Elamur Research⌕
☰
home / papers / 2008 / Context-Sensitive Convolution Tree Kernel for Pronoun Resolution

Context-Sensitive Convolution Tree Kernel for Pronoun Resolution

Summary

This work tackles pronoun resolution, the task of linking a third-person pronoun to its antecedent, using tree kernels over syntactic parses. The authors identify two weaknesses in earlier kernel-based work built on Collins and Duffy's convolution tree kernel (Yang et al. 2006). First, enumerated sub-trees ignore everything outside themselves. Second, there is no principled way to choose which part of the parse tree to compare. Their solution has two parts. A dynamic-expansion scheme builds the tree span by starting from the shortest anaphor–candidate path, then adding compatible competing candidates and predicate paths, and finally pruning redundant unary nodes. A context-sensitive convolution tree kernel then also counts sub-trees together with their ancestor node paths. Texts are parsed with the Charniak parser, and classification uses SVMLight. On the ACE 2003 corpus, both components beat earlier tree span schemes and the standard kernel. The dynamic scheme is notably more robust when the antecedent lies in an earlier sentence. The study shows that targeted, context-aware syntactic structure helps more than simply including more of the tree.

Key points

  • Evaluation covers third-person pronouns in ACE 2003. Candidate NPs come from the current and two preceding sentences and are filtered by number, person, and gender agreement, leaving about 7 candidates per anaphor. Results are reported as F-measure.
  • Including ancestor paths of length m = 3 (parent and grandparent context) gives about 2.2 F more than m = 1. Longer paths bring no further gain, likely because of parsing errors.
  • The dynamic-expansion span outperforms Min-, Simple-, and Full-Expansion by roughly 2.4, 1.2, and 2.1 F, respectively.
  • Ablation from the full system (83.0 F): removing competitor information costs 1.9, removing predicate information costs 0.8, and removing pruning costs 0.2.
  • For antecedents two sentences away, the dynamic scheme scores 54.2 F, compared with 43.8 for Simple-Expansion and 35.4 for Full-Expansion.

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

7 papers of our library cite it (of the 140673 whose bibliographies we read); 6 of them have a page here, 1 not yet.

Cite

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

@inproceedings{zhou2008context,
  title = {{Context-Sensitive Convolution Tree Kernel for Pronoun Resolution}},
  author = {Zhou, GuoDong and Kong, Fang and Zhu, QiaoMing},
  booktitle = {Proceedings of the Third International Joint Conference on Natural Language Processing: Volume-I},
  year = {2008},
  url = {https://aclanthology.org/I08-1004/}
}

References

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

  1. Aone C and Bennett W.W. (1995). Evaluating automated and manual acquisition of anaphora resolution strategies. ACL’1995:122-129. find our card
  2. Bergsma S. and Lin D.K.(2006). Bootstrapping path-based pronoun resolution. COLING-ACL’2006: 33-40. find our card
  3. Charniak E. (2001). Immediate-head Parsing for Language Models. ACL’2001: 129-137. find our card
  4. Toulouse, France Collins M. (1999) Head-driven statistical models for natural language parsing. Ph.D. Thesis. University of Pennsylvania. find
  5. Collins M. and Duffy N. (2001). Convolution Kernels for Natural Language. NIPS’2001: 625-632. find
  6. Cambridge, MA Culotta A. and Sorensen J. (2004). Dependency tree kernels for relation extraction. ACL’2004. 423-429. 21-26 July 2004. Barcelona, Spain. find
  7. Haussler D. (1999). Convolution Kernels on Discrete Structures. Technical Report UCS-CRL-99-10, University of California, Santa Cruz. find
  8. Hobbs J. (1978). Resolving pronoun references. Lingua. 44:339-352. find our card
  9. Joachims T. (1998). Text Categorization with Support Vector Machine: learning with many relevant features. ECML-1998: 137-142. find
  10. Chemnitz, Germany Lappin S. and Leass H. (1994). An algorithm for pronominal anaphora resolution. Computational Linguistics. 20(4):526-561. find our card
  11. Mitkov R. (1998). Robust pronoun resolution with limited knowledge. COLING-ACL’1998:869-875. Montreal, Canada. find our card
  12. Moschitti A. (2004). A study on convolution kernels for shallow semantic parsing. ACL’2004:335-342. find
  13. Pradhan S., Hacioglu K., Krugler V., Ward W., Martin J.H. and Jurafsky D. (2005). Support Vector Learning for Semantic Argument Classification. Machine Learning. 60(1):11-39. find
  14. Soon W. Ng H.T.and Lim D. (2001). A machine learning approach to creference resolution of noun phrases. Computational Linguistics. 27(4): 521-544. find
  15. Yang X.F., Zhou G.D., Su J. and Tan C.L., Coreference Resolution Using Competition Learning Approach, ACL’2003):176-183. Sapporo, Japan, 7-12 July 2003. find our card
  16. Yang X.F., Su J. and Tan C.L. (2006). Kernel-based pronoun resolution with structured syntactic knowledge. COLING-ACL’2006: 41-48. find our card
  17. Zelenko D., Aone C. and Richardella. (2003). Kernel methods for relation extraction. Journal of Machine Learning Research. 3(Feb):1083-1106. find
  18. Zhang M., Zhang J., Su J. and Zhou G.D. (2006). A Composite Kernel to Extract Relations between Entities with both Flat and Structured Features. COLING-ACL-2006: 825-832. find
  19. Sydney, Australia Zhou G.D., Su J. Zhang J. and Zhang M. (2005). Exploring various knowledge in relation extraction. ACL’2005. 427-434. 25-30 June, Ann Arbor, Mich-gan, USA. find
  20. Zhou G.D., Zhang M., Ji D.H. and Zhu Q.M. (2007). Tree Kernel-based Relation Extraction with Context-Sensitive Structured Parse Tree Information. EMNLP-CoNLL’2007 find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/zhou-2008-context-sensitive-convolution-tree-kernel/