Elamur Research⌕
☰
home / papers / 2006 / Kernel-Based Pronoun Resolution with Structured Syntactic Knowledge

Kernel-Based Pronoun Resolution with Structured Syntactic Knowledge

Summary

Choosing the right antecedent for a pronoun depends heavily on syntax. Earlier systems, however, reduced parse trees to hand-picked flat features such as grammatical role, and those features may miss useful structure. Yang, Su and Tan let the learner work with the parse tree itself. Each pronoun–candidate pair receives a tree fragment as a structured feature. A convolution tree kernel scores how similar two fragments are. A composite kernel then merges that score with conventional features inside an SVM classifier, following the Soon et al. / Ng and Cardie learning framework. Sentence parses are linked under one root, so pairs that span sentences can still share a tree. The authors compare three ways of cutting out the fragment, varying how much context they keep. On the ACE data, covering three domains, adding the tree feature raises success rates by roughly 5–8%. The gain holds across training sizes and parsers. The work shows that kernels can discover syntactic cues for anaphora without manual feature design.

Key points

  • Three tree-fragment variants are tested: Min-Expansion, Simple-Expansion and Full-Expansion. The Simple version (S Simple) is used in the main comparisons, both alone and combined with the baseline features (NORM+S Simple).
  • Evaluation uses the ACE corpus in three domains: NWire, NPaper and BNews. With Charniak's parser, NORM+S Simple reaches success rates of 79.2, 82.7 and 82.3 respectively.
  • Removing the Subject and Object features from the baseline lowers success by as much as 5.7% on NWire. With the tree feature present, the loss falls to 1.9% (NWire) and 1.1% (BNews), and NPaper improves slightly.
  • The Luo and Zitouni (2005) flat syntactic features, which are based on binding theory and c-command, add at most 1.3%. This gain is smaller than the one from the structured feature.
  • Swapping Charniak (2000) for Collins (1999, Model II) changes results by less than 2% in every domain, so the approach holds up across standard parsers.

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

25 papers of our library cite it (of the 140673 whose bibliographies we read); 10 of them have a page here, 15 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.3115/1220175.1220181

@inproceedings{yang2006kernel,
  title = {{Kernel-Based Pronoun Resolution with Structured Syntactic Knowledge}},
  author = {Yang, Xiaofeng and Su, Jian and Tan, Chew Lim},
  booktitle = {Proceedings of the 21st International Conference on Computational Linguistics and 44th Annual Meeting of the Association for Computational Linguistics},
  year = {2006},
  doi = {10.3115/1220175.1220181},
  url = {https://doi.org/10.3115/1220175.1220181}
}

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. C. Aone and S. W. Bennett. 1995. Evaluating automated and manual acquisition of anaphora resolution strategies. In Proceedings of the 33rd Annual Meeting of the Association for Compuational Linguistics, pages 122–129. find our card
  2. E. Charniak. 2000. A maximum-entropy-inspired parser. In Proceedings of North American chapter of the Association for Computational Linguistics annual meeting, pages 132–139. find our card
  3. M. Collins and N. Duffy. 2002. New ranking algorithms for parsing and tagging: kernels over discrete structures and the voted perceptron. In Proceedings of the 40th Annual Meeting of the Association As in their pulic reports on Section 23 of WSJ TreeBank, Charniak (2000)’s parser achieves 89.6% reca… find
  4. M. Collins. 1999. Head-Driven Statistical Models for Natural Language Parsing. Ph.D. thesis, University of Pennsylvania. find
  5. J. Hobbs. 1978. Resolving pronoun references. Lingua, 44:339–352. find our card
  6. T. Joachims. 1999. Making large-scale svm learning practical. In Advances in Kernel Methods - Support Vector Learning. MIT Press. find
  7. F. Keller and M. Lapata. 2003. Using the web to obtain freqencies for unseen bigrams. Computational Linguistics, 29(3):459–484. find
  8. C. Kennedy and B. Boguraev. 1996. Anaphora for everyone: pronominal anaphra resolution without a parser. In Proceedings of the 16th International Conference on Computational Linguistics, pages 113–118, Copenhagen, Denmark. find
  9. S. Lappin and H. Leass. 1994. An algorithm for pronominal anaphora resolution. Computational Linguistics, 20(4):525–561. find our card
  10. X. Luo and I. Zitouni. 2005. Milti-lingual coreference resolution with syntactic features. In Proceedings of Human Language Techonology conference and Conference on Empirical Methods in Natural Language Processing (HLT/EMNLP), pages 660–667. find
  11. 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
  12. A. Moschitti. 2004. A study on convolution kernels for shallow semantic parsing. In Proceedings of the 42nd Annual Meeting of the Association for Computational Linguistics (ACL’04), pages 335–342. find
  13. 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
  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. V. Vapnik. 1995. The Nature of Statistical Learning Theory. Springer. find
  16. X. Yang, J. Su, G. Zhou, and C. Tan. 2004. Improving pronoun resolution by incorporating coreferential information of candidates. In Proceedings of 42th Annual Meeting of the Association for Computational Linguistics, pages 127–134, Barcelona. find our card
  17. D. Zelenko, C. Aone, and A. Richardella. 2003. Kernel methods for relation extraction. Journal of Machine Learning Research, 3(6):1083 – 1106. find
  18. S. Zhao and R. Grishman. 2005. Extracting relations with integrated information using kernel methods. In Proceedings of 43rd Annual Meeting of the Association for Computational Linguistics (ACL05), pages 419–426. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/papers/yang-2006-kernel-based-pronoun-resolution-structured/