Elamur Research⌕
☰
home / papers / 2011 / A Pronoun Anaphora Resolution System based on Factorial Hidden Markov Models

A Pronoun Anaphora Resolution System based on Factorial Hidden Markov Models

Summary

The paper addresses pronoun anaphora resolution, the task of linking a pronoun to the earlier mention it refers to, using a supervised generative approach built on factorial hidden Markov models (FHMMs). The design is inspired by human incremental processing. The model reads text word by word and maintains a small hidden buffer of recently introduced entities, which serves as an explicit short-term discourse memory. Each hidden state has three coupled components: a referent operation (none, new or old mention), coreference features stored for each buffered entity, and a part-of-speech tag. As a result, POS tagging and pronoun resolution are decoded together. Pronouns are modeled as being generated by buffer entries, and noun phrases are represented by their head words. On the ACE corpus (BNEWS, NPAPER and NWIRE genres), the system achieved the highest average score among the compared systems, outperforming the mention-ranking model of Denis and Baldridge (2007) and the unsupervised emPronouns system. The authors argue that sequential models with an explicit entity memory are a promising alternative to pairwise and ranking approaches.

Key points

  • Each buffered mention carries four coreference features: index (recency order), named-entity type, number and gender. Gender and number distribution data came from Shane Bergsma.
  • The FHMM system averaged 76.3, compared with 75.2 for the Denis and Baldridge ranking (RK) model, even when RK used a 10-sentence window. RK was 0.5 points better only on NWIRE.
  • A paired Student's t-test against emPronouns showed statistically significant gains on all three genres (p < 0.001). emPronouns was run with its original model files and was not retrained on ACE.
  • In the training data, a mention distance of 14 covered every pronoun–antecedent pair. Very distant antecedents remain a source of errors because the buffer size is limited.
  • Other errors came from plural candidates that could not be distinguished by gender, coordinate noun phrases that lack a single head word, and sparse or missing features. Sparsity was handled with add-one smoothing and deleted interpolation.

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

5 papers of our library cite it (of the 140673 whose bibliographies we read); 2 of them have a page here, 3 not yet.

Cite

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

@inproceedings{li2011pronoun,
  title = {{A Pronoun Anaphora Resolution System based on Factorial Hidden Markov Models}},
  author = {Li, Dingcheng and Miller, Tim and Schuler, William},
  booktitle = {Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies},
  year = {2011},
  url = {https://aclanthology.org/P11-1117/}
}

References

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

  1. S Bergsma. 2005. Automatic acquisition of gender information for anaphora resolution. page 342353. Springer. find
  2. Eugene Charniak and Micha Elsner. 2009. Em works for pronoun anaphora resolution. In Proceedings of the Conference of the European Chapter of the Association for Computational Linguistics (EACL-09), Athens, Greece. find our card
  3. Noam Chomsky. 1981. Lectures on government and binding. Foris, Dordercht. find
  4. H.H. Clark and CJ Sengul. 1979. In search of referents for nouns and pronouns. Memory & Cognition, 7(1):35–41. find
  5. P. Denis and J. Baldridge. 2007. A ranking approach to pronoun resolution. In Proc. IJCAI. find
  6. Kevin Duh. 2005. Jointly labeling multiple sequences: a factorial HMM approach. In ACL ’05: Proceedings of the ACL Student Research Workshop, pages 19–24, Ann Arbor, Michigan. find
  7. Zoubin Ghahramani and Michael I. Jordan. 1997. Factorial hidden markov models. Machine Learning, 29:1– 31. find
  8. A. Haghighi and D. Klein. 2007. Unsupervised coreference resolution in a nonparametric bayesian model. In Proceedings of the 45th annual meeting on Association for Computational Linguistics, page 848. find our card
  9. A. Haghighi and D. Klein. 2009. Simple coreference resolution with rich syntactic and semantic features. In Proceedings of the 2009 Conference on Empirical Methods in Natural Language Processing: Volume 3-Volume 3, pages 1152–1161. Association for Computational Linguistics. find our card
  10. A. Haghighi and D. Klein. 2010. Coreference resolution in a modular, entity-centered model. In Human Language Technologies: The 2010 Annual Conference of the North American Chapter of the Association for Computational Linguistics, pages 385–393. Association for Computational Linguistics. find our card
  11. L. Hasler, C. Orasan, and K. Naumann. 2006. NPs for events: Experiments in coreference annotation. In Proceedings of the 5th edition of the International Conference on Language Resources and Evaluation (LREC2006), pages 1167–1172. Citeseer. find our card
  12. Dan Klein and Christopher D. Manning. 2003. Accurate unlexicalized parsing. In Proceedings of the 41st Annual Meeting of the Association for Computational Linguistics, pages 423–430, Sapporo, Japan. find our card
  13. X Luo. 2005. On coreference resolution performance metrics. pages 25–32. Association for Computational Linguistics Morristown, NJ, USA. find our card
  14. A. McCallum and B. Wellner. 2003. Toward conditional models of identity uncertainty with application to proper noun coreference. In IJCAI Workshop on Information Integration on the Web. Citeseer. find
  15. David McClosky, Eugene Charniak, and Mark Johnson. 2008. BLLIP North American News Text, Complete. Linguistic Data Consortium. LDC2008T13. find
  16. T.S. Morton. 2000. Coreference for NLP applications. In Proceedings of the 38th Annual Meeting on Association for Computational Linguistics, pages 173–180. Association for Computational Linguistics. find
  17. V. Ng. 2008. Unsupervised models for coreference resolution. In Proceedings of the Conference on Empirical Methods in Natural Language Processing, pages 640– 649. Association for Computational Linguistics. find our card
  18. US NIST. 2003. The ACE 2003 Evaluation Plan. US National Institute for Standards and Technology (NIST), Gaithersburg, MD.[online, pages 2003–08. find
  19. L. Qiu, M.Y. Kan, and T.S. Chua. 2004. A public reference implementation of the rap anaphora resolution algorithm. Arxiv preprint cs/0406031. find our card
  20. X. Yang, J. Su, G. Zhou, and C.L. Tan. 2004. Improving pronoun resolution by incorporating coreferential information of candidates. In Proceedings of the 42nd Annual Meeting on Association for Computational Linguistics, page 127. Association for Computational Linguistics. find our card

↑ back to top

This page in Markdown →