Elamur Research⌕
☰
home / papers / 2014 / Cross-lingual Coreference Resolution of Pronouns

Cross-lingual Coreference Resolution of Pronouns

Summary

Novák and Žabokrtský study coreference resolution (CR) on a bitext. Their goal is to resolve English 3rd-person central pronouns (personal, possessive and reflexive) more accurately by drawing on the aligned Czech translation. The two languages handle pronouns differently. Czech has grammatical gender, drops subject pronouns and uses reflexive possessives. Those differences can supply useful hints to an English resolver. The authors use PCEDT, but replace its manual annotation with fully automatic Treex analysis. They keep gold data only for coreference links and for their own pronoun alignment annotation. They extend an English mention-ranking CR model with features projected from the Czech side and with one indicator of a Czech-predicted coreference link. Missing Czech counterparts make standard word alignment unreliable, so they also train a supervised pronoun aligner. This aligner beats the GIZA++-based alignment. The best cross-lingual configuration improves on both a monolingual baseline and a pure projection approach. The authors describe this as the first machine-learning attempt at cross-lingual CR. They also argue it is relevant to discourse-aware SMT.

Key points

  • Data: Prague Czech-English Dependency Treebank (PCEDT) contains about 1.2M words per language in nearly 50,000 sentence pairs. Its English side is the Penn Treebank WSJ section. All layers were re-annotated automatically in the Treex framework.
  • Baseline alignment: MGIZA++ (a multi-threaded GIZA++) was trained on CzEng 1.0, which is roughly 300 times larger than PCEDT. The authors added a supervised pronoun aligner that takes the GIZA++ output as one of its features. It wins in intrinsic evaluation and also in downstream CR.
  • Feature sets compared: EN (monolingual), EN + CS (projected Czech features), EN + CS-COREF (Czech coreference indicator) and all three combined. Each is tested with both the original and the supervised alignment.
  • Results:
    • With supervised alignment, every cross-lingual variant beats the baseline and all systems using the original alignment.
    • The best system gains more than 1 point absolute; read from Table 3, eval F1 rises from about 70.2 to about 71.2.
    • This gain is significant at p ≤ 0.1, but not at p ≤ 0.05.
    • Projection-only CR performs far below the baseline.
  • Error analysis: The analysis used 1,988 dev instances.
    • EN + CS fixed 49 cases and broke 23.
    • EN + CS-COREF fixed 63 and broke 39.
    • The gains came mainly from pleonastic it, possessives with Czech reflexive counterparts, and gender/number agreement.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

Fields: Linguistics & languages

Topics: Coreference resolution

Cited by

10 papers of our library cite it (of the 140673 whose bibliographies we read); 3 of them have a page here, 7 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work.

@inproceedings{novak2014cross,
  title = {{Cross-lingual Coreference Resolution of Pronouns}},
  author = {Novák, Michal and Žabokrtský, Zdeněk},
  booktitle = {Proceedings of COLING 2014, the 25th International Conference on Computational Linguistics: Technical Papers},
  year = {2014},
  url = {https://aclanthology.org/C14-1003/}
}

References

37 references, 10 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Ondřej Bojar, Zdeněk Žabokrtský, Ondřej Dušek, Petra Galuščáková, Martin Majliš, David Mareček, Jiřı́ Maršı́k, Michal Novák, Martin Popel, and Aleš Tamchyna. 2012. The joy of parallelism with CzEng 1.0. In Proceedings of LREC 2012, Istanbul, Turkey. European Language Resources Associ… find
  2. David Burkett and Dan Klein. 2008. Two languages are better than one (for syntactic parsing). In Proceedings of the Conference on Empirical Methods in Natural Language Processing, Stroudsburg, PA, USA. Association for Computational Linguistics. find
  3. David Burkett, John Blitzer, and Dan Klein. 2010. Joint parsing and alignment with weakly synchronized grammars. In Human Language Technologies: The 2010 Annual Conference of the North American Chapter of the Association for Computational Linguistics, Stroudsburg, PA, USA. Association for Computatio… find
  4. Eugene Charniak and Micha Elsner. 2009. EM works for pronoun anaphora resolution. In Proceedings of the 12th Conference of the European Chapter of the Association for Computational Linguistics, Stroudsburg, PA, USA. Association for Computational Linguistics. find our card
  5. Dev P R 60.28 79.14 Eval P R 63.28 78.78 A 60.97 F 68.43 A 63.72 F 70.19 36.55 61.77 61.97 62.27 41.98 61.07 61.15 61.33 24.66 80.45 80.23 80.96 31.07 69.44 69.40 69.79 33.33 64.30 64.07 64.03 42.38 63.74 63.72 63.59 21.58 79.62 78.62 79.57 28.60 70.80 70.39 70.69 36.60 62.27 62.17 62.68 41.38 61.42… find
  6. CNC. 2005. Czech national corpus – SYN2005. Prague, Czech Republic. Institute of the Czech National Corpus. find
  7. Dipanjan Das and Slav Petrov. 2011. Unsupervised part-of-speech tagging with bilingual graph-based projections. In Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies - Volume 1, Stroudsburg, PA, USA. Association for Computational Ling… find
  8. José Guilherme Camargo de Souza and Constantin Orsan. 2011. Can projected chains in parallel corpora help coreference resolution? In Proceedings of the 8th International Conference on Anaphora Processing and Applications, Berlin, Heidelberg. Springer-Verlag. find
  9. Pascal Denis and Jason Baldridge. 2007. A ranking approach to pronoun resolution. In Proceedings of the 20th International Joint Conference on Artifical Intelligence, San Francisco, CA, USA. Morgan Kaufmann Publishers Inc. find
  10. Qin Gao and Stephan Vogel. 2008. Parallel implementations of word alignment tool. In Software Engineering, Testing, and Quality Assurance for Natural Language Processing, Stroudsburg, PA, USA. Association for Computational Linguistics. find
  11. Liane Guillou, Christian Hardmeier, Aaron Smith, Jrg Tiedemann, and Bonnie Webber. 2014. ParCor 1.0: A parallel pronoun-coreference corpus to support statistical MT. In Proceedings of the 9th International Conference on Language Resources and Evaluation (LREC-2014), Reykjavik, Iceland. European Lang… find our card
  12. Jan Hajič, Eva Hajičová, Jarmila Panevová, Petr Sgall, Ondřej Bojar, Silvie Cinková, Eva Fučı́ková, Marie Mikulová, Petr Pajas, Jan Popelka, Jiřı́ Semecký, Jana Šindlerová, Jan Štěpánek, Josef Toman, Zdeka Urešová, and Zdeněk Žabokrtský. 2012. Announcing Prague Czech-English Dep… find
  13. Jan Hajič. 2004. Disambiguation of Rich Inflection (Computational Morphology of Czech). Nakladatelstvı́ Karolinum. find
  14. Sanda M. Harabagiu and Steven J. Maiorano. 2000. Multilingual coreference resolution. In Proceedings of the Sixth Conference on Applied Natural Language Processing, Stroudsburg, PA, USA. Association for Computational Linguistics. find our card
  15. Martin Wittorff Haulrich. 2012. Data-driven bitext dependency parsing and alignment. Ph.D. thesis, Copenhagen Business School. find
  16. Eduard Hovy, Mitchell Marcus, Martha Palmer, Lance Ramshaw, and Ralph Weischedel. 2006. OntoNotes: the 90% solution. In Proceedings of the Human Language Technology Conference of the NAACL, Companion Volume: Short Papers, Stroudsburg, PA, USA. Association for Computational Linguistics. find
  17. Sungchul Kim, Kristina Toutanova, and Hwanjo Yu. 2012. Multilingual named entity recognition using parallel data and metadata from Wikipedia. In Proceedings of the 50th Annual Meeting of the Association for Computational Linguistics: Long Papers - Volume 1, Stroudsburg, PA, USA. Association for Comp… find
  18. Qi Li, Haibo Li, Heng Ji, Wen Wang, Jing Zheng, and Fei Huang. 2012. Joint bilingual name tagging for parallel corpora. In Proceedings of the 21st ACM International Conference on Information and Knowledge Management, New York, NY, USA. ACM. find
  19. Linguistic Data Consortium. 1999. Penn Treebank 3. LDC99T42. find
  20. Ruslan Mitkov and Catalina Barbu. 2003. Using bilingual corpora to improve pronoun resolution. Languages in contrast, 4(2). find
  21. Vincent Ng. 2010. Supervised noun phrase coreference research: The first fifteen years. In Proceedings of the 48th Annual Meeting of the Association for Computational Linguistics, Stroudsburg, PA, USA. Association for Computational Linguistics. find our card
  22. Giang Linh Nguy, Václav Novák, and Zdeněk Žabokrtský. 2009. Comparison of classification and ranking approaches to pronominal anaphora resolution in Czech. In Proceedings of the SIGDIAL 2009 Conference, London, UK. The Association for Computational Linguistics. find our card
  23. Michal Novák, Anna Nedoluzhko, and Zdeněk Žabokrtský. 2013. Translation of “it” in a deep syntax framework. In 51st Annual Meeting of the Association for Computational Linguistics Proceedings of the Workshop on Discourse in Machine Translation, Sofija, Bulgaria. Omnipress, Inc. find our card
  24. Franz Josef Och and Hermann Ney. 2000. Improved statistical alignment models. In Proceedings of the 38th Annual Meeting on Association for Computational Linguistics, Stroudsburg, PA, USA. Association for Computational Linguistics. find our card
  25. Maciej Ogrodniczuk. 2013. Translation- and projection-based unsupervised coreference resolution for Polish. In Language Processing and Intelligent Information Systems, number 7912, Berlin / Heidelberg. Springer. find
  26. Martin Popel and Zdeněk Žabokrtský. 2010. TectoMT: Modular NLP framework. In Lecture Notes in Artificial Intelligence, Proceedings of the 7th International Conference on Advances in Natural Language Processing (IceTAL 2010), volume 6233, Berlin / Heidelberg. Springer. find
  27. Oana Postolache, Dan Cristea, and Constantin Orasan. 2006. Transferring coreference chains through word alignment. In Proceedings of the Fifth International Conference on Language Resources and Evaluation, Genoa, Italy. European Language Resources Association. find our card
  28. Randolph Quirk, Sidney Greenbaum, Geoffrey Leech, and Jan Svartvik. 1985. A Comprehensive Grammar of the English Language. Longman. find
  29. Altaf Rahman and Vincent Ng. 2009. Supervised models for coreference resolution. In Proceedings of the 2009 Conference on Empirical Methods in Natural Language Processing: Volume 2 - Volume 2, Stroudsburg, PA, USA. Association for Computational Linguistics. find our card
  30. Altaf Rahman and Vincent Ng. 2012. Translation-based projection for multilingual coreference resolution. In Proceedings of the 2012 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Stroudsburg, PA, USA. Association for Computatio… find our card
  31. Rudolf Rosa, Ondřej Dušek, David Mareček, and Martin Popel. 2012. Using parallel features in parsing of machine-translated sentences for correction of grammatical errors. In Proceedings of the Sixth Workshop on Syntax, Semantics and Structure in Statistical Translation, Stroudsburg, PA, USA. Asso… find
  32. Petr Sgall, Eva Hajičová, and Jarmila Panevová. 1986. The Meaning of the Sentence in Its Semantic and Pragmatic Aspects. D. Reidel Publishing Company, Dordrecht. find
  33. Petr Sgall. 1967. Generativnı́ popis jazyka a česká deklinace. Academia, Prague, Czech Republic. find
  34. David A. Smith and Noah A. Smith. 2004. Bilingual parsing with factored estimation: Using English to parse Korean. In Proceedings of the 2004 Conference on Empirical Methods in Natural Language Processing. Association for Computational Linguistics. find
  35. Kateřina Veselovská, Giang Linh Nguy, and Michal Novák. 2012. Using Czech-English parallel corpora in automatic identification of “it”. In The Fifth Workshop on Building and Using Comparable Corpora, İstanbul, Turkey. European Language Resources Association. find
  36. Piek Vossen, editor. 1998. EuroWordNet: A Multilingual Database with Lexical Semantic Networks. Kluwer Academic Publishers, Norwell, MA, USA. find
  37. Tao Zhuang and Chengqing Zong. 2010. Joint inference for bilingual semantic role labeling. In Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing, Stroudsburg, PA, USA. Association for Computational Linguistics. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/novak-2014-cross-lingual-coreference-resolution-pronouns/