Elamur Research⌕
☰
home / papers / 2004 / Improving Pronoun Resolution by Incorporating Coreferential Information of Candidates

Improving Pronoun Resolution by Incorporating Coreferential Information of Candidates

Summary

Most learning-based pronoun resolvers judge each anaphor–candidate pair in isolation and ignore how salient the candidate is in the surrounding discourse. Inspired by centering theory and Strube's S-List model, Yang et al. add this missing context to a supervised resolver. They extend the Soon et al. (2001) framework, which uses a C5.0 decision tree and 14 knowledge-poor features. Each instance becomes a triple: anaphor, candidate, and the candidate's closest antecedent. The antecedent's properties are encoded as ten new "backward features."

The authors tested the approach on the MUC-6 and MUC-7 coreference corpora in two settings. In the first, gold antecedents were supplied. In the second, a separate noun phrase resolution module found the antecedents automatically. Training the classifier on the same noisy, automatically derived backward features it meets at test time still beat the baseline, even though that NP module had low recall.

The work shows that a candidate's coreference history is a learnable, useful salience signal. It also shows that matching training and testing conditions matters when upstream components are imperfect.

Key points

  • Backward features: Features 15–24 describe the candidate's closest antecedent, covering NP type, NE type, human likelihood, first-NP and subject position, apposition, and a flag for candidates with no antecedent.
  • Oracle setting: With correct antecedents, the success rate rose over the baseline by 4.7% on MUC-6 and 3.5% on MUC-7.
  • Real resolution (RealResolve-3):
    • Success improved by 4.7% on MUC-6 and 1.8% on MUC-7.
    • Recall improved by 2.1% and 2.8%, with no significant loss of precision.
    • Classifiers trained on perfect features (RealResolve-1/2) lost recall and fell below the baseline.
  • Non-pronoun module quality: Recall/precision was 62.2%/78.8% on MUC-6 and 50.1%/75.4% on MUC-7. The model tolerates low recall as long as precision stays high.
  • Classifier refinement: The iterative refining algorithm converged in its second round with an unchanged tree. This is likely because most candidates are non-pronouns (89.1% on MUC-6, 83.7% on MUC-7).

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

13 papers of our library cite it (of the 140673 whose bibliographies we read); 8 of them have a page here, 5 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.3115/1218955.1218972

@inproceedings{yang2004improving,
  title = {{Improving Pronoun Resolution by Incorporating Coreferential Information of Candidates}},
  author = {Yang, Xiaofeng and Su, Jian and Zhou, Guodong and Tan, Chew-Lim},
  booktitle = {Proceedings of the 42nd Annual Meeting of the Association for Computational Linguistics (ACL-04)},
  year = {2004},
  doi = {10.3115/1218955.1218972},
  url = {https://doi.org/10.3115/1218955.1218972}
}

References

19 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. Brennan, M. Friedman, and C. Pollard. 1987. A centering approach to pronouns. In Proceedings of the 25th Annual Meeting of the Association for Compuational Linguistics, pages 155–162. find
  2. N. Ge, J. Hale, and E. Charniak. 1998. A statistical approach to anaphora resolution. In Proceedings of the 6th Workshop on Very Large Corpora. find our card
  3. B. Grosz, A. Joshi, and S. Weinstein. 1983. Providing a unified account of definite noun phrases in discourse. In Proceedings of the 21st Annual meeting of the Association for Computational Linguistics, pages 44–50. find
  4. B. Grosz, A. Joshi, and S. Weinstein. 1995. Centering: a framework for modeling the local coherence of discourse. Computational Linguistics, 21(2):203–225. find our card
  5. R. Iida, K. Inui, H. Takamura, and Y. Matsumoto. 2003. Incorporating contextual cues in trainable models for coreference resolution. In Proceedings of the 10th Conference of EACL, Workshop ”The Computational Treatment of Anaphora”. find
  6. R. Mitkov. 1998. Robust pronoun resolution with limited knowledge. In Proceedings of the 17th Int. Conference on Computational Linguistics, pages 869–875. find our card
  7. R. Mitkov. 1999. Anaphora resolution: The state of the art. Technical report, University of Wolverhampton. find
  8. MUC-6. 1995. Proceedings of the Sixth Message Understanding Conference. Morgan Kaufmann Publishers, San Francisco, CA. find
  9. MUC-7. 1998. Proceedings of the Seventh Message Understanding Conference. Morgan Kaufmann Publishers, San Francisco, CA. find
  10. V. Ng and C. Cardie. 2002. Improving machine learning approaches to coreference resolution. In Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics, pages 104–111, Philadelphia. find our card
  11. M. Paul, K. Yamamoto, and E. Sumita. 1999. Corpus-based anaphora resolution towards antecedent preference. In Proceedings of the 37th Annual Meeting of the Association for Computational Linguistics, Workshop ”Coreference and It’s Applications”, pages 47–52. find
  12. J. R. Quinlan. 1993. C4.5: Programs for machine learning. Morgan Kaufmann Publishers, San Francisco, CA. find our card
  13. C. Sidner. 1981. Focusing for interpretation of pronouns. American Journal of Computational Linguistics, 7(4):217–231. find
  14. W. Soon, H. Ng, and D. Lim. 2001. A machine learning approach to coreference resolution of noun phrases. Computational Linguistics, 27(4):521–544. find our card
  15. M. Strube and C. Muller. 2003. A machine learning approach to pronoun resolution in spoken dialogue. In Proceedings of the 41st Annual Meeting of the Association for Computational Linguistics, pages 168–175, Japan. find our card
  16. M. Strube. 1998. Never look back: An alternative to centering. In Proceedings of the 17th Int. Conference on Computational Linguistics and 36th Annual Meeting of ACL, pages 1251–1257. find
  17. J. R. Tetreault. 2001. A corpus-based evaluation of centering and pronoun resolution. Computational Linguistics, 27(4):507–520. find our card
  18. M. Vilain, J. Burger, J. Aberdeen, D. Connolly, and L. Hirschman. 1995. A model-theoretic coreference scoring scheme. In Proceedings of the Sixth Message understanding Conference (MUC-6), pages 45–52, San Francisco, CA. Morgan Kaufmann Publishers. find our card
  19. X. Yang, G. Zhou, J. Su, and C. Tan. 2003. Coreference resolution using competition learning approach. In Proceedings of the 41st Annual Meeting of the Association for Computational Linguistics, Japan. find our card

↑ back to top

This page in Markdown →