment]	0.0	1	1	1	1
7527	130	integrate q/a application into environment	[integrating Q/A applications into environments]	0.0	5	1	1	1
7528	130	recent advance	[recent advances]	0.0	2	1	1	1
7529	130	computer-assisted academic writing	[computer-assisted academic writing]	0.0	3	1	1	1
7530	130	complete lexical distribution of knowledge	[Complete lexical distribution of knowledge]	0.0	5	1	1	1
7531	130	performance (typically	[performance (typically]	0.0	2	1	1	1
7532	130	understand technology	[understanding technology]	0.0	2	1	1	1
7533	130	polynomial time approximation	[polynomial time approximations]	0.0	3	1	1	1
7534	130	construction as basis	[construction as basis]	0.0	3	1	1	1
7535	130	graph's	[graph's]	0.0	1	1	1	1
7536	130	three method construct word vector .e	[three methods constructing word vectors .e]	0.0	6	1	1	1
7537	130	Arabic'	[Arabic's]	0.0	1	1	1	1
7538	130	grammar checker	[grammar checker]	0.0	2	1	1	1
7539	130	frequently good predictor	[frequently good predictors]	0.0	3	1	1	1
7540	130	simple set	[simpler set]	0.0	2	1	1	1
7541	130	novel Ie paradigm	[novel IE paradigm]	0.0	3	1	1	1
7542	130	type of the error	[types of the errors]	0.0	4	1	1	1
7543	130	state stemmer	[state stemmer]	0.0	2	1	1	1
7544	130	predict the translation of the word	[predicting the translation of the words]	0.0	6	1	1	1
7545	130	way identify argument structure	[way identifying argument structures]	0.0	4	1	1	1
7546	130	efficacy apply evaluation technique	[efficacy applying evaluation techniques]	0.0	4	1	1	1
7547	130	interactive question-answering q/a system	[interactive question-answering Q/A system]	0.0	4	1	1	1
7548	130	generation component	[generation component]	0.0	2	1	1	1
7549	130	126 ,610 sentence	[126 ,610 sentences]	0.0	3	1	1	1
7550	130	evidence from a feature	[evidence from an features]	0.0	4	1	1	1
7551	130	contribute to the word match	[contributing to the word matches]	0.0	5	1	1	1
7552	130	atn	[ATN]	0.0	1	1	1	1
7553	130	system output	[system output]	0.0	2	1	1	1
7554	130	portion use a symbol	[portion using a symbol]	0.0	4	1	1	1
7555	130	correctness	[correctness]	0.0	1	1	1	1
7556	130	space of inference type	[space of inference types]	0.0	4	1	1	1
7557	130	neat form	[neat form]	0.0	2	1	1	1
7558	130	language modeling	[language modeling]	0.0	2	1	1	1
7559	130	evaluate qa system performance	[evaluating QA system performance]	0.0	4	1	1	1
7560	130	different strategy one utilize knowledge-based mechanism	[different strategies one utilizing knowledge-based mechanisms]	0.0	6	1	1	1
7561	130	bilingual corpus a commonly available resource	[bilingual corpora a commonly available resource]	0.0	6	1	1	1
7562	130	generalization of chart parsing corresponding	[generalization of chart Parsing corresponding]	0.0	5	1	1	1
7563	130	overlap	[overlapping]	0.0	1	1	1	1
7564	130	i efficient parsing language	[i efficient parsing language]	0.0	4	1	1	1
7565	130	single tagging result	[single tagging result]	0.0	3	1	1	1
7566	130	.0% relative improvement over system	[.0% relative improvement over system]	0.0	5	1	1	1
7567	130	.5% agreement with a state	[.5% agreement with a state]	0.0	5	1	1	1
7568	130	predict segment boundary	[predicting segment boundaries]	0.0	3	1	1	1
7569	130	for scoring set on the basis	[for scoring sets on the basis]	0.0	6	1	1	1
7570	130	parallel sentence	[parallel sentences]	0.0	2	1	1	1
7571	130	increase parsing time complexity	[increasing parsing time complexity]	0.0	4	1	1	1
7572	130	assignment of descriptor	[assignment of descriptors]	0.0	3	1	1	1
7573	130	revision support system	[revision support systems]	0.0	3	1	1	1
7574	130	iv modeling	[iv modeling]	0.0	2	1	1	1
7575	130	use a pos tagger	[using a POS tagger]	0.0	4	1	1	1
7576	130	two new disambiguation system one	[two new disambiguation systems one]	0.0	5	1	1	1
7577	130	general-purpose language understand technology	[general-purpose language understanding technology]	0.0	4	1	1	1
7578	130	re-utterance	[re-utterance]	0.0	1	1	1	1
7579	130	underpin	[underpinning]	0.0	1	1	1	1
7580	130	knowledge-based acquisition	[knowledge-based acquisition]	0.0	2	1	1	1
7581	130	Word	[Word]	0.0	1	1	1	1
7582	130	usage	[usage]	0.0	1	1	1	1
7583	130	combination of hand-built symbolic resource	[combination of hand-built symbolic resources]	0.0	5	1	1	1
7584	130	additional source	[additional source]	0.0	2	1	1	1
7585	130	several kind of knowledge	[several kinds of knowledge]	0.0	4	1	1	1
7586	130	two distinct domain	[two distinct domains]	0.0	3	1	1	1
7587	130	use triplet of word	[using triplets of words]	0.0	4	1	1	1
7588	130	adopt a mt-based paraphrasing technique	[adopting a MT-based paraphrasing technique]	0.0	5	1	1	1
7589	130	bitext of the law	[bitexts of the laws]	0.0	4	1	1	1
7590	130	combination of symbolic resource	[combination of symbolic resources]	0.0	4	1	1	1
7591	130	posteriori estimation	[posteriori estimation]	0.0	2	1	1	1
7592	130	optimal free indexation algorithm	[optimal free indexation algorithm]	0.0	4	1	1	1
7593	130	Grammar LFG	[Grammars LFG]	0.0	2	1	1	1
7594	130	generalization corresponding	[generalization corresponding]	0.0	2	1	1	1
7595	130	significant efficiency gain for the algorithm	[significant efficiency gains for the algorithm]	0.0	6	1	1	1
7596	130	state-of-the-art evaluation measure for machine translation	[state-of-the-art evaluation measures for machine translation]	0.0	6	1	1	1
7597	130	insight into the potential	[insight into the potential]	0.0	4	1	1	1
7598	130	start with a small parallel corpus	[starting with a small parallel corpus]	0.0	6	1	1	1
7599	130	contain word token	[containing word tokens]	0.0	3	1	1	1
7600	130	progress from generation of true text	[progressing from generation of true text]	0.0	6	1	1	1
7601	130	care	[CARE]	0.0	1	1	1	1
7602	130	easy exploitation	[easy exploitation]	0.0	2	1	1	1
7603	130	build upon the head-oriented notion	[building upon the head-oriented notions]	0.0	5	1	1	1
7604	130	utterance of speech from site	[utterances of speech from sites]	0.0	5	1	1	1
7605	130	sentence-plan-generator SPG	[sentence-plan-generator SPG]	0.0	2	1	1	1
7606	130	bilingual corpus a resource	[bilingual corpora a resource]	0.0	4	1	1	1
7607	130	corpus a more commonly available resource	[corpora a more commonly available resource]	0.0	6	1	1	1
7608	130	target language word	[target language word]	0.0	3	1	1	1
7609	130	robust parsing model	[robust parsing models]	0.0	3	1	1	1
7610	130	word contribute	[words contributing]	0.0	2	1	1	1
7611	130	technique make analysis robust---an scheduling parser	[techniques making analysis robust---an scheduling parser]	0.0	6	1	1	1
7612	130	contextual clue such as editing term	[contextual clues such as editing terms]	0.0	6	1	1	1
7613	130	special type	[special type]	0.0	2	1	1	1
7614	130	large improvement	[larger improvements]	0.0	2	1	1	1
7615	130	two system one	[two systems one]	0.0	3	1	1	1
7616	130	linguistic tool	[linguistic tools]	0.0	2	1	1	1
7617	130	robust efficient parsing language	[Robust efficient parsing language]	0.0	4	1	1	1
7618	130	sentence alignment	[sentence alignment]	0.0	2	1	1	1
7619	130	involve extraction	[involving extraction]	0.0	2	1	1	1
7620	130	machine translation tool kit	[machine translation tool kit]	0.0	4	1	1	1
7621	130	two successive learner	[two successive learners]	0.0	3	1	1	1
7622	130	sample from a entity-oriented language definition	[samples from an entity-oriented language definition]	0.0	6	1	1	1
7623	130	acquire new stem	[acquiring new stems]	0.0	3	1	1	1
7624	130	seed sentence	[seed sentences]	0.0	2	1	1	1
7625	130	disambigation	[disambigation]	0.0	1	1	1	1
7626	130	way deal	[ways dealing]	0.0	2	1	1	1
7627	130	specific move	[specific move]	0.0	2	1	1	1
7628	130	understanding of language	[understanding of language]	0.0	3	1	1	1
7629	130	include conditioning on word	[including conditioning on words]	0.0	4	1	1	1
7630	130	speculative claim to the contrary	[speculative claims to the contrary]	0.0	5	1	1	1
7631	130	good indicator for the prediction task	[better indicators for the prediction task]	0.0	6	1	1	1
7632	130	Hierarchical Graph HDAG Kernel	[Hierarchical Graph HDAG Kernel]	0.0	4	1	1	1
7633	130	template-based approach	[template-based approaches]	0.0	2	1	1	1
7634	130	IBM mt system	[IBM MT system]	0.0	3	1	1	1
7635	130	combinatorics of free indexation	[combinatorics of free indexation]	0.0	4	1	1	1
7636	130	attribute sequence of the hdag	[attribute sequences of the HDAGs]	0.0	5	1	1	1
7637	130	German corpus of srh	[German corpus of SRHs]	0.0	4	1	1	1
7638	130	leverage language technology	[leveraging language technology]	0.0	3	1	1	1
7639	130	transition graph	[transition graphs]	0.0	2	1	1	1
7640	130	processing of utterance	[processing of utterances]	0.0	3	1	1	1
7641	130	especially on the change	[especially on the changes]	0.0	4	1	1	1
7642	130	predicate with honorifics	[predicate with honorifics]	0.0	3	1	1	1
7643	130	about 110 word	[about 110 words]	0.0	3	1	1	1
7644	130	explore English-Chinese bitext in a volume	[exploring English-Chinese bitexts in a volume]	0.0	6	1	1	1
7645	130	accord to the precision metric	[according to the precision metric]	0.0	5	1	1	1
7646	130	use parsing strategy	[using parsing strategies]	0.0	3	1	1	1
7647	130	complex database	[complex databases]	0.0	2	1	1	1
7648	130	feature for paraphrase classification	[features for paraphrase classification]	0.0	4	1	1	1
7649	130	use a little speech from speaker	[using a little speech from speakers]	0.0	6	1	1	1
7650	130	acquisition of grammar	[acquisition of grammars]	0.0	3	1	1	1
7651	130	new way identify argument structure	[new way identifying argument structures]	0.0	5	1	1	1
7652	130	size of the database	[size of the databases]	0.0	4	1	1	1
7653	130	aim for complete interpretation	[aiming for complete interpretations]	0.0	4	1	1	1
7654	130	word sense WSD	[word sense WSD]	0.0	3	1	1	1
7655	130	35 .0% relative improvement	[35 .0% relative improvement]	0.0	4	1	1	1
7656	130	examination of the computational complexity	[examination of the computational complexity]	0.0	5	1	1	1
7657	130	expression of natural language	[expressions of natural languages]	0.0	4	1	1	1
7658	130	explicit information from dictionary MRD's	[explicit information from dictionaries MRD's]	0.0	5	1	1	1
7659	130	output of tool	[output of tools]	0.0	3	1	1	1
7660	130	mean equivalence	[meaning equivalence]	0.0	2	1	1	1
7661	130	draft	[drafts]	0.0	1	1	1	1
7662	130	equivalence of the formalism	[equivalence of the formalisms]	0.0	4	1	1	1
7663	130	problem predict segment boundary	[problem predicting segment boundaries]	0.0	4	1	1	1
7664	130	harvesting English-Chinese bitext of the law	[harvesting English-Chinese bitexts of the laws]	0.0	6	1	1	1
7665	130	RCG	[RCG]	0.0	1	2	2	1
7666	130	unsegmented Arabic corpus	[unsegmented Arabic corpus]	0.0	3	1	1	1
7667	130	build a stemmer	[building a stemmer]	0.0	3	1	1	1
7668	130	individually high-precision rule	[individually high-precision rules]	0.0	3	1	1	1
7669	130	good performance (typically	[best performance (typically]	0.0	3	1	1	1
7670	130	evaluation of word sense algorithm	[evaluation of word sense algorithms]	0.0	5	1	1	1
7671	130	fundamentally different strategy utilize mechanism	[fundamentally different strategies utilizing mechanisms]	0.0	5	1	1	1
7672	130	first-order statistical class grammar	[first-order statistical class grammar]	0.0	4	1	1	1
7673	130	omission	[omissions]	0.0	1	1	1	1
7674	130	parse reduction	[parse reduction]	0.0	2	1	1	1
7675	130	ACE	[ACE]	0.0	1	1	1	1
7676	130	amount of variation superior	[amount of variation superior]	0.0	4	1	1	1
7677	130	result from experiment	[Results from experiments]	0.0	3	1	1	1
7678	130	translation memory	[translation memory]	0.0	2	1	1	1
7679	130	construct a phonological system speech recognition	[constructing a phonological system speech recognition]	0.0	6	1	1	1
7680	130	attribute such as size	[attributes such as size]	0.0	4	1	1	1
7681	130	semantic representation USR	[semantic representation USR]	0.0	3	1	1	1
7682	130	most man like woman	[most men like women]	0.0	4	1	1	1
7683	130	visualization	[visualization]	0.0	1	1	1	1
7684	130	precise adverb	[precise adverbs]	0.0	2	1	1	1
7685	130	parsing of Korean final language	[parsing of Korean final language]	0.0	5	1	1	1
7686	130	usually in a manner	[usually in a manner]	0.0	4	1	1	1
7687	130	error reduction of 4	[error reduction of 4]	0.0	4	1	1	1
7688	130	two standard output	[two standard outputs]	0.0	3	1	1	1
7689	130	analysis component	[analysis component]	0.0	2	1	1	1
7690	130	output of linguistic tool	[output of linguistic tools]	0.0	4	1	1	1
7691	130	many question	[many questions]	0.0	2	1	1	1
7692	130	ambiguity packing disambiguation technique	[ambiguity packing disambiguation techniques]	0.0	4	1	1	1
7693	130	training resource	[training resources]	0.0	2	1	1	1
7694	130	determine the meaning from user	[determining the meaning from users]	0.0	5	1	1	1
7695	130	i robust efficient parsing language	[i Robust efficient parsing language]	0.0	5	1	1	1
7696	130	parsing choice of the parser	[parsing choices of the parser]	0.0	5	1	1	1
7697	130	method for scoring system output	[methods for scoring system output]	0.0	5	1	1	1
7698	130	good strategy	[best strategies]	0.0	2	1	1	1
7699	130	parallel corpus ,000 word	[parallel corpus ,000 words]	0.0	4	1	1	1
7700	130	experiment corresponding to domain	[experiments corresponding to domains]	0.0	4	1	1	1
7701	130	translation of natural language	[translation of natural language]	0.0	4	1	1	1
7702	130	continuous markov model	[continuous Markov models]	0.0	3	1	1	1
7703	130	provide insight	[providing insights]	0.0	2	1	1	1
7704	130	confines	[confines]	0.0	1	1	1	1
7705	130	three method	[three methods]	0.0	2	1	1	1
7706	130	parser/generator for LFG	[parser/generator for LFG]	0.0	3	1	1	1
7707	130	represent the meaning through the use	[representing the meaning through the use]	0.0	6	1	1	1
7708	130	conversion	[conversion]	0.0	1	1	1	1
7709	130	fine-grained modeling of unknown word feature	[fine-grained modeling of unknown word features]	0.0	6	1	1	1
7710	130	abstract for learning	[abstracts for learning]	0.0	3	1	1	1
7711	130	language search engine	[language search engines]	0.0	3	1	1	1
7712	130	new technology of personal computer	[new technology of personal computers]	0.0	5	1	1	1
7713	130	measure with judgment	[measure with judgment]	0.0	3	1	1	1
7714	130	research on new problem include vocabulary	[research on new problems including vocabulary]	0.0	6	1	1	1
7715	130	build the editor	[building the editor]	0.0	3	1	1	1
7716	130	extension of pre-exist component	[extensions of pre-existing components]	0.0	4	1	1	1
7717	130	use include conditioning on word	[use including conditioning on words]	0.0	5	1	1	1
7718	130	a attribute	[a attributes]	0.0	2	1	1	1
7719	130	statistic of model	[statistics of models]	0.0	3	1	1	1
7720	130	model use associative processor	[model using associative processors]	0.0	4	1	1	1
7721	130	sample from a language definition	[samples from an language definition]	0.0	5	1	1	1
7722	130	speed-up	[speed-up]	0.0	1	1	1	1
7723	130	SemCor datum set	[SemCor data sets]	0.0	3	1	1	1
7724	130	unique property of grammar TAG	[unique properties of grammars TAG]	0.0	5	1	1	1
7725	130	pass protocol	[passing protocols]	0.0	2	1	1	1
7726	130	automatically predict segment boundary	[automatically predicting segment boundaries]	0.0	4	1	1	1
7727	130	predicate argument structure	[predicate argument structures]	0.0	3	1	1	1
7728	130	application back	[application back]	0.0	2	1	1	1
7729	130	surface English	[surface English]	0.0	2	1	1	1
7730	130	variation in the phrasing	[variations in the phrasing]	0.0	4	1	1	1
7731	130	domain acoustic model with classifier	[domain acoustic models with classifiers]	0.0	5	1	1	1
7732	130	theory of anaphora resolution	[theories of anaphora resolution]	0.0	4	1	1	1
7733	130	complete sentence	[complete sentences]	0.0	2	1	1	1
7734	130	numbering system	[numbering system]	0.0	2	1	1	1
7735	130	neighbour	[neighbour]	0.0	1	1	1	1
7736	130	grammaticality 99% sentence equivalence in meaning	[grammaticality 99% sentences equivalence in meaning]	0.0	6	1	1	1
7737	130	corpus a much more available resource	[corpora a much more available resource]	0.0	6	1	1	1
7738	130	feature of the tree	[features of the tree]	0.0	4	1	1	1
7739	130	transition graph for a type	[transition graphs for a type]	0.0	5	1	1	1
7740	130	two contribution continuous speech recognition	[two contributions continuous speech recognition]	0.0	5	1	1	1
7741	130	change	[changes]	0.0	1	1	1	1
7742	130	many other nlp problem	[many other NLP problems]	0.0	4	1	1	1
7743	130	primarily on explicit information from dictionary	[primarily on explicit information from dictionaries]	0.0	6	1	1	1
7744	130	Markov probability	[Markov probability]	0.0	2	1	1	1
7745	130	expressiveness of the formalism	[expressiveness of the formalisms]	0.0	4	1	1	1
7746	130	disambiguation system	[disambiguation systems]	0.0	2	1	1	1
7747	130	start with a small corpus	[starting with a small corpus]	0.0	5	1	1	1
7748	130	.0% relative improvement	[.0% relative improvement]	0.0	3	1	1	1
7749	130	primarily on information	[primarily on information]	0.0	3	1	1	1
7750	130	one element	[one element]	0.0	2	1	1	1
7751	130	identify unknown word	[identifying unknown words]	0.0	3	1	1	1
7752	130	build a Arabic stemmer	[building a Arabic stemmer]	0.0	4	1	1	1
7753	130	language with inflection	[languages with inflection]	0.0	3	1	1	1
7754	130	word form	[word forms]	0.0	2	2	2	1
7755	130	large Chinese newspaper corpus	[large Chinese newspaper corpora]	0.0	4	1	1	1
7756	130	connect referent	[connecting referents]	0.0	2	1	1	1
7757	130	machine dictionary	[machine dictionaries]	0.0	2	1	1	1
7758	130	interactive method	[interactive method]	0.0	2	1	1	1
7759	130	analysis of the English coordinate sentence	[analysis of the English coordinate sentences]	0.0	6	1	1	1
7760	130	definition of Minimalist grammar	[definition of Minimalist grammars]	0.0	4	1	1	1
7761	130	knowledge about knowledge	[knowledge about knowledge]	0.0	3	1	1	1
7762	130	learning phrase word	[learning phrases words]	0.0	3	1	1	1
7763	130	use Network classifier	[using Network classifiers]	0.0	3	1	1	1
7764	130	two word-sense system	[two word-sense systems]	0.0	3	1	1	1
7765	130	information from dictionary	[information from dictionaries]	0.0	3	1	1	1
7766	130	build spelling-checker for language	[building spelling-checkers for languages]	0.0	4	1	1	1
7767	130	software infrastructure	[software infrastructure]	0.0	2	1	1	1
7768	130	omission of argument	[omissions of arguments]	0.0	3	1	1	1
7769	130	measure with human judgment	[measure with human judgment]	0.0	4	1	1	1
7770	130	architecture with a variety	[architecture with a variety]	0.0	4	1	1	1
7771	130	iii system development to domain	[iii system development to domains]	0.0	5	1	1	1
7772	130	problem combine language model LMs	[problem combining language models LMs]	0.0	5	1	1	1
7773	130	natural sublanguage with text structure	[natural sublanguages with text structure]	0.0	5	1	1	1
7774	130	edge with a valid semantic interpretation	[edges with a valid semantic interpretation]	0.0	6	1	1	1
7775	130	srh	[SRHs]	0.0	1	1	1	1
7776	130	browse the repository	[browsing the repository]	0.0	3	1	1	1
7777	130	independent event-based extractive summarization approach	[independent event-based extractive summarization approaches]	0.0	5	1	1	1
7778	130	aspect of language learning	[aspects of language learning]	0.0	4	1	1	1
7779	130	answer capability on text	[answering capability on text]	0.0	4	1	1	1
7780	130	problem numeric-valued attribute such as size	[problems numeric-valued attributes such as size]	0.0	6	1	1	1
7781	130	attribute such as size b perspective-taking	[attributes such as size b perspective-taking]	0.0	6	1	1	1
7782	130	fully-connected statistical class grammar	[fully-connected statistical class grammar]	0.0	4	1	1	1
7783	130	Hierarchical Acyclic Graph HDAG	[Hierarchical Acyclic Graph HDAG]	0.0	4	1	1	1
7784	130	small parallel corpus word	[small parallel corpus words]	0.0	4	1	1	1
7785	130	probabilistic character recognition ocr model	[probabilistic character recognition OCR model]	0.0	5	1	1	1
7786	130	use include conditioning on consecutive word	[use including conditioning on consecutive words]	0.0	6	1	1	1
7787	130	i efficient parsing of Korean language	[i efficient parsing of Korean language]	0.0	6	1	1	1
7788	130	bigram language	[bigram language]	0.0	2	1	1	1
7789	130	practical way	[practical ways]	0.0	2	1	1	1
7790	130	grammatical formalism tree Grammar	[grammatical formalisms Tree Grammars]	0.0	4	1	1	1
7791	130	progress into the noisy output	[progressing into the noisy output]	0.0	5	1	1	1
7792	130	experience with TACITUS	[experience with TACITUS]	0.0	3	1	1	1
7793	130	lack of method	[lack of methods]	0.0	3	1	1	1
7794	130	provably optimal indexation algorithm	[provably optimal indexation algorithm]	0.0	4	1	1	1
7795	130	class grammar	[class grammar]	0.0	2	1	1	1
7796	130	use a corpus	[using an corpus]	0.0	3	1	1	1
7797	130	several key factor include limitation	[several key factors including limitations]	0.0	5	1	1	1
7798	130	powerful tool	[powerful tools]	0.0	2	1	1	1
7799	130	mission	[mission]	0.0	1	1	1	1
7800	130	simple means heuristic learning	[simple means heuristic learning]	0.0	4	1	1	1
7801	130	unsupervised algorithm acquire stem	[unsupervised algorithm acquiring stems]	0.0	4	1	1	1
7802	130	effect of segment order	[effects of segment order]	0.0	4	1	1	1
7803	130	reappraise this counterexample	[reappraising these counterexamples]	0.0	3	1	1	1
7804	130	unique author	[unique author]	0.0	2	1	1	1
7805	130	several key factor include inherent limitation	[several key factors including inherent limitations]	0.0	6	1	1	1
7806	130	datum source	[data sources]	0.0	2	1	1	1
7807	130	basic goal	[basic goal]	0.0	2	1	1	1
7808	130	paradigm for interaction	[paradigm for interaction]	0.0	3	1	1	1
7809	130	modeling of unknown word feature	[modeling of unknown word features]	0.0	5	1	1	1
7810	130	sum	[sum]	0.0	1	3	3	1
7811	130	system speech recognition	[system speech recognition]	0.0	3	1	1	1
7812	130	sentence in non-parallel corpus	[sentences in non-parallel corpora]	0.0	4	1	1	1
7813	130	part of text processing software	[part of text processing software]	0.0	5	1	1	1
7814	130	output of a exist parser	[output of an existing parser]	0.0	5	1	1	1
7815	130	part of a study	[part of a study]	0.0	4	1	1	1
7816	130	therefore a good choice	[therefore a good choice]	0.0	4	1	1	1
7817	130	constraint from the text	[constraints from the text]	0.0	4	1	1	1
7818	130	application of Unification categorial Grammar	[application of Unification Categorial Grammar]	0.0	5	1	1	1
7819	130	i efficient parsing	[i efficient parsing]	0.0	3	1	1	1
7820	130	large amount of work	[large amount of work]	0.0	4	1	1	1
7821	130	interaction of dialog	[interactions of dialogs]	0.0	3	1	1	1
7822	130	intention component	[intention component]	0.0	2	1	1	1
7823	130	real-time demonstration	[real-time demonstration]	0.0	2	1	1	1
7824	130	sentence-plan-generator	[sentence-plan-generator]	0.0	1	1	1	1
7825	130	use feature as evidence	[using features as evidence]	0.0	4	1	1	1
7826	130	analysis more robust---an scheduling parser	[analysis more robust---an scheduling parser]	0.0	5	1	1	1
7827	130	standard use of BLEU	[standard use of BLEU]	0.0	4	1	1	1
7828	130	tv show	[TV shows]	0.0	2	1	1	1
7829	130	story	[story]	0.0	1	1	1	1
7830	130	symbol	[symbol]	0.0	1	1	1	1
7831	130	bus	[bus]	0.0	1	1	1	1
7832	130	achieve similar performance to mixture	[achieving similar performance to mixtures]	0.0	5	1	1	1
7833	130	frequency	[frequency]	0.0	1	1	1	1
7834	130	ii quality translation	[ii quality translation]	0.0	3	1	1	1
7835	130	missile	[missiles]	0.0	1	1	1	1
7836	130	SCF datum	[SCF data]	0.0	2	1	1	1
7837	130	generate	[generating]	0.0	1	1	1	1
7838	130	parallel corpus a resource	[parallel corpora a resource]	0.0	4	1	1	1
7839	130	find more datum	[finding more data]	0.0	3	1	1	1
7840	130	train successive learner	[training successive learners]	0.0	3	1	1	1
7841	130	practical efficiency	[practical efficiency]	0.0	2	1	1	1
7842	130	apply model	[applying models]	0.0	2	1	1	1
7843	130	plan COMPLEX a computational lexicon	[planning COMPLEX a computational lexicon]	0.0	5	1	1	1
7844	130	detailed linguistic information	[detailed linguistic information]	0.0	3	1	1	1
7845	130	CCLINC Coalition Language System	[CCLINC Coalition Language System]	0.0	4	1	1	1
7846	130	computer-assisted writing	[computer-assisted writing]	0.0	2	1	1	1
7847	130	extension of component	[extensions of components]	0.0	3	1	1	1
7848	130	newswire text	[newswire text]	0.0	2	1	1	1
7849	130	use seed	[using seeds]	0.0	2	1	1	1
7850	130	issue in Chinese language processing	[issue in Chinese Language Processing]	0.0	5	1	1	1
7851	130	empirical mt research	[empirical MT research]	0.0	3	1	1	1
7852	130	overt case marker	[overt case markers]	0.0	3	1	1	1
7853	130	Prolog cf query optimisation	[Prolog cf query optimisation]	0.0	4	1	1	1
7854	130	framework less restrictive	[framework less restrictive]	0.0	3	1	1	1
7855	130	ambiguity resolution	[ambiguity resolution]	0.0	2	1	1	1
7856	130	use strategy	[using strategies]	0.0	2	1	1	1
7857	130	Travel	[Travel]	0.0	1	2	2	1
7858	130	pose problem for reader	[posing problems for readers]	0.0	4	1	1	1
7859	130	strength of antecedence	[strength of antecedence]	0.0	3	1	1	1
7860	130	effectiveness of this approach	[effectiveness of this approach]	0.0	4	1	1	1
7861	130	use less memory	[using less memory]	0.0	3	1	1	1
7862	130	triplet	[triplets]	0.0	1	1	1	1
7863	130	start with a parallel corpus word	[starting with a parallel corpus words]	0.0	6	1	1	1
7864	130	two style perform inference	[Two styles performing inference]	0.0	4	1	1	1
7865	130	Lincoln Laboratory	[Lincoln Laboratory]	0.0	2	1	1	1
7866	130	man like mostly young woman	[men like mostly young women]	0.0	5	1	1	1
7867	130	speaker group	[speaker group]	0.0	2	1	1	1
7868	130	robust parsing of Korean final language	[Robust parsing of Korean final language]	0.0	6	1	1	1
7869	130	Grammar approach	[Grammars approach]	0.0	2	1	1	1
7870	130	harvesting English-Chinese bitext from the web	[harvesting English-Chinese bitexts from the Web]	0.0	6	1	1	1
7871	130	attain a design overhead	[attaining a design overheads]	0.0	4	1	1	1
7872	130	response to each user	[responses to each user]	0.0	4	1	1	1
7873	130	interesting combination of symbolic resource	[interesting combination of symbolic resources]	0.0	5	1	1	1
7874	130	example from a bi-directional fragment	[example from an bi-directional fragment]	0.0	5	1	1	1
7875	130	a numeric-valued attribute	[a numeric-valued attributes]	0.0	3	1	1	1
7876	130	language question	[language question]	0.0	2	1	1	1
7877	130	good predictor of document descriptor	[good predictors of document descriptors]	0.0	5	1	1	1
7878	130	generative probabilistic model of language	[generative probabilistic model of language]	0.0	5	1	1	1
7879	130	discourse phenomenon	[discourse phenomena]	0.0	2	1	1	1
7880	130	grammar generation	[grammar generation]	0.0	2	1	1	1
7881	130	block selection	[block selection]	0.0	2	1	1	1
7882	130	different way	[different ways]	0.0	2	1	1	1
7883	130	computational theory of reasoning	[computational theory of reasoning]	0.0	4	1	1	1
7884	130	iii Rapid system development via acquisition	[iii Rapid system development via acquisition]	0.0	6	1	1	1
7885	130	construct word vector	[constructing word vectors]	0.0	3	1	1	1
7886	130	result of a user evaluation	[results of a user evaluation]	0.0	5	1	1	1
7887	130	new technology of powerful computer	[new technology of powerful computers]	0.0	5	1	1	1
7888	130	computer with integral graphic display	[computers with integral graphics displays]	0.0	5	1	1	1
7889	130	strategy one utilize primarily knowledge-based mechanism	[strategies one utilizing primarily knowledge-based mechanisms]	0.0	6	1	1	1
7890	130	i parsing of Korean final language	[i parsing of Korean final language]	0.0	6	1	1	1
7891	130	equivalence in meaning	[equivalence in meaning]	0.0	3	1	1	1
7892	130	set of extract	[set of extracts]	0.0	3	1	1	1
7893	130	sets(	[sets(]	0.0	1	1	1	1
7894	130	combination strategy	[combination strategies]	0.0	2	1	1	1
7895	130	behaviour	[behaviour]	0.0	1	1	1	1
7896	130	include the influence	[including the influence]	0.0	3	1	1	1
7897	130	modeling of word feature	[modeling of word features]	0.0	4	1	1	1
7898	130	segment order-sensitive method	[segment order-sensitive methods]	0.0	3	1	1	1
7899	130	statistic on co-occurrence pattern	[statistics on co-occurrence patterns]	0.0	4	1	1	1
7900	130	underspecified representation USR	[underspecified representation USR]	0.0	3	1	1	1
7901	130	preference of approach	[preference of approach]	0.0	3	1	1	1
7902	130	dependency network	[dependency network]	0.0	2	1	1	1
7903	130	young man like mostly woman	[young men like mostly women]	0.0	5	1	1	1
7904	130	material Roget's Thesaurus	[material Roget's Thesaurus]	0.0	3	1	1	1
7905	130	cost of generation system	[cost of generation systems]	0.0	4	1	1	1
7906	130	information from machine-readable dictionary	[information from machine-readable dictionaries]	0.0	4	1	1	1
7907	130	pooling of the speech datum	[pooling of the speech data]	0.0	5	1	1	1
7908	130	occurrence of a morpheme	[occurrences of a morpheme]	0.0	4	1	1	1
7909	130	unique property of grammar	[unique properties of grammars]	0.0	4	1	1	1
7910	130	prior in model	[priors in models]	0.0	3	1	1	1
7911	130	relationship between the formalism	[relationship between the formalisms]	0.0	4	1	1	1
7912	130	language interface	[language interface]	0.0	2	5	5	1
7913	130	single semantic frame	[single semantic frame]	0.0	3	1	1	1
7914	130	computer text	[computer text]	0.0	2	1	1	1
7915	130	word-level alignment model	[word-level alignment models]	0.0	3	1	1	1
7916	130	information related	[information related]	0.0	2	1	1	1
7917	130	representative chinese-to-english SMT model	[representative Chinese-to-English SMT model]	0.0	4	1	1	1
7918	130	such mt cbmt	[such MT CBMT]	0.0	3	1	1	1
7919	130	unique property	[unique properties]	0.0	2	1	1	1
7920	130	monitor the collection of 12,000 utterance	[monitoring the collection of 12,000 utterances]	0.0	6	1	1	1
7921	130	reliable basis	[reliable basis]	0.0	2	1	1	1
7922	130	event-based extractive mutli-document summarization approach	[event-based extractive mutli-document summarization approaches]	0.0	5	1	1	1
7923	130	integral graphic display	[integral graphics displays]	0.0	3	1	1	1
7924	130	alternative speech recognition hypothese SRH	[alternative speech recognition hypotheses SRH]	0.0	5	1	1	1
7925	130	include jointly conditioning on multiple word	[including jointly conditioning on multiple words]	0.0	6	1	1	1
7926	130	four effect the reduction	[four effects the reduction]	0.0	4	1	1	1
7927	130	one superiority of this method	[one superiority of this method]	0.0	5	1	1	1
7928	130	Arabic-English translation task	[Arabic-English translation task]	0.0	3	1	1	1
7929	130	reduce the amount	[reducing the amount]	0.0	3	1	1	1
7930	130	efficacy	[efficacy]	0.0	1	1	1	1
7931	130	discourse segmentation	[discourse segmentation]	0.0	2	1	1	1
7932	130	extension of this basic idea	[extensions of this basic idea]	0.0	5	1	1	1
7933	130	directly on word sense	[directly on word sense]	0.0	4	1	1	1
7934	130	provide insight into the the complexity	[providing insights into the the complexity]	0.0	6	1	1	1
