Elamur Research⌕
☰
home / papers / 2008 / Specialized Models and Ranking for Coreference Resolution

Specialized Models and Ranking for Coreference Resolution

Summary

The paper looks at whether supervised coreference resolution can be improved without moving to complex joint or global inference. Denis and Baldridge change the standard mention-pair classifier pipeline in two ways. First, they replace binary classification with a log-linear ranker, trained with TADM, that scores all candidate antecedents of a mention against each other. Second, they train separate models for different kinds of referring expressions, drawing on linguistic accessibility theory. Rankers always choose some antecedent, so a discourse-status (anaphoricity) classifier first decides which mentions should be sent to them. On the ACE corpus, the combined system (specialized rankers plus the discourse-status filter) beats an otherwise identical classifier baseline by more than 3 F-points on MUC, B³ and CEAF. The authors credit the gains to sharper antecedent selection and to a better division of labor across mention types. The work shows that simple, efficient local models can rival heavier global approaches. Oracle experiments point to the discourse-status component as the main remaining bottleneck.

Key points

  • The specialized models cover five mention types: third-person pronouns, speech pronouns, proper names, definite descriptions, and others. ACE has 19,322 training anaphors and 4,599 test anaphors.
  • The specialized ranker system (RANK+DS+SP) improves on the standard classifier by 3.2 F-points on MUC, 3.1 on B³ and 3.6 on CEAF.
  • The discourse-status classifier reaches 80.8% accuracy, compared with a 59.7% majority-class baseline.
  • Wrong antecedents for true anaphors drop to 17.9% with RANK+DS+SP, against 23.1% for the plain CLASS baseline.
  • Using perfect discourse status (an oracle) gives large further gains. With a perfect resolver but the real discourse-status classifier, F-scores are capped at 88.1 (MUC), 85.2 (B³) and 79.7 (CEAF).

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: Discourse analysis · Coreference resolution

Cited by

54 papers of our library cite it (of the 140673 whose bibliographies we read); 31 of them have a page here, 23 not yet.

Cite

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

@inproceedings{denis2008specialized,
  title = {{Specialized Models and Ranking for Coreference Resolution}},
  author = {Denis, Pascal and Baldridge, Jason},
  booktitle = {Proceedings of the 2008 Conference on Empirical Methods in Natural Language Processing},
  year = {2008},
  doi = {10.3115/1613715.1613797},
  url = {https://doi.org/10.3115/1613715.1613797}
}

References

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

  1. M. Ariel. 1988. Referring and accessibility. Journal of Linguistics, pages 65–87. find
  2. N. Asher, P. Denis, and B. Reese. 2006. Names and pops and discourse structure. In Workshop on Constraints in Discourse, Maynooth, Ireland. find
  3. A. Bagga and B. Baldwin. 1998. Algorithms for scoring coreference chains. In Proceedings of LREC 1998, pages 563–566. find our card
  4. We strongly advocate that coreference results should never be presented in terms of MUC scores alone. find
  5. E. Charniak and M. Johnson. 2005. Coarse-to-fine nbest parsing and maxent discriminative reranking. In Proceedings of ACL 2005, Ann Arbor, Michigan. find
  6. M. Collins and N. Duffy. 2002. New ranking algorithms for parsing and tagging: Kernels over discrete structures and the voted perceptron. In Proceedings of ACL 2002, pages 263–270, Philadelphia, PA. find
  7. H. Daumé III and D. Marcu. 2005. A large-scale exploration of effective global features for a joint entity detection and tracking model. In Proceedings of HLT-EMNLP 2005, Vancouver, Canada. find
  8. P. Denis and J. Baldridge. 2007a. Joint determination of anaphoricity and coreference resolution using integer programming. In Proceedings of HLT-NAACL 2007, Rochester, NY. find our card
  9. P. Denis and J. Baldridge. 2007b. A ranking approach to pronoun resolution. In Proceedings of IJCAI 2007, Hyderabad, India. find
  10. Pascal Denis. 2007. New Learning Models for Robust Reference Resolution. Ph.D. thesis, The University of Texas at Austin. find
  11. R. Elwell and J. Baldridge. 2008. Discourse connective argument identification with connective specific rankers. In Proceedings of the International Conference on Semantic Computing, Santa Clara, CA. find
  12. B. Grosz, A. Joshi, and S. Weinstein. 1995. Centering: A framework for modelling the local coherence of discourse. Computational Linguistics, 2(21). find
  13. J. K. Gundel, N. Hedberg, and R. Zacharski. 1993. Cognitive status and the form of referring expressions in discourse. Language, 69:274–307. find
  14. A. Haghighi and D. Klein. 2007. Unsupervised coreference resolution in a nonparametric Bayesian model. In Proceedings ACL 2007, pages 848–855, Prague, Czech Republic. find our card
  15. A. Kehler, D. Appelt, L. Taylor, and A. Simma. 2004. The (non)utility of predicate-argument frequencies for pronoun interpretation. In Proceedings of HLT-NAACL 2004. find
  16. X. Luo. 2005. On coreference resolution performance metrics. In Proceedings of HLT-NAACL 2005, pages 25–32. find our card
  17. R. Malouf. 2002. A comparison of algorithms for maximum entropy parameter estimation. In Proceedings of the Sixth Workshop on Natural Language Learning, pages 49–55, Taipei, Taiwan. find our card
  18. A. McCallum and B. Wellner. 2004. Conditional models of identity uncertainty with application to noun coreference. In Proceedings of NIPS 2004. find
  19. J. F. McCarthy and W. G. Lehnert. 1995. Using decision trees for coreference resolution. In IJCAI, pages 1050–1055. find
  20. T. Morton. 2000. Coreference for NLP applications. In Proceedings of ACL 2000, Hong Kong. find
  21. V. Ng and C. Cardie. 2002a. Identifying anaphoric and non-anaphoric noun phrases to improve coreference resolution. In Proceedings of COLING 2002. find
  22. V. Ng and C. Cardie. 2002b. Improving machine learning approaches to coreference resolution. In Proceedings of ACL 2002, pages 104–111. find our card
  23. V. Ng. 2004. Learning noun phrase anaphoricity to improve coreference resolution: Issues in representation and optimization. In Proceedings of ACL 2004. find
  24. V. Ng. 2005a. Machine learning for coreference resolution: From local classification to global ranking. In Proceedings of ACL 2005, pages 157–164, Ann Arbor, MI. find our card
  25. V. Ng. 2005b. Supervised ranking for pronoun resolution: Some recent improvements. In Proceedings of AAAI 2005. find
  26. V. Ng. 2007. Semantic class induction and coreference resolution. In Proceedings of ACL 2007. find our card
  27. M. Osborne and J. Baldridge. 2004. Ensemble-based active learning for parse selection. In Proceedings of HLT-NAACL 2004, pages 89–96, Boston, MA. find
  28. E. F. Prince. 1981. Toward a taxonomy of given-new information. In P. Cole, editor, Radical Pragmatics, pages 223–255. Academic Press, New York. find
  29. D. Ravichandran, E. Hovy, and F. J. Och. 2003. Statistical QA - classifier vs re-ranker: What’s the difference? In Proceedings of the ACL Workshop on Multilingual Summarization and Question Answering– Machine Learning and Beyond. find
  30. W. M. Soon, H. T. Ng, and D. Lim. 2001. A machine learning approach to coreference resolution of noun phrases. Computational Linguistics, 27(4):521–544. find our card
  31. K. Toutanova, P. Markova, and C. Manning. 2004. The leaf projection path view of parse trees: Exploring string kernels for HPSG parse selection. In Proceedings of EMNLP 2004, pages 166–173, Barcelona. find
  32. O. Uryupina. 2004. Linguistically motivated sample selection for coreference resolution. In Proceedings of DAARC 2004, Furnas. find
  33. R. van der Sandt. 1992. Presupposition projection as anaphora resolution. Journal of Semantics, 9:333– 377. find
  34. M. Vilain, J. Burger, J. Aberdeen, D. Connolly, and L. Hirschman. 1995. A model-theoretic coreference scoring scheme. In Proceedings fo the 6th Message Understanding Conference (MUC-6), pages 45– 52, San Mateo, CA. Morgan Kaufmann. find our card
  35. X. Yang, G. Zhou, J. Su, and C.L. Tan. 2003. Coreference resolution using competitive learning approach. In Proceedings of ACL 2003, pages 176–183. find

↑ back to top

This page in Markdown →