Elamur Research⌕
☰
home / papers / 2016 / Unsupervised Ranking Model for Entity Coreference Resolution

Unsupervised Ranking Model for Entity Coreference Resolution

Summary

Entity coreference resolution groups the noun-phrase mentions in a text that point to the same real-world entity. The strongest systems are supervised ranking models, but they depend on costly hand-annotated corpora that resource-poor languages rarely have. Earlier unsupervised methods trailed both supervised and rule-based systems.

The authors present the first unsupervised ranking model for this task. It is generative and, in the spirit of IBM Model 1, chooses the most probable antecedent for each mention independently. Each mention also receives a deterministic resolution mode: string-matching, precise-construct, or attribute-matching. These modes are borrowed from sieves in the Stanford multi-sieve system, and each one uses its own feature set. Parameters are learned with EM on unannotated Gigaword news text.

On the CoNLL-2012 English test data the model reaches 58.44% CoNLL F1. That is 3.01 points above the Stanford deterministic system, and it also beats the supervised IMS system. The result shows that annotation-free coreference can substantially narrow the gap to supervised accuracy.

Key points

  • Training data: the APW and NYT sections of the Gigaword Corpus (1994–2010), about 1.6B words in 3.6M documents. Duplicates and documents shorter than 3 sentences were removed. Development and test sets are the CoNLL-2012 English data drawn from OntoNotes, with automatically detected mentions and predicted parses.
  • Mode-dependent distance: sentence distance is modelled only in attribute-matching mode, which covers cases such as pronouns. The other two modes treat antecedent position as uniform.
  • Learning: EM ran for 10 iterations from uniform initialization, and the best development scores came around iteration 5. Each iteration took about 12 hours on 10 CPUs.
  • Gains over unsupervised and deterministic baselines (CoNLL F1, dev/test):
    • Stanford: +2.93 / +3.01
    • Multigraph: +1.41 / +1.77
    • MIR: +2.62 / +3.02
    • The model leads all three on every metric reported: MUC, B³, CEAFm, CEAFe and BLANC.
  • Comparison with supervised systems: the model beats IMS by 1.02 points (58.44 vs 57.42 on test). It still trails the best supervised systems, for example Non-Linear at 63.39.

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

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

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/n16-1116 · arXiv 1603.04553

@inproceedings{ma2016unsupervised,
  title = {{Unsupervised Ranking Model for Entity Coreference Resolution}},
  author = {Ma, Xuezhe and Liu, Zhengzhong and Hovy, Eduard},
  booktitle = {Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies},
  year = {2016},
  doi = {10.18653/v1/n16-1116},
  eprint = {1603.04553},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.18653/v1/n16-1116}
}

References

39 references, 25 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Amit Bagga and Breck Baldwin. 1998. Algorithms for scoring coreference chains. In The first international conference on language resources and evaluation workshop on linguistics coreference, volume 1, pages 563–566. find our card
  2. Citeseer. Shane Bergsma and Dekang Lin. 2006. Bootstrapping path-based pronoun resolution. In Proceedings of ACL-2006, pages 33–40, Sydney, Australia, July. Association for Computational Linguistics. Anders Björkelund and Richárd Farkas. 2012. Data-driven multilingual coreference resolution using… find our card
  3. Peter F Brown, Vincent J Della Pietra, Stephen A Della Pietra, and Robert L Mercer. 1993. The mathematics of statistical machine translation: Parameter estimation. Computational linguistics, 19(2):263–311. find
  4. Nathanael Chambers and Dan Jurafsky. 2008. Unsupervised learning of narrative event chains. In Proceedings of ACL-2008: HLT, pages 789–797, Columbus, Ohio, June. Association for Computational Linguistics. find
  5. Nathanael Chambers and Dan Jurafsky. 2009. Unsupervised learning of narrative schemas and their participants. In Proceedings of ACL-2009, pages 602–610, Suntec, Singapore, August. Association for Computational Linguistics. find
  6. Eugene Charniak and Micha Elsner. 2009. EM works for pronoun anaphora resolution. In Proceedings of EACL 2009, pages 148–156, Athens, Greece, March. find our card
  7. Colin Cherry and Shane Bergsma. 2005. An Expectation Maximization approach to pronoun resolution. In Proceedings of CoNLL-2005, pages 88–95, Ann Arbor, Michigan, June. find our card
  8. Kevin Clark and Christopher D. Manning. 2015. Entity-centric coreference resolution with model stacking. In Proceedings of ACL-IJCNLP-2015, pages 1405–1415, Beijing, China, July. Association for Computational Linguistics. find our card
  9. Arthur P Dempster, Nan M Laird, and Donald B Rubin. 1977. Maximum likelihood from incomplete data via the em algorithm. Journal of the royal statistical society. Series B (methodological), pages 1–38. find
  10. Greg Durrett and Dan Klein. 2013. Easy victories and uphill battles in coreference resolution. In Proceedings of EMNLP-2013, pages 1971–1982, Seattle, Washington, USA, October. Association for Computational Linguistics. find our card
  11. Greg Durrett and Dan Klein. 2014. A joint model for entity analysis: Coreference, typing, and linking. In Proceedings of the Transactions of the Association for Computational Linguistics. Eraldo Fernandes, Cı́cero dos Santos, and Ruy Milidiú. 2012. Latent structure perceptron with feature induction… find our card
  12. Computational Linguistics. Aria Haghighi and Dan Klein. 2007. Unsupervised coreference resolution in a nonparametric bayesian model. In Proceedings of ACL-2007, pages 848–855, Prague, Czech Republic, June. Association for Computational Linguistics. find our card
  13. Aria Haghighi and Dan Klein. 2009. Simple coreference resolution with rich syntactic and semantic features. In Proceedings of EMNLP-2009, pages 1152–1161, Singapore, August. Association for Computational Linguistics. find our card
  14. Aria Haghighi and Dan Klein. 2010. Coreference resolution in a modular, entity-centered model. In Proceedings of NAACL-2010, pages 385–393, Los Angeles, California, June. Association for Computational Linguistics. find our card
  15. Eduard Hovy, Mitchell Marcus, Martha Palmer, Lance Ramshaw, and Ralph Weischedel. 2006. Ontonotes: The 90% solution. In Proceedings of NAACL-2006, pages 57–60, New York City, USA, June. Association for Computational Linguistics. find
  16. Heng Ji and Dekang Lin. 2009. Gender and animacy knowledge discovery from web-scale n-grams for unsupervised person mention detection. In Proceedings of PACLIC-2009, pages 220–229. find
  17. Heng Ji, HT Dang, J Nothman, and B Hachey. 2014. Overview of tac-kbp2014 entity discovery and linking tasks. In Proc. Text Analysis Conference (TAC2014). find
  18. Heeyoung Lee, Yves Peirsman, Angel Chang, Nathanael Chambers, Mihai Surdeanu, and Dan Jurafsky. 2011. Stanford’s multi-pass sieve coreference resolution system at the conll-2011 shared task. In Proceedings of CoNLL-2011: Shared Task, pages 28–34, Portland, Oregon, USA, June. Association for Computat… find our card
  19. Heeyoung Lee, Angel Chang, Yves Peirsman, Nathanael Chambers, Mihai Surdeanu, and Dan Jurafsky. 2013. Deterministic coreference resolution based on entity-centric, precision-ranked rules. Comput. Linguist., 39(4):885–916, December. find our card
  20. Xiaoqiang Luo. 2005. On coreference resolution performance metrics. In Proceedings of EMNLP-2005, pages 25–32, Vancouver, British Columbia, Canada, October. Association for Computational Linguistics. find our card
  21. Xuezhe Ma and Fei Xia. 2014. Unsupervised dependency parsing with transferring distribution via parallel guidance and entropy regularization. In Proceedings of ACL-2014, pages 1337–1348, Baltimore, Maryland, June. find
  22. Sebastian Martschat and Michael Strube. 2015. Latent structures for coreference resolution. Transactions of the Association for Computational Linguistics, 3:405– 418. find our card
  23. Sebastian Martschat. 2013. Multigraph clustering for unsupervised coreference resolution. In ACL-2013: Student Research Workshop, pages 81–88, Sofia, Bulgaria, August. Association for Computational Linguistics. find our card
  24. Joseph F McCarthy and Wendy G Lehnert. 1995. Using decision trees for conference resolution. In Proceedings of IJCAI-1995, pages 1050–1055. find
  25. Morgan Kaufmann Publishers Inc. Nafise Sadat Moosavi and Michael Strube. 2014. Unsupervised coreference resolution by utilizing the most informative relations. In Proceedings of COLING-2014, pages 644–655. find
  26. Vincent Ng. 2008. Unsupervised models for coreference resolution. In Proceedings of EMNLP-2008, pages 640–649, Honolulu, Hawaii, October. Association for Computational Linguistics. find our card
  27. Vincent Ng. 2010. Supervised noun phrase coreference research: The first fifteen years. In Proceedings of ACL-2010, pages 1396–1411, Uppsala, Sweden, July. Association for Computational Linguistics. find our card
  28. Robert Parker, David Graff, Junbo Kong, Ke Chen, and Kazuaki Maeda. 2011. English gigaword fifth edition. Linguistic Data Consortium, LDC2011T07. find
  29. Hoifung Poon and Pedro Domingos. 2008. Joint unsupervised coreference resolution with Markov Logic. In Proceedings of EMNLP-2008, pages 650–659, Honolulu, Hawaii, October. Association for Computational Linguistics. find our card
  30. Sameer Pradhan, Alessandro Moschitti, Nianwen Xue, Olga Uryupina, and Yuchen Zhang. 2012. Conll-2012 shared task: Modeling multilingual unrestricted coreference in ontonotes. In Proceedings of EMNLP-CoNLL-2012 - Shared Task, pages 1–40, Jeju Island, Korea, July. Association for Computational Linguis… find our card
  31. Lev Ratinov and Dan Roth. 2012. Learning-based multi-sieve co-reference resolution with knowledge. In Proceedings of EMNLP-CoNLL-2012, pages 1234–1244, Jeju Island, Korea, July. Association for Computational Linguistics. find our card
  32. Marta Recasens and Eduard Hovy. 2011. Blanc: Implementing the rand index for coreference evaluation. Natural Language Engineering, 17(04):485–510. find
  33. 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
  34. Marc Vilain, John Burger, John Aberdeen, Dennis Connolly, and Lynette Hirschman. 1995. A model-theoretic coreference scoring scheme. In Proceedings of the 6th conference on Message understanding, pages 45–52. Association for Computational Linguistics. find our card
  35. Ben Wellner, Andrew McCallum, Fuchun Peng, and Michael Hay. 2004. An integrated, conditional model of information extraction and coreference with application to citation matching. In Proceedings of the 20th conference on Uncertainty in artificial intelligence, pages 593–601. find
  36. AUAI Press. Sam Wiseman, Alexander M. Rush, Stuart Shieber, and Jason Weston. 2015. Learning anaphoricity and antecedent ranking features for coreference resolution. In Proceedings of ACL-IJCNLP-2015, pages 1416– 1426, Beijing, China, July. Association for Computational Linguistics. find our card
  37. Xiaofeng Yang, Jian Su, Jun Lang, Chew Lim Tan, Ting Liu, and Sheng Li. 2008a. An entity-mention model for coreference resolution with inductive logic programming. In Proceedings of ACL-2008, pages 843– 851. find our card
  38. Xiaofeng Yang, Jian Su, and Chew Lim Tan. 2008b. A twin-candidate model for learning-based anaphora resolution. Computational Linguistics, 34(3):327–356. find our card
  39. Appendix A. Derivation of Model Learning Formally, we iteratively estimate the model parameters θ, employing the following EM algorithm: E-step: Compute the posterior probabilities of C, P (C\|D; θ), based on the current θ. θ 0 M-step: Calculate the new the expected complete E P (C\|D;θ) [log P (D, C;… find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/ma-2016-unsupervised-ranking-model-entity-coreference/