Elamur Research⌕
☰
home / papers / 2004 / An NP-Cluster Based Approach to Coreference Resolution

An NP-Cluster Based Approach to Coreference Resolution

Summary

Most supervised coreference systems decide whether two noun phrases co-refer by classifying NP pairs. The weakness is that a single mention, such as a pronoun or a short demonstrative phrase, often says little about the entity it denotes. Yang et al. instead link each new NP to an already-built coreferential cluster, which pools descriptive evidence from all earlier mentions. Each instance combines the current NP, a candidate cluster and a reference NP from that cluster. The model uses 24 features: the 18 pairwise features of a Soon et al. (2001)-style baseline, plus six cluster-level ones covering agreement, cluster size and string similarity. Both systems use C5.0 decision trees and are tested on biomedical MEDLINE abstracts. Compared with the NP-NP baseline, the cluster approach raises recall by 4.6% and precision by 1.3%. String matching against the whole cluster, or against its longest member, contributes most of the gain. The paper presents this as the first supervised method to resolve coreference by relating NPs to clusters rather than to single antecedents.

Key points

  • Baseline: an NP-NP classifier with 18 features, including two new token-overlap measures, StrSim1 and StrSim2. Links are chosen by Most Recent First (MRF) or Best First (BF) selection, with a confidence threshold of 0.5.
  • New cluster features: Cluster NumAgree, Cluster GenAgree, Cluster SemAgree, Cluster Length, Cluster StrSim and Cluster StrLNPSim. The last one compares the NP with the cluster's longest member.
  • Feature ablation (BF strategy):
    • The baseline-equivalent setting scores R 80.3 / P 77.5 / F 78.9.
    • Adding Cluster StrSim alone raises F by 2.7%.
    • Adding Cluster StrLNPSim alone raises F by 2.1%.
    • Combining the two gives the best result: R 84.9 / P 78.9 / F 81.8.
  • Cluster Length adds only 0.1% F and gives no benefit, or slightly lowers F, when combined with the string features.
  • Limitation: clusters are built incrementally, so text following an NP is ignored. The authors plan future work on globally optimal cluster linking.

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

18 papers of our library cite it (of the 140673 whose bibliographies we read); 7 of them have a page here, 11 not yet.

Cite

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

@inproceedings{yang2004np,
  title = {{An NP-Cluster Based Approach to Coreference Resolution}},
  author = {Yang, Xiaofeng and Su, Jian and Zhou, GuoDong and Tan, Chew Lim},
  booktitle = {COLING 2004: Proceedings of the 20th International Conference on Computational Linguistics},
  year = {2004},
  url = {https://aclanthology.org/C04-1033/}
}

References

17 references, 8 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. C. Aone and S. W. Bennett. 1995. Evaluating automated and manual acquistion of anaphora resolution strategies. In Proceedings of the 33rd Annual Meeting of the Association for Compuational Linguistics, pages 122–129. find
  2. A. Bagga and A. Biermann. 1998. Entity-based cross document coreferencing using the vector space model. In Proceedings of the 36th Annual Meeting of the Association for Computational Linguisticsthe 17th International Conference on Computational Linguistics, pages 79–85. find
  3. C. Cardie and K. Wagstaff. 1999. Noun phrase coreference as clustering. In Proceedings of the Joint Conference on Empirical Methods in NLP and Very Large Corpora. find our card
  4. J. Castano, J. Zhang, and J. Pustejovsky. 2002. Anaphora resolution in biomedical literature. In International Symposium on Reference Resolution, Alicante, Spain. find
  5. C. Gooi and J. Allan. 2004. Cross-document coreference on a large scale corpus. In Proceedings of 2004 Human Language Technology conference / North American chapter of the Association for Computational Linguistics annual meeting. find our card
  6. S. Harabagiu, R. Bunescu, and S. Maiorano. 2001. Text knowledge mining for coreference resolution. In Proceedings of the 2nd Annual Meeting of the North America Chapter of the Association for Compuational Linguistics, pages 55–62. find
  7. S. Lappin and H. Leass. 1994. An algorithm for pronominal anaphora resolution. Computational Linguistics, 20(4):525–561. find our card
  8. J. McCarthy and Q. Lehnert. 1995. Using decision trees for coreference resolution. In Proceedings of the 14th International Conference on Artificial Intelligences, pages 1050–1055. find
  9. V. Ng and C. Cardie. 2002a. Combining sample selection and error-driven pruning for machine learning of coreference rules. In Proceedings of the conference on Empirical Methods in Natural Language Processing, pages 55–62, Philadelphia. find our card
  10. V. Ng and C. Cardie. 2002b. 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. J. R. Quinlan. 1993. C4.5: Programs for machine learning. Morgan Kaufmann Publishers, San Francisco, CA. find our card
  12. D. Shen, J. Zhang, G. Zhou, J. Su, and C. Tan. 2003. Effective adaptation of hidden markov model-based named-entity recognizer for biomedical domain. In Proceedings of ACL03 Workshop on Natural Language Processing in Biomedicine, Japan. find
  13. 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
  14. M. Strube, S. Rapp, and C. Muller. 2002. The influence of minimum edit distance on reference resolution. In Proceedings of the Conference on Empirical Methods in Natural Language Processing, pages 312–319, Philadelphia. find
  15. 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
  16. X. Yang, G. Zhou, J. Su, and C. Tan. 2004. Improving noun phrase coreference resolution by matching strings. In Proceedings of the 1st International Joint Conference on Natural Language Processing, Hainan. find
  17. G. Zhou and J. Su. 2002. Named Entity recognition using a HMM-based chunk tagger. In Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics, Philadelphia. find

↑ back to top

This page in Markdown →