Elamur Research⌕
☰
home / papers / 2005 / Multi-Lingual Coreference Resolution With Syntactic Features

Multi-Lingual Coreference Resolution With Syntactic Features

Summary

The paper examines whether structural cues from automatically generated parse trees can improve statistical coreference resolution in several languages. Luo and Zitouni (IBM T.J. Watson) build on a maximum entropy entity-mention framework. In it, a link model scores attaching a mention to an existing entity, and a start model scores creating a new entity. Instead of writing rules and weighting them by hand, they convert parse-tree structure into trainable features, so the learner can assign low weight to signals coming from noisy parses.

Guided by binding theory, they approximate governing category indirectly. They use a c-command test plus counts of NPs, VPs and clauses along the c-command path. They also add dependency features (shared heads and head POS at one or two levels, direct modification) and apposition/same-parent features.

Experiments use ACE data for Arabic, Chinese and English. The features give statistically significant F-measure gains for Arabic and English but barely affect Chinese, which the authors link to properties such as pro-drop. The study shows that linguistically motivated features designed on English can carry over to other languages when learned from machine parses.

Key points

  • Binding-theory features: binary c-command (ccmd) between the minimal NPs of a candidate and a pronoun, plus NPcount, VPcount and Scount along the c-command path. These implicitly encode governing category and clause boundaries.
  • Parsers are statistical models trained on treebanks: the Penn Treebank (English), the Chinese treebank of Xia et al. (2000), and the Arabic treebank of Maamouri and Bies (2004).
  • ACE 2004 blind test, F-measure (baseline → +synt):
    • Arabic: 80.1 → 81.5
    • Chinese: 84.7 → 84.7
    • English: 80.6 → 82.0
    • Test sets contain about 10–11K mentions per language.
  • ACE-Value gains are smaller (Arabic 88.0 → 88.9, Chinese 92.7 → 92.8, English 90.9 → 91.6) because the metric heavily discounts pronouns. For this reason, contrastive results are reported with F-measure.
  • The improvement is concentrated on targeted mention types: English pronoun F-measure rises from 69.5% to 73.7% on the ACE 2004 test set.

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

12 papers of our library cite it (of the 140673 whose bibliographies we read); 6 of them have a page here, 6 not yet.

Cite

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

@inproceedings{luo2005multi,
  title = {{Multi-Lingual Coreference Resolution With Syntactic Features}},
  author = {Luo, Xiaoqiang and Zitouni, Imed},
  booktitle = {Proceedings of Human Language Technology Conference and Conference on Empirical Methods in Natural Language Processing},
  year = {2005},
  doi = {10.3115/1220575.1220658},
  url = {https://doi.org/10.3115/1220575.1220658}
}

References

25 references, 13 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Amit Bagga and Breck Baldwin. 1998. Algorithms for scoring coreference chains. In Proceedings of the Linguistic Coreference Workshop at The First International Conference on Language Resources and Evaluation (LREC’98), pages 563–566. find our card
  2. Santorini Beatrice and Anthony Kroch. 2000. The syntax of natural language: An online introduction using the Trees program. www.ling.upenn.edu/beatrice/syntax-textbook. find
  3. 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
  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. Liliane Haegeman. 1994. Introduction to Government and Binding Theory. Basil Blackwell Inc., 2nd edition. find
  6. J. Hobbs. 1976. Pronoun resolution. Technical report, Dept. of Computer Science, CUNY, Technical Report TR76-1. find
  7. C.-T. James Huang. 1984. On the distribution and reference of empty pronouns. Linguistic Inquiry, 15:531–574. find
  8. Andrew Kehler, Douglas Appelt, Lara Taylor, and Aleksandr Simma. 2004. The (Non)utility of predicate-argument frequencies for pronoun interpretation. In Daniel Marcu Susan Dumais and Salim Roukos, editors, HLT-NAACL 2004: Main Proceedings, Boston, Massachusetts, USA, May 2 - May 7. Association for C… find
  9. Shalom Lappin and Herbert J. Leass. 1994. An algorithm for pronominal anaphora resolution. Computational Linguistics, 20(4), December. find our card
  10. Xiaoqiang Luo, Abe Ittycheriah, Hongyan Jing, Nanda Kambhatla, and Salim Roukos. 2004. A mention-synchronous coreference resolution algorithm based on the bell tree. In Proc. of ACL. find our card
  11. Xiaoqiang Luo. 2005. On coreference resolution performance metrics. In Procs. of HLT/EMNLP. find our card
  12. Mohamed Maamouri and Ann Bies. 2004. Developing an Arabic treebank: Methods, guidelines, procedures, and tools. In Proceedings of the Workshop on Computational Approaches to Arabic Script-based Languages, COLING. find our card
  13. M. Marcus, B. Santorini, and M. Marcinkiewicz. 1993. Building a large annotated corpus of English: the Penn treebank. Computational Linguistics, 19(2):313–330. find our card
  14. Michael McCord. 1993. Heuristics for broad-coverage natural language parsing. In Proc. ARPA Human Language Technology Workshop. find
  15. R. Mitkov. 1998. Robust pronoun resolution with limited knowledge. In Procs. of the 36th ACL/17th COLING, pages 869–875. find our card
  16. Thomas S. Morton. 2000. Coreference for NLP applications. In In Proceedings of the 38th Annual Meeting of the Association for Computational Linguistics. find
  17. Vincent Ng and Claire Cardie. 2002. Improving machine learning approaches to coreference resolution. In Proc. of ACL, pages 104–111. find our card
  18. NIST. 2004. Proceedings of ACE Evaluation and PI Meeting 2004 Workshop, Alexandria, VA, September. find
  19. NIST. 2004. The ACE evaluation plan. www.nist.gov/speech/tests/ace/index.htm. find
  20. Adwait Ratnaparkhi. 1999. Learning to parse natural language with maximum entropy models. Machine Learning, 34:151–178. find
  21. 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
  22. Roland Stuckardt. 2001. Design and enhanced evaluation of a robust anaphor resolution algorithm. Computational Linguistics, 27(4). find
  23. 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
  24. F. Xia, M. Palmer, N. Xue, M.E. Okurowski, J. Kovarik, F.D. Chiou, S. Huang, T. Kroch, and M. Marcus. 2000. Developing guidelines and ensuring consistency for Chinese text annotation. In Proc of the 2nd Intl. Conf. on Language Resources and Evaluation (LREC 2000). find
  25. Xiaofeng Yang, Guodong Zhou, Jian Su, and Chew Lim Tan. 2003. Coreference resolution using competition learning approach. In Proc. of ACL. find our card

↑ back to top

This page in Markdown →