Elamur Research⌕
☰
home / papers / 2004 / A Mention-Synchronous Coreference Resolution Algorithm Based On the Bell Tree

A Mention-Synchronous Coreference Resolution Algorithm Based On the Bell Tree

Summary

Coreference resolution means grouping the mentions in a document (names, nominals and pronouns) into entities. Earlier statistical systems first scored pairs of mentions and then merged them greedily. Each linking decision was made locally and never revisited. Luo et al. instead treat the whole process as a search over the Bell tree. Mentions are added one at a time from left to right, and each one either joins an existing partial entity or starts a new one. Every leaf of the tree is one possible partition of the mentions, so resolving coreference becomes finding the highest-scoring path from the root to a leaf. A binary Maximum Entropy model gives the probability of linking a mention to an entity. The probability of starting a new entity comes from those linking scores, adjusted by a tunable start penalty. A beam search with pruning keeps the search tractable. The authors compare a mention-pair model with an entity-mention model and test both on ACE 2002/2003 data and MUC6. They report state-of-the-art ACE results in three languages and competitive MUC6 results. The paper gives an incremental, globally ranked alternative to greedy pairwise linking.

Key points

  • The search space is the Bell tree. Its number of leaves equals the Bell number of the mention count, which grows very quickly; for example, three mentions already give five possible partitions.
  • Every hypothesis takes the same number of link or start steps. This makes it fair to compare paths by multiplying their local probabilities, and it lets mentions be added incrementally.
  • No separate model for starting a new entity is trained. That probability is derived from the linking probabilities, and a start penalty trades off missed entities against false alarms.
  • Two Maximum Entropy variants are compared:
    • a mention-pair model, where a mention's score against an entity is its best score against any single mention in that entity;
    • an entity-mention model, which computes features over the whole entity, such as a canonical mention and the minimum distance.
  • Features cover string match, acronyms, edit distance, distances in tokens, sentences and mentions, POS and apposition from a parser, counts, and pronoun attributes.

The results section was cut from the text you provided, so I left out specific scores rather than guess them.

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

84 papers of our library cite it (of the 140673 whose bibliographies we read); 49 of them have a page here, 35 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.3115/1218955.1218973

@inproceedings{luo2004mention,
  title = {{A Mention-Synchronous Coreference Resolution Algorithm Based On the Bell Tree}},
  author = {Luo, Xiaoqiang and Ittycheriah, Abe and Jing, Hongyan and Kambhatla, Nanda and Roukos, Salim},
  booktitle = {Proceedings of the 42nd Annual Meeting of the Association for Computational Linguistics (ACL-04)},
  year = {2004},
  doi = {10.3115/1218955.1218973},
  url = {https://doi.org/10.3115/1218955.1218973}
}

References

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

  1. E.T. Bell. 1934. Exponential numbers. Amer. Math. Monthly, pages 411–419. find
  2. Adam L. Berger, Stephen A. Della Pietra, and Vincent J. Della Pietra. 1996. A maximum entropy approach to natural language processing. Computational Linguistics, 22(1):39–71, March. find our card
  3. R Florian, H Hassan, A Ittycheriah, H Jing, N Kambhatla, X Luo, N Nicolov, and S Roukos. 2004. A statistical model for multilingual entity detection and tracking. In Daniel Marcu Susan Dumais and Salim Roukos, editors, HLT-NAACL 2004: Main Proceedings, pages 1–8, Boston, Massachusetts, USA, May 2 -… find
  4. Niyu Ge, John Hale, and Eugene Charniak. 1998. A statistical approach to anaphora resolution. In Proc. of the sixth Workshop on Very Large Corpora. find our card
  5. Sanda M. Harabagiu, Razvan C. Bunescu, and Steven J. Maiorano. 2001. Text and knowledge mining for coreference resolution. In Proc. of NAACL. find
  6. J. Hobbs. 1976. Pronoun resolution. Technical report, Dept. of Computer Science, CUNY, Technical Report TR76-1. find
  7. A. Ittycheriah, L. Lita, N. Kambhatla, N. Nicolov, S. Roukos, and M. Stys. 2003. Identifying and tracking entity mentions in a maximum entropy framework. In HLT-NAACL 2003: Short Papers, May 27 - June 1. find
  8. Andrew Kehler. 1997. Probabilistic coreference in information extraction. In Proc. of EMNLP. find our card
  9. Andrew McCallum and Ben Wellner. 2003. Toward conditional models of identity uncertainty with application to proper noun coreference. In IJCAI Workshop on Information Integration on the Web. find
  10. R. Mitkov. 1998. Robust pronoun resolution with limited knowledge. In Procs. of the 17th Internaltional Conference on Computational Linguistics, pages 869–875. find our card
  11. Thomas S. Morton. 2000. Coreference for NLP applications. In In Proceedings of the 38th Annual Meeting of the Association for Computational Linguistics. find
  12. MUC-6. 1995. Proceedings of the Sixth Message Understanding Conference(MUC-6), San Francisco, CA. Morgan Kaufmann. find
  13. Vincent Ng and Claire Cardie. 2002. Improving machine learning approaches to coreference resolution. In Proc. of ACL, pages 104–111. find our card
  14. NIST. 2003. The ACE evaluation plan. www.nist.gov/speech/tests/ace/index.htm. find
  15. Adwait Ratnaparkhi. 1997. A Linear Observed Time Statistical Parser Based on Maximum Entropy Models. In Second Conference on Empirical Methods in Natural Language Processing, pages 1 – 10. find
  16. Wee Meng Soon, Hwee Tou Ng, and Chung Yong Lim. 2001. A machine learning approach to coreference resolution of noun phrases. Computational Linguistics, 27(4):521–544. find our card
  17. M. Vilain, J. Burger, J. Aberdeen, D. Connolly, , and L. Hirschman. 1995. A model-theoretic coreference scoring scheme. In In Proc. of MUC6, pages 45–52. find our card
  18. Xiaofeng Yang, Guodong Zhou, Jian Su, and Chew Lim Tan. 2003. Coreference resolution using  competition learning approach. In Proc. of the ACL. & Q find our card

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/luo-2004-mention-synchronous-coreference-resolution-algorithm/