Elamur Research⌕
☰
home / papers / 2015 / Joint Event Trigger Identification and Event Coreference Resolution with Structured Perceptron

Joint Event Trigger Identification and Event Coreference Resolution with Structured Perceptron

Summary

Event extraction and event coreference resolution are usually handled in sequence: triggers are detected first and then linked. In that setup, a weak trigger detector can miss mentions such as abstract nominals or pronouns, and coreference evidence could otherwise have helped identify them. Araki and Mitamura address this with a single document-level model that predicts event triggers and coreference links together. They frame the output as an event graph, in which nodes are triggers (possibly multi-token) and edges are coreference links. The graph is learned with an averaged structured perceptron and beam search. Their incremental joint decoder combines segment-based trigger decoding with best-first clustering, which avoids the difficulty of synchronizing partial token-level labels with coreference decisions. Experiments on the biology-domain ProcessBank corpus compare the joint model against a two-stage pipeline that uses the same features and learner. The joint model improves coreference quality, mainly by reducing false positives, while trigger identification stays roughly unchanged. The work shows that the two tasks can benefit from being modeled jointly.

Key points

  • Data: ProcessBank, a set of 200 paragraphs taken from Campbell and Reece's (2005) Biology textbook, split into 120 train, 30 dev and 50 test paragraphs. It contains 1403 event triggers and 131 coreference links, and about 13.4% of the triggers span multiple tokens.
  • Event coreference on test: the joint model reaches 31.91 BLANC F1 versus 25.05 for the pipeline (+6.9) and 47.45 CoNLL F1 versus 45.66 (+1.8). Most of the gain comes from higher precision.
  • Trigger identification is nearly identical: the joint model scores 60.21 F1 and the baseline 60.68. The authors attribute this to the small corpus.
  • Training details:
    • The standard perceptron update was kept; early update performed worse, and max-violation update gave similar results.
    • Beam width was set to k=1 because the development set has only 28 coreferences.
    • Features come from Stanford CoreNLP, MATE, OpenNLP, Nomlex and Levin verb classes.
  • Future work includes scaling to ACE 2005 (likely requiring mini-batch distributed training) and adding event types, arguments and subevent relations.

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

14 papers of our library cite it (of the 140673 whose bibliographies we read); 5 of them have a page here, 9 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/d15-1247

@inproceedings{araki2015joint,
  title = {{Joint Event Trigger Identification and Event Coreference Resolution with Structured Perceptron}},
  author = {Araki, Jun and Mitamura, Teruko},
  booktitle = {Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing},
  year = {2015},
  doi = {10.18653/v1/d15-1247},
  url = {https://doi.org/10.18653/v1/d15-1247}
}

References

43 references, 18 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. James Allan. 2002. Topic Detection and Tracking: Event-based Information Organization. Kluwer Academic Publishers. find
  2. Amit Bagga and Breck Baldwin. 1998. Algorithms for scoring coreference chains. In Proceedings of LREC 1998 Workshop on Linguistics Coreference, pages 563–566. find our card
  3. Cosmin Adrian Bejan and Sanda M. Harabagiu. 2014. Unsupervised event coreference resolution. Computational Linguistics, 40(2):311–347. find our card
  4. Jonathan Berant, Vivek Srikumar, Pei-Chun Chen, Abby Vander Linden, Brittany Harding, Brad Huang, Peter Clark, and Christopher D. Manning. 2014. Modeling biological processes for reading comprehension. In Proceedings of EMNLP 2014, pages 1499–1510. find
  5. Daniel M. Bikel and Vittorio Castelli. 2008. Event matching using the transitive closure of dependency relations. In Proceedings of ACL 2008, pages 145– 148. find
  6. Anders Björkelund and Richárd Farkas. 2012. Data-driven multilingual coreference resolution using resolver stacking. In Proceedings of EMNLP/CoNLL 2012, pages 49–55. find our card
  7. Anders Björkelund and Jonas Kuhn. 2014. Learning structured perceptrons for coreference resolution with latent antecedents and non-local features. In Proceedings of ACL 2014, pages 47–57. find our card
  8. Anders Björkelund, Love Hafdell, and Pierre Nugues. 2009. Multilingual semantic role labeling. In Proceedings of CoNLL 2009, pages 43–48. find our card
  9. Bernd Bohnet and Joakim Nivre. 2012. A transition-based system for joint part-of-speech tagging and labeled non-projective dependency parsing. In Proceedings of EMNLP/CoNLL 2012, pages 1455– 1465. find
  10. Neil Campbell and Jane Reece. 2005. Biology. Benjamin Cummings. find
  11. Michael Collins and Brian Roark. 2004. Incremental parsing with the perceptron algorithm. In Proceedings of ACL 2004, pages 111–118. find our card
  12. Michael Collins. 2002. Discriminative training methods for Hidden Markov Models: Theory and experiments with perceptron algorithms. In Proceedings of EMNLP 2002, pages 1–8. find our card
  13. Marie-Catherine de Marneffe, Anna N. Rafferty, and Christopher D. Manning. 2008. Finding contradictions in text. In Proceedings of ACL-HLT 2008, pages 1039–1047. find
  14. Pascal Denis and Jason Baldridge. 2009. Global joint models for coreference resolution and named entity classification. Procesamiento del Lenguaje Natural, 42:87–96. find
  15. Yu Hong, Jianfeng Zhang, Bin Ma, Jianmin Yao, Guodong Zhou, and Qiaoming Zhu. 2011. Using cross-entity inference to improve event extraction. In Proceedings of ACL-HLT 2011, pages 1127–1136. find
  16. Liang Huang, Suphan Fayong, and Yang Guo. 2012. Structured perceptron with inexact search. In Proceedings of NAACL-HLT 2012, pages 142–151. find our card
  17. Heng Ji and Ralph Grishman. 2008. Refining event extraction through cross-document inference. In Proceedings of ACL-HLT 2008, pages 254–262. find
  18. Heng Ji and Ralph Grishman. 2011. Knowledge base population: Successful approaches and challenges. In Proceedings of ACL-HLT 2011, pages 1148–1158. find
  19. Richard Johansson and Pierre Nugues. 2008. Dependency-based semantic role labeling of Prop-Bank. In Proceedings of EMNLP 2008, pages 69– 78. find
  20. Heeyoung Lee, Marta Recasens, Angel Chang, Mihai Surdeanu, and Dan Jurafsky. 2012. Joint entity and event coreference resolution across documents. In Proceedings of EMNLP/CoNLL 2012, pages 489– 500. find our card
  21. Heeyoung Lee, Angel Chang, Yves Peirsman, Nathanael Chambers, Mihai Surdeanu, and Dan Jurafsky. 2013. Deterministic coreference resolution based on entity-centric, precision-ranked rules. Computational Linguistics, 39(4):885–916. find our card
  22. Beth Levin. 1993. English Verb Classes and Alternation: A Preliminary Investigation. The University of Chicago Press. find
  23. Qi Li and Heng Ji. 2014. Incremental joint extraction of entity mentions and relations. In Proceedings of ACL 2014, pages 402–412. find
  24. Wenjie Li, Mingli Wu, Qin Lu, Wei Xu, and Chunfa Yuan. 2006. Extractive summarization using inter- and intra- event relevance. In Proceedings of ACL/COLING 2006, pages 369–376. find
  25. Qi Li, Heng Ji, and Liang Huang. 2013. Joint event extraction via structured prediction with global features. In Proceedings of ACL 2013, pages 73–82. find
  26. Shasha Liao and Ralph Grishman. 2010. Using document level cross-event inference to improve event extraction. In Proceedings of ACL 2010, pages 789– 797. find
  27. Zhengzhong Liu, Jun Araki, Eduard Hovy, and Teruko Mitamura. 2014. Supervised within-document event coreference using information propagation. In Proceedings of LREC 2014. find our card
  28. Xiaoqiang Luo, Sameer Pradhan, Marta Recasens, and Eduard Hovy. 2014. An extension of BLANC to system mentions. In Proceedings of ACL 2014, pages 24–29. find
  29. Xiaoqiang Luo. 2005. On coreference resolution performance metrics. In Proceedings of HLT/EMNLP 2005, pages 25–32. find our card
  30. Catherine Macleod, Ralph Grishman, Adam Meyers, Leslie Barrett, and Ruth Reeves. 1998. Nomlex: A lexicon of nominalizations. In Proceedings of EU-RALEX 1998, pages 187–193. find
  31. Christopher Manning, Mihai Surdeanu, John Bauer, Jenny Finkel, Steven Bethard, and David McClosky. 2014. The Stanford CoreNLP natural language processing toolkit. In Proceedings ACL 2014: System Demonstrations, pages 55–60. find our card
  32. David McClosky, Mihai Surdeanu, and Christopher Manning. 2011. Event extraction as dependency parsing. In Proceedings of ACL-HLT 2011, pages 1626–1635. find
  33. Vincent Ng and Claire Cardie. 2002. Improving machine learning approaches to coreference resolution. In Proceedings of ACL 2002, pages 104–111. find our card
  34. Hoifung Poon and Lucy Vanderwende. 2010. Joint inference for knowledge extraction from biomedical literature. In Proceedings of NAACL-HLT 2010, pages 813–821. find
  35. Sameer Pradhan, Xiaoqiang Luo, Marta Recasens, Eduard Hovy, Vincent Ng, and Michael Strube. 2014. Scoring coreference partitions of predicted mentions: A reference implementation. In Proceedings of ACL 2014, pages 30–35. find our card
  36. Lev Ratinov and Dan Roth. 2009. Design challenges and misconceptions in named entity recognition. In Proceedings of CoNLL 2009, pages 147–155. find our card
  37. Marta Recasens and Eduard Hovy. 2011. BLANC: Implementing the Rand index for coreference evaluation. Natural Language Engineering, 17(4):485– 510. find
  38. Sebastian Riedel and Andrew McCallum. 2011. Fast and robust joint models for biomedical event extraction. In Proceedings of EMNLP 2011, pages 1–12. find
  39. Deepak Venugopal, Chen Chen, Vibhav Gogate, and Vincent Ng. 2014. Relieving the computational bottleneck: Joint inference for event extraction with high-dimensional features. In Proceedings of EMNLP 2014, pages 831–843. find
  40. Marc Vilain, John Burger, John Aberdeen, Dennis Connolly, and Lynette Hirschman. 1995. A model-theoretic coreference scoring scheme. In Proceedings of MUC-6, pages 45–52. find our card
  41. Yue Zhang and Stephen Clark. 2008a. Joint word segmentation and POS tagging using a single perceptron. In Proceedings of ACL-HLT 2008, pages 888– 896. find
  42. Yue Zhang and Stephen Clark. 2008b. A tale of two parsers: Investigating and combining graph-based and transition-based dependency parsing. In Proceedings of EMNLP 2008, pages 562–571. find our card
  43. Kai Zhao and Liang Huang. 2013. Minibatch and parallelization for online large margin structured learning. In Proceedings of NAACL-HLT 2013, pages 370–379. find

↑ back to top

This page in Markdown →