Elamur ResearchβŒ•
☰
home / papers / 2010 / Unsupervised Event Coreference Resolution with Rich Linguistic Features

Unsupervised Event Coreference Resolution with Rich Linguistic Features

Summary

Most earlier event coreference systems were supervised pairwise classifiers. They needed costly annotated data, they were tied to one domain or language, and they could not model how events are distributed across a document or topic. Bejan and Harabagiu treat coreference as fully unsupervised clustering and propose two generative nonparametric Bayesian models. The first extends the hierarchical Dirichlet process (HDP) so that each event mention can carry many features, not only its word. The second combines an infinite factorial HMM with an infinite HMM (iFHMM-iHMM). This lets it handle an unbounded feature space, choose salient feature values automatically and capture discourse-level ordering. In both models the number of events is left open. Each mention is described by lexical, class (TimeML event class), WordNet and semantic-role features (PropBank and FrameNet, aligned statistically). The authors evaluate within- and cross-document coreference on ACE and on ECB, a newly annotated corpus with documents grouped into topics. Both models clearly beat two baselines. This matters because it shows rich-feature event clustering can work without labeled training data.

Key points

  • Models: two HDP variants (HDP_flat and HDP_struct) whose feature types were tuned by hand, and the iFHMM-iHMM, which selects features automatically. The iFHMM-iHMM's scores mostly fall between those of the two HDP variants. It was tested with four filtering schemes (unfiltered, discrete, median, uniform), and none was clearly better than the others.
  • Features: head word and lemma, neighbouring words and neighbouring event mentions, POS, TimeML event class, WordNet synonym-closure and lexicographer-file clusters, PropBank arguments mapped to FrameNet frame elements, semantic frame, and feature combinations.
  • Feature selection: the iFHMM-iHMM kept at most 1.5% of distinct feature values. Its best results used about 0.05% on ACE (Ξ±β€² = 10) and 0.91% on ECB (Ξ±β€² = 150).
  • Sampling vs. using all features: in an experiment on a feature subset, BΒ³ F1 rose from 55.0 to about 80–82 on ACE within-document, from 73.6 to about 89 on ECB within-document, and from 64.5 to about 78–79 on ECB cross-document.
  • Comparison and errors: the closest prior system, the supervised approach of Chen and Ji (2009), reached 92.2% BΒ³ F on ACE, but it used gold event mentions and a hand-tuned threshold. The most common errors involve mentions whose arguments are left implicit, and participants or locations that are linked only by part-of relations.

Our summary of the paper, in our own words β€” not its text. The work belongs to its authors; read it at the links above.

Fields: Linguistics & languages

Topics: Coreference resolution Β· Semantic role labeling

Cited by

91 papers of our library cite it (of the 140673 whose bibliographies we read); 35 of them have a page here, 56 not yet.

Cite

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

@inproceedings{bejan2010unsupervised,
  title = {{Unsupervised Event Coreference Resolution with Rich Linguistic Features}},
  author = {Bejan, Cosmin and Harabagiu, Sanda},
  booktitle = {Proceedings of the 48th Annual Meeting of the Association for Computational Linguistics},
  year = {2010},
  url = {https://aclanthology.org/P10-1143/}
}

References

40 references, 13 of them with a card of ours. β†— leads to the work, Β«findΒ» searches for one we have no id for yet.

  1. ACE-Event. 2005. ACE (Automatic Content Extraction) English Annotation Guidelines for Events, version 5.4.3 2005.07.01. find
  2. David Ahn. 2006. The stages of event extraction. In Proceedings of the Workshop on Annotating and Reasoning about Time and Events, pages 1–8. find
  3. James Allan, Jaime Carbonell, George Doddington, Jonathan Yamron, and Yiming Yang. 1998. Topic Detection and Tracking Pilot Study: Final Report. In Proceedings of the Broadcast News Understanding and Transcription Workshop, pages 194–218. find
  4. Amit Bagga and Breck Baldwin. 1998. Algorithms for Scoring Coreference Chains. In Proceedings of the 1st International Conference on Language Resources and Evaluation (LREC-1998). find our card
  5. Amit Bagga and Breck Baldwin. 1999. Cross-Document Event Coreference: Annotations, Experiments, and Observations. In Proceedings of the ACL Workshop on Coreference and its Applications, pages 1–8. find our card
  6. Collin F. Baker, Charles J. Fillmore, and John B. Lowe. 1998. The Berkeley FrameNet project. In Proceedings of the 36th Annual Meeting of the Association for Computational Linguistics and 17th International Conference on Computational Linguistics (COLING-ACL). find our card
  7. Matthew J. Beal, Zoubin Ghahramani, and Carl Edward Rasmussen. 2002. The Infinite Hidden Markov Model. In Advances in Neural Information Processing Systems 14 (NIPS). find
  8. Cosmin Adrian Bejan and Sanda Harabagiu. 2008. A Linguistic Resource for Discovering Event Structures and Resolving Event Coreference. In Proceedings of the Sixth International Conference on Language Resources and Evaluation (LREC). find our card
  9. Cosmin Adrian Bejan and Chris Hathaway. 2007. UTD-SRL: A Pipeline Architecture for Extracting Frame Semantic Structures. In Proceedings of the Fourth International Workshop on Semantic Evaluations (SemEval), pages 460–463. find our card
  10. Cosmin Adrian Bejan, Matthew Titsworth, Andrew Hickl, and Sanda Harabagiu. 2009. Nonparametric Bayesian Models for Unsupervised Event Coreference Resolution. In Advances in Neural Information Processing Systems 23 (NIPS). find
  11. Cosmin Adrian Bejan. 2007. Deriving Chronological Information from Texts through a Graph-based Algorithm. In Proceedings of the 20th Florida Artificial Intelligence Research Society International Conference (FLAIRS), Applied Natural Language Processing track. find
  12. Zheng Chen and Heng Ji. 2009. Graph-based Event Coreference Resolution. In Proceedings of the 2009 Workshop on Graph-based Methods for Natural Language Processing (TextGraphs-4), pages 54– 57. find our card
  13. Donald Davidson, 1969. The Individuation of Events. In N. Rescher et al., eds., Essays in Honor of Carl G. Hempel, Dordrecht: Reidel. Reprinted in D. Davidson, ed., Essays on Actions and Events, 2001, Oxford: Clarendon Press. find
  14. Donald Davidson, 1985. Reply to Quine on Events, pages 172–176. In E. LePore and B. McLaughlin, eds., Actions and Events: Perspectives on the Philosophy of Donald Davidson, Oxford: Blackwell. find
  15. Marie-Catherine de Marneffe, Anna N. Rafferty, and Christopher D. Manning. 2008. Finding Contradictions in Text. In Proceedings of the 46th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies (ACL-HLT), pages 1039–1047. find
  16. Christiane Fellbaum. 1998. WordNet: An Electronic Lexical Database. MIT Press. find
  17. Thomas S. Ferguson. 1973. A Bayesian Analysis of Some Nonparametric Problems. The Annals of Statistics, 1(2):209–230. find
  18. Charles J. Fillmore. 1982. Frame Semantics. In Linguistics in the Morning Calm. find
  19. Stuart Geman and Donald Geman. 1984. Stochastic relaxation, Gibbs distributions and the Bayesian restoration of images. IEEE Transactions on Pattern Analysis and Machine Intelligence, 6:721–741. find
  20. Zoubin Ghahramani, T. L. Griffiths, and Peter Sollich, 2007. Bayesian Statistics 8, chapter Bayesian non-parametric latent feature models, pages 201–225. Oxford University Press. find
  21. Tom Griffiths and Zoubin Ghahramani. 2006. Infinite Latent Feature Models and the Indian Buffet Process. In Advances in Neural Information Processing Systems 18 (NIPS), pages 475–482. find
  22. Aria Haghighi and Dan Klein. 2007. Unsupervised Coreference Resolution in a Nonparametric Bayesian Model. In Proceedings of the 45th Annual Meeting of the Association of Computational Linguistics (ACL), pages 848–855. find our card
  23. Aria Haghighi, Andrew Ng, and Christopher Manning. 2005. Robust Textual Inference via Graph Matching. In Proceedings of Human Language Technology Conference and Conference on Empirical Methods in Natural Language Processing (HLT-EMNLP), pages 387–394. find
  24. Laura Hasler and Constantin Orasan. 2009. Do coreferential arguments make event mentions coreferential? In Proceedings of the 7th Discourse Anaphora and Anaphor Resolution Colloquium (DAARC 2009). find
  25. Kevin Humphreys, Robert Gaizauskas, and Saliha Azzam. 1997. Event coreference for information extraction. In Proceedings of the Workshop on Operational Factors in Practical, Robust Anaphora Resolution for Unrestricted Texts, 35th Meeting of ACL, pages 75–81. find our card
  26. John B. Lowe, Collin F. Baker, and Charles J. Fillmore. 1997. A frame-semantic approach to semantic annotation. In Proceedings of the SIGLEX Workshop on Tagging Text with Lexical Semantics: Why, What, and How?, pages 18–24. find our card
  27. Xiaoqiang Luo. 2005. On coreference resolution performance metrics. In Proceedings of the Human Language Technology Conference and Conference on Empirical Methods in Natural Language Processing (EMNLP-2005), pages 25–32. find our card
  28. Jeff Malpas. 2009. Donald Davidson. In The Stanford Encyclopedia of Philosophy (Fall 2009 Edition), Edward N. Zalta (ed.), http://plato.stan ford.edu/archives/fall2009/entries/davidson/. β†—
  29. Srini Narayanan and Sanda Harabagiu. 2004. Question Answering Based on Semantic Structures. In Proceedings of the 20th International Conference on Computational Linguistics (COLING), pages 693– 701. find
  30. Radford M. Neal. 2003. Slice Sampling. The Annals of Statistics, 31:705–741. find
  31. Vincent Ng. 2008. Unsupervised Models for Coreference Resolution. In Proceedings of the 2008 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 640–649. find our card
  32. Martha Palmer, Daniel Gildea, and Paul Kingsbury. 2005. The Proposition Bank: An Annotated Corpus of Semantic Roles. Computational Linguistics, 31(1):71–105. find our card
  33. Hoifung Poon and Pedro Domingos. 2008. Joint Unsupervised Coreference Resolution with Markov Logic. In Proceedings of the 2008 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 650–659. find our card
  34. James Pustejovsky, Jose Castano, Bob Ingria, Roser Sauri, Rob Gaizauskas, Andrea Setzer, and Graham Katz. 2003a. TimeML: Robust Specification of Event and Temporal Expressions in Text. In Proceedings of the Fifth International Workshop on Computational Semantics (IWCS). find
  35. James Pustejovsky, Patrick Hanks, Roser Sauri, Andrew See, Robert Gaizauskas, Andrea Setzer, Dragomir Radev, Beth Sundheim, David Day, Lisa Ferro, and Marcia Lazo. 2003b. The TimeBank Corpus. In Corpus Linguistics, pages 647–656. find
  36. W. V. O. Quine, 1985. Events and Reification, pages 162–171. In E. LePore and B. P. McLaughlin, eds., Actions and Events: Perspectives on the philosophy of Donald Davidson, Oxford: Blackwell. Reprinted in R. Casati and A. C. Varzi, eds., Events, 1996, pages 107–116, Aldershot: Dartmouth. find
  37. Lawrence R. Rabiner. 1989. A Tutorial on Hidden Markov Models and Selected Applications in Speech Recognition. In Proceedings of the IEEE, pages 257–286. find
  38. Yee Whye Teh, Michael Jordan, Matthew Beal, and David Blei. 2006. Hierarchical Dirichlet Processes. Journal of the American Statistical Association, 101(476):1566–1581. find
  39. Jurgen Van Gael, Y. Saatci, Yee Whye Teh, and Zoubin Ghahramani. 2008a. Beam Sampling for the Infinite Hidden Markov Model. In Proceedings of the 25th Annual International Conference on Machine Learning (ICML), pages 1088–1095. find
  40. Jurgen Van Gael, Yee Whye Teh, and Zoubin Ghahramani. 2008b. The Infinite Factorial Hidden Markov Model. In Advances in Neural Information Processing Systems 21 (NIPS). find

↑ back to top