Bartek

This page offers the official Creative Commons Attribution 3.0 Unported License release of the Bartek, a statistical tool for performing Coreference Resolution (CR). By downloading the Bartek package you accept the conditions of that licence.

Principal developer: Mateusz Kopeć
Authors: Mateusz Kopeć, Bartłomiej Nitoń, Maciej Ogrodniczuk
License: CC BY v.3

Documentation

Manual file of Bartek 1.0, in English: in preparation

Downloads

Bartek 1.0 is available to download in two versions:

• standalone jar

• source code

Bartek comes with default models trained on the full Polish Coreference Corpus. It also contains compiled resources based on [[pl.wikipedia.org|Polish Wikipedia] and [[http://plwordnet.pwr.wroc.pl/wordnet/|plWordnet].

Online demo

Multiservice will soon offer an online demo of Bartek.

You may also want to see other Polish Coreference Tools.

Citing

When using Bartek, please cite the following article:

Maciej Ogrodniczuk, Katarzyna Głowińska, Mateusz Kopeć, Agata Savary, and Magdalena Zawisławska. Polish Coreference Corpus. In Zygmunt Vetulani, editor, Proceedings of the 6th Language & Technology Conference: Human Language Technologies as a Challenge for Computer Science and Linguistics, pages 494–498, Poznań, Poland, 2013. Wydawnictwo Poznańskie, Fundacja Uniwersytetu im. Adama Mickiewicza.