Elamur Research⌕
☰
home / papers / 2006 / Bootstrapping Path-Based Pronoun Resolution

Bootstrapping Path-Based Pronoun Resolution

Summary

The paper targets a gap in third-person pronoun resolution. Standard systems rely on gender/number agreement and syntactic constraints, so they cannot distinguish pairs like "John needs his friend" (coreferent) from "John needs his support" (not coreferent). Bergsma and Lin propose path coreference, which scores the dependency path between a candidate noun and a pronoun. Each path's coreference likelihood is learned by bootstrapping from a large news corpus parsed with Minipar. The method counts how often a path's endpoints agree or clash in gender/number and turns these counts into a probability. Highly coreferent paths are then reused to mine probabilistic gender/number statistics for nouns, which in turn refine the path counts. These statistical features are combined with conventional cues in an SVM classifier that ranks antecedent candidates. Evaluated on the ANC, AQT and MUC datasets, the new features give statistically significant improvements on the two larger sets. The approach captures world knowledge and corrects some parser errors that Binding Theory alone cannot handle.

Key points

  • Path coreference score. For each path p, C(p) = N_S(p) / (N_S(p) + N_D(p)), where N_S and N_D count same-group and different-group terminals. Pronouns are split into seven gender/number/person groups.
  • Path normalisation and two-stage extraction.
    • Paths use lemmatised verbs and nouns, and named entities are replaced by their Minipar category.
    • Off-path modifiers are dropped, and paths are capped at eight nodes.
    • Extraction first uses pronoun–pronoun paths, then re-bootstraps with noun–pronoun paths.
  • Gender/number lexicon. Bootstrapping from coreferent paths yields what the authors call the most complete and accurate probabilistic gender table available at the time. Gender turned out to be the strongest single probabilistic feature.
  • Results. On ANC and AQT, each probabilistic feature produced a significant gain under McNemar's test (p=0.05). On the smaller MUC set the differences were not significant, although the relative contributions of the features were consistent.
  • Extensions. Preliminary experiments suggest path statistics can also help detect pleonastic pronouns and cataphora. Expanding paths with distributionally similar words did not help.

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

64 papers of our library cite it (of the 140673 whose bibliographies we read); 30 of them have a page here, 34 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.3115/1220175.1220180

@inproceedings{bergsma2006bootstrapping,
  title = {{Bootstrapping Path-Based Pronoun Resolution}},
  author = {Bergsma, Shane and Lin, Dekang},
  booktitle = {Proceedings of the 21st International Conference on Computational Linguistics and 44th Annual Meeting of the Association for Computational Linguistics},
  year = {2006},
  doi = {10.3115/1220175.1220180},
  url = {https://doi.org/10.3115/1220175.1220180}
}

References

20 references, 12 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Chinatsu Aone and Scott William Bennett. 1995. Evaluating automated and manual acquisition of anaphora resolution strategies. In Proceedings of the 33rd Annual Meeting of the Association for Computational Linguistics, pages 122– 129. find our card
  2. Catalina Barbu and Ruslan Mitkov. 2001. Evaluation tool for rule-based anaphora resolution methods. In Proceedings of the 39th Annual Meeting of the Association for Computational Linguistics, pages 34–41. find our card
  3. David L. Bean and Ellen Riloff. 2004. Unsupervised learning of contextual role knowledge for coreference resolution. In HLT-NAACL, pages 297–304. find our card
  4. Shane Bergsma. 2005. Automatic acquisition of gender information for anaphora resolution. In Proceedings of the Eighteenth Canadian Conference on Artificial Intelligence (Canadian AI’2005), pages 342–353. find
  5. Colin Cherry and Shane Bergsma. 2005. An expectation maximization approach to pronoun resolution. In Proceedings of the Ninth Conference on Natural Language Learning (CoNLL-2005), pages 88–95. find our card
  6. Kenneth Ward Church and Patrick Hanks. 1989. Word association norms, mutual information, and lexicography. In Proceedings of the 27th Annual Meeting of the Association for Computational Linguistics (ACL’89), pages 76–83. find our card
  7. Ido Dagan and Alan Itai. 1990. Automatic processing of large corpora for the resolution of anaphora references. In Proceedings of the 13th International Conference on Computational Linguistics (COLING-90), volume 3, pages 330–332, Helsinki, Finland. find our card
  8. Niyu Ge, John Hale, and Eugene Charniak. 1998. A statistical approach to anaphora resolution. In Proceedings of the Sixth Workshop on Very Large Corpora, pages 161–171. find our card
  9. Liliane Haegeman. 1994. Introduction to Government & Binding theory: Second Edition. Basil Blackwell, Cambridge, UK. find
  10. Thorsten Joachims. 1999. Making large-scale SVM learning practical. In B. Schölkopf and C. Burges, editors, Advances in Kernel Methods. MIT-Press. find
  11. Andrew Kehler, Douglas Appelt, Lara Taylor, and Aleksandr Simma. 2004. The (non)utility of predicate-argument frequencies for pronoun interpretation. In Proceedings of HLT/NAACL-04, pages 289–296. find
  12. Shalom Lappin and Herbert J. Leass. 1994. An algorithm for pronominal anaphora resolution. Computational Linguistics, 20(4):535–561. find our card
  13. Dekang Lin and Patrick Pantel. 2001. Discovery of inference rules for question answering. Natural Language Engineering, 7(4):343–360. find
  14. Dekang Lin. 1998. Dependency-based evaluation of MINI-PAR. In Proceedings of the Workshop on the Evaluation of Parsing Systems, First International Conference on Language Resources and Evaluation. find
  15. George A. Miller, Richard Beckwith, Christiane Fellbaum, Derek Gross, and Katherine J. Miller. 1990. Introduction to WordNet: an on-line lexical database. International Journal of Lexicography, 3(4):235–244. find
  16. Ruslan Mitkov. 1997. Factors in anaphora resolution: they are not the only things that matter. a case study based on two different approaches. In Proceedings of the ACL ’97 / EACL ’97 Workshop on Operational Factors in Practical, Robust Anaphora Resolution, pages 14–21. find our card
  17. MUC-7. 1997. Coreference task definition (v3.0, 13 Jul 97). In Proceedings of the Seventh Message Understanding Conference (MUC-7). find
  18. Vincent Ng and Claire 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. find our card
  19. Wee Meng Soon, Hwee Tou Ng, and Daniel Chung Yong Lim. 2001. A machine learning approach to coreference resolution of noun phrases. Computational Linguistics, 27(4):521–544. find our card
  20. Xiaofeng Yang, Jian Su, and Chew Lim Tan. 2005. Improving pronoun resolution using statistics-based semantic compatibility information. In Proceedings of the 43rd Annual Meeting of the Association for Computational Linguistics (ACL’05), pages 165–172, June. find our card

↑ back to top