Elamur Research⌕
☰
home / papers / 2010 / Dependency-driven Anaphoricity Determination for Coreference Resolution

Dependency-driven Anaphoricity Determination for Coreference Resolution

Summary

Coreference systems waste effort and make errors when they search for antecedents of noun phrases that actually introduce new entities. Kong et al. address anaphoricity determination, which means deciding in advance whether a mention is anaphoric, with a tree kernel classifier over syntactic parses. Their main contribution is a way to prune the full parse tree around each mention. Guided by constituent dependencies seen from both syntactic and semantic angles, the method keeps the nodes and paths relevant to the mention and discards noisy material. The result is a dependency-driven dynamic syntactic parse tree (D-DSPT). On the ACE 2003 corpus, D-DSPT outperforms earlier tree structures on the anaphoricity task. When the resulting classifier is placed in front of a learning-based coreference resolver, F1 rises clearly in all three domains, giving the best reported results at the time. The study shows that carefully shaped structural input matters for kernel methods, and that an explicit anaphoricity filter can substantially help coreference resolution.

Key points

  • The D-DSPT module improves coreference F1 over a baseline without anaphoricity determination by 4.1, 3.9 and 5.0 points, reaching 63.2 (NWIRE), 67.4 (NPAPER) and 61.6 (BNEWS) on ACE 2003. The gains come mostly from higher precision with only a small loss in recall.
  • Gold-standard anaphoricity would add 6.4, 6.1 and 7.0 F1 over the baseline. D-DSPT therefore still trails this upper bound by roughly 2 points per domain (2.3, 2.2, 2.0).
  • Improvements appear for pronominal, definite and indefinite NPs. Proper NPs gain almost nothing because the baseline's abbreviation feature already handles them well.
  • The system beats Zhou and Kong (2009), which used a dynamically-extended tree, by 1.4, 2.2 and 2.0 F1. It also outperforms Ng (2009), which used graph minimum cuts.
  • The authors present this as the first work to exploit constituent dependencies directly in tree kernel-based anaphoricity determination. The approach builds on the convolution tree kernel and on the dynamic parse trees of Qian et al. (2008) for relation extraction.

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

Cite

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

@inproceedings{kong2010dependency,
  title = {{Dependency-driven Anaphoricity Determination for Coreference Resolution}},
  author = {Kong, Fang and Zhou, Guodong and Qian, Longhua and Zhu, Qiaoming},
  booktitle = {Proceedings of the 23rd International Conference on Computational Linguistics (Coling 2010)},
  year = {2010},
  url = {https://aclanthology.org/C10-1068/}
}

References

31 references, 14 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. D. Bean and E. Riloff 1999. Corpus-based Identification of Non-Anaphoric Noun Phrases. find our card
  2. S. Bergsma, D. Lin and R. Goebel 2008. Distributional Identification of Non-referential Pronouns. ACL’2008 find
  3. C. Cherry and S. Bergsma. 2005. An expectation maximization approach to pronoun resolution. CoNLL’2005 find our card
  4. M. Collins and N. Duffy. 2001. Covolution kernels for natural language. NIPS̓2001 find
  5. M. Denber 1998. Automatic Resolution of Anapho-ria in English. Technical Report, Eastman Ko-dakCo. find
  6. P. Denis and J. Baldridge. 2007. Global, joint determination of anaphoricity and coreference resolution using integer programming. NAACL/HLT’2007 find our card
  7. R. Evans 2001. Applying machine learning toward an automatic classification of it. Literary and Linguistic Computing, 16(1):45-57 find
  8. F. Kong, G.D. Zhou and Q.M. Zhu. 2009 Employing the Centering Theory in Pronoun Resolution from the Semantic Perspective. EMNLP’2009 find our card
  9. F. Kong, Y.C. Li, G.D. Zhou and Q.M. Zhu. 2009. Exploring Syntactic Features for Pronoun Resolution Using Context-Sensitive Convolution Tree Kernel. IALP’2009 find
  10. S. Lappin and J. L. Herbert. 1994. An algorithm for pronominal anaphora resolution. Computational Linguistics, 20(4) find our card
  11. J.H. Li. G.D. Zhou, H. Zhao, Q.M. Zhu and P.D. Qian. Improving nominal SRL in Chinese language with verbal SRL information and automatic predicate recognition. EMNLP '2009 find
  12. X. Luo. 2007. Coreference or not: A twin model for coreference resolution. NAACL-HLT’2007 find our card
  13. V. Ng and C. Cardie 2002. Identify Anaphoric and Non-Anaphoric Noun Phrases to Improve Coreference Resolution. COLING’2002 find
  14. V. Ng and C. Cardie 2002. Improving machine learning approaches to coreference resolution. ACL’2002 find our card
  15. V. Ng 2004. Learning Noun Phrase Anaphoricity to Improve Coreference Resolution: Issues in Representation and Optimization. find
  16. V. Ng 2009. Graph-cut based anaphoricity determination for coreference resolution. NAACL’2009 find our card
  17. L.H. Qian, G.D. Zhou, F. Kong, Q.M. Zhu and P.D. Qian. 2008. Exploiting constituent dependencies for tree kernel-based semantic relation extraction. COLING’2008 find
  18. W. M. Soon, H. T. Ng and D. Lim 2001. A machine learning approach to coreference resolution of noun phrase. Computational Linguistics, 27(4):521-544. find
  19. V. Stoyanov, N. Gilbert, C. Cardie and E. Riloff. 2009. Conundrums in Noun Phrase Coreference Resolution: Making Sense of the State-of-the Art. ACL’2009 find our card
  20. B. L. Webber. 1979. A Formal Approach to Discourse Anaphora. Garland Publishing, Inc. find
  21. X.F. Yang, G.D. Zhou, J. Su and C.L. Chew. 2003. Coreference Resolution Using Competition Learning Approach. ACL’2003 find our card
  22. X.F. Yang, J. Su and C.L. Chew. 2005. A Twin Candidate Model of Coreference Resolution with Non-Anaphor Identification Capability. IJCNLP’2005 find our card
  23. X.F. Yang, J. Su and C.L. Chew. 2006. Kernel-based pronoun resolution with structured syntactic knowledge. COLING-ACL’2006 find our card
  24. X.F. Yang, J. Su and C.L. Tan 2008. A Twin-Candidate Model for Learning-Based Anaphora Resolution. Computational Linguistics 34(3):327-356 find our card
  25. M. Zhang, J. Zhang, J. Su and G.D. Zhou. 2006. A composite kernel to extract relations between entities with both flat and structured features. COLING/ACL’2006 find
  26. S. Zhao and R. Grishman. 2005. Extracting relations with integered information using kernel methods. ACL’2005 find
  27. D. Zelenko, A. Chinatsu and R. Anthony. 2003. Kernel methods for relation extraction. Machine Learning Researching 3(2003):1083-1106 find
  28. G.D. Zhou, F. Kong and Q.M. Zhu. 2008. Context-sensitive convolution tree kernel for pronoun resolution. IJCNLP’2008 find our card
  29. G.D. Zhou and F. Kong. 2009. Global Learning of Noun Phrase Anaphoricity in Coreference Resolution via Label Propagetion. EMNLP’2009 find
  30. G.D. Zhou and J. Su. 2002. Named Entity recognition using a HMM-based chunk tagger. ACL’2002 find
  31. G.D. Zhou, M. Zhang, D.H. Ji and Q.M. Zhu. 2007. Tree kernel-based relation extraction with context-sensitive structured parse tree information. EMNLP/CoNLL’2007 find

↑ back to top

This page in Markdown →