Elamur Research⌕
☰
home / papers / 2013 / A Structured Distributional Semantic Model for Event Co-reference

A Structured Distributional Semantic Model for Event Co-reference

Summary

Standard distributional semantic models treat context as a bag of words, so they cannot tell "Booth killed Lincoln" apart from "Lincoln killed Booth." They also compose meaning poorly. The authors propose a Structured DSM (SDSM) in which each word is a matrix: one axis covers dependency relations, the other covers the context vocabulary.

The matrices come from a PropStore, a database of dependency triples drawn from a parsed Simple English Wikipedia. The PropStore also records lemmas, POS tags, WordNet supersenses and sentence IDs. To compose two words linked by a relation, SDSM keeps only the sentences where both words appear with that relation. The context gathered from those sentences becomes the phrase's representation.

The model is tested on judging whether two event mentions corefer. Each event is an event–agent–patient triple, compared at three levels of composition. SDSM outperforms window-based vectors, SENNA embeddings and simple additive or multiplicative composition. This suggests that syntax-aware structure helps distributional meaning in compositional settings.

Key points

  • The PropStore stores ⟨w1, r, w2⟩ triples with sentence indices. Parsing used Tratz and Hovy (2011), supersenses came from sst-light, and phrase queries were capped at three words because of sparsity.
  • The authors computed 159 similarity features per event pair. These vary in composition level (full, partial, none), granularity (lemma vs. SST), distance metric and normalization. A J48 decision tree performed best among the classifiers tried.
  • Two datasets were used:
    • IC Event Coreference Corpus: 100 violent-event news articles, with 2214 coreferent and 2214 non-coreferent pairs.
    • ECB: 482 Google News documents in 45 topics, giving 11039 positive and 33459 negative instances.
  • On IC, SDSM reached F-1 0.922 and accuracy 0.906, versus 0.865 and 0.835 for SENNA. On ECB, SDSM reached F-1 0.564 and accuracy 0.843, versus the best baseline's 0.528 and 0.834.
  • In the ablation study, agent–action partial composition and individual word features were the most predictive. Patient features and full-composition features helped least, likely because of sparsity.

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); 3 of them have a page here, 4 not yet.

Cite

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

@inproceedings{goyal2013structured,
  title = {{A Structured Distributional Semantic Model for Event Co-reference}},
  author = {Goyal, Kartik and Jauhar, Sujay Kumar and Li, Huiying and Sachan, Mrinmaya and Srivastava, Shashank and Hovy, Eduard},
  booktitle = {Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)},
  year = {2013},
  url = {https://aclanthology.org/P13-2083/}
}

References

32 references, 9 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Marco Baroni and Alessandro Lenci. 2010. Distributional memory: A general framework for corpus-based semantics. Comput. Linguist., 36(4):673–721, December. find our card
  2. Marco Baroni and Roberto Zamparelli. 2010. Nouns are vectors, adjectives are matrices: representing adjective-noun constructions in semantic space. In Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing, EMNLP ’10, pages 1183–1193, Stroudsburg, PA, USA. Association… find
  3. Cosmin Adrian Bejan and Sanda Harabagiu. 2010. Unsupervised event coreference resolution with rich linguistic features. In Proceedings of the 48th Annual Meeting of the Association for Computational Linguistics, ACL ’10, pages 1412–1422, Stroudsburg, PA, USA. Association for Computational Linguistic… find our card
  4. Massimiliano Ciaramita and Yasemin Altun. 2006. Broad-coverage sense disambiguation and information extraction with a supersense sequence tagger. In Proceedings of the 2006 Conference on Empirical Methods in Natural Language Processing, EMNLP ’06, pages 594–602, Stroudsburg, PA, USA. Association for… find
  5. Ronan Collobert, Jason Weston, Léon Bottou, Michael Karlen, Koray Kavukcuoglu, and Pavel Kuksa. 2011. Natural language processing (almost) from scratch. J. Mach. Learn. Res., 999888:2493–2537, November. find our card
  6. Georgiana Dinu and Mirella Lapata. 2010. Measuring distributional similarity in context. In Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing, EMNLP ’10, pages 1162–1172, Stroudsburg, PA, USA. Association for Computational Linguistics. find
  7. Katrin Erk and Sebastian Padó. 2008. A structured vector space model for word meaning in context. In Proceedings of the Conference on Empirical Methods in Natural Language Processing, EMNLP ’08, pages 897–906, Stroudsburg, PA, USA. Association for Computational Linguistics. find
  8. Katrin Erk. 2007. A simple, similarity-based model for selectional preferences. find
  9. Christiane Fellbaum. 1998. WordNet: An Electronic Lexical Database. Bradford Books. find
  10. John R. Firth. 1957. A Synopsis of Linguistic Theory, 1930-1955. Studies in Linguistic Analysis, pages 1– 32. find
  11. Kartik. Goyal, Sujay Kumar Jauhar, Mrinmaya Sachan, Shashank Srivastava, Huiying Li, and Eduard Hovy. 2013. A structured distributional semantic model : Integrating structure with semantics. In Proceedings of the 1st Continuous Vector Space Models and their Compositionality Workshop at the conferenc… find
  12. Edward Grefenstette, Mehrnoosh Sadrzadeh, Stephen Clark, Bob Coecke, and Stephen Pulman. 2011. Concrete sentence spaces for compositional distributional models of meaning. In Proceedings of the Ninth International Conference on Computational Semantics, IWCS ’11, pages 125–134, Stroudsburg, PA, USA.… find
  13. Emiliano Guevara. 2010. A regression model of adjective-noun compositionality in distributional semantics. In Proceedings of the 2010 Workshop on GEometrical Models of Natural Language Semantics, GEMS ’10, pages 33–37, Stroudsburg, PA, USA. Association for Computational Linguistics. find
  14. Aria Haghighi and Dan Klein. 2009. Simple coreference resolution with rich syntactic and semantic features. In Proceedings of the 2009 Conference on Empirical Methods in Natural Language Processing: Volume 3 - Volume 3, EMNLP ’09, pages 1152– 1161, Stroudsburg, PA, USA. Association for Computational… find our card
  15. E.H. Hovy, T. Mitamura, M.F. Verdejo, J. Araki, and A. Philpot. 2013. Events are not simple: Identity, non-identity, and quasi-identity. In Proceedings of the 1st Events Workshop at the conference of the HLT-NAACL 2013. find
  16. Heeyoung Lee, Marta Recasens, Angel Chang, Mihai Surdeanu, and Dan Jurafsky. 2012. Joint entity and event coreference resolution across documents. In Proceedings of the 2012 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning, EMNLP-CoNLL… find our card
  17. Christopher D. Manning, Prabhakar Raghavan, and Hinrich Schütze. 2008. Introduction to Information Retrieval. Cambridge University Press, New York, NY, USA. find
  18. Diana McCarthy and John Carroll. 2003. Disambiguating nouns, verbs, and adjectives using automatically acquired selectional preferences. Comput. Linguist., 29(4):639–654, December. find
  19. Diana McCarthy, Rob Koeling, Julie Weeds, and John Carroll. 2004. Finding predominant word senses in untagged text. In Proceedings of the 42nd Annual Meeting on Association for Computational Linguistics, ACL ’04, Stroudsburg, PA, USA. Association for Computational Linguistics. find our card
  20. Jeff Mitchell and Mirella Lapata. 2008. Vector-based models of semantic composition. In Proceedings of ACL-08: HLT, pages 236–244. find
  21. Patrick Pantel and Dekang Lin. 2000. Word-for-word glossing with contextually similar words. In Proceedings of the 1st North American chapter of the Association for Computational Linguistics conference, NAACL 2000, pages 78–85, Stroudsburg, PA, USA. Association for Computational Linguistics. find
  22. Karthik Raghunathan, Heeyoung Lee, Sudarshan Rangarajan, Nathanael Chambers, Mihai Surdeanu, Dan Jurafsky, and Christopher Manning. 2010. A multi-pass sieve for coreference resolution. In Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing, EMNLP ’10, pages 492–501… find our card
  23. Sebastian Rudolph and Eugenie Giesbrecht. 2010. Compositional matrix-space models of language. In Proceedings of the 48th Annual Meeting of the Association for Computational Linguistics, ACL ’10, pages 907–916, Stroudsburg, PA, USA. Association for Computational Linguistics. find
  24. Hinrich Schütze. 1998. Automatic word sense discrimination. Comput. Linguist., 24(1):97–123. find our card
  25. Diarmuid Ó Séaghdha and Anna Korhonen. 2011. Probabilistic models of similarity in syntactic context. In Proceedings of the Conference on Empirical Methods in Natural Language Processing, EMNLP ’11, pages 1047–1057, Stroudsburg, PA, USA. Association for Computational Linguistics. find
  26. Richard Socher, Brody Huval, Christopher D. Manning, and Andrew Y. Ng. 2012. Semantic compositionality through recursive matrix-vector spaces. In Proceedings of the 2012 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning, EMNLP-CoNLL ’12,… find
  27. Veselin Stoyanov, Nathan Gilbert, Claire Cardie, and Ellen Riloff. 2009. Conundrums in noun phrase coreference resolution: making sense of the state-of-the-art. In Proceedings of the Joint Conference of the 47th Annual Meeting of the ACL and the 4th International Joint Conference on Natural Language… find our card
  28. Stefanie Tellex, Boris Katz, Jimmy J. Lin, Aaron Fernandes, and Gregory Marton. 2003. Quantitative evaluation of passage retrieval algorithms for question answering. In SIGIR, pages 41–47. find
  29. Stefan Thater, Hagen Fürstenau, and Manfred Pinkal. 2010. Contextualizing semantic representations using syntactically enriched vector models. In Proceedings of the 48th Annual Meeting of the Association for Computational Linguistics, ACL ’10, pages 948–957, Stroudsburg, PA, USA. Association for Com… find
  30. Stephen Tratz and Eduard Hovy. 2011. A fast, accurate, non-projective, semantically-enriched parser. In Proceedings of the Conference on Empirical Methods in Natural Language Processing, EMNLP ’11, pages 1257–1268, Stroudsburg, PA, USA. Association for Computational Linguistics. find
  31. Hila Weisman, Jonathan Berant, Idan Szpektor, and Ido Dagan. 2012. Learning verb inference rules from linguistically-motivated evidence. In Proceedings of the 2012 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning, EMNLP-CoNLL ’12, pages… find
  32. S. K. M. Wong and Vijay V. Raghavan. 1984. Vector space model of information retrieval: a reevaluation. In Proceedings of the 7th annual international ACM SIGIR conference on Research and development in information retrieval, SIGIR ’84, pages 167–185, Swinton, UK. British Computer Society. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/goyal-2013-structured-distributional-semantic-model-event/