Specialized Models and Ranking for Coreference Resolution

Abstract : This paper investigates two strategies for improving coreference resolution: (1) training separate models that specialize in particular types of mentions (e.g., pronouns versus proper nouns) and (2) using a ranking loss function rather than a classification function. In addition to being conceptually simple, these modifications of the standard single-model, classification-based approach also deliver significant performance improvements. Specifically, we show that on the ACE corpus both strategies produce f-score gains of more than 3% across the three coreference evaluation metrics (MUC, B^3, and CEAF).
Type de document :
Communication dans un congrès
The Association for Computational Linguistics. Empirical Methods on Natural Language Processing, 2008, Honolulu, Hawaï, United States. pp.660-669, 2008, Proceedings of the 2008 Conference on Empirical Methods in Natural Language Processing. 〈http://www.aclweb.org/anthology/D/D08/D08-1069.pdf〉
Liste complète des métadonnées

https://hal.inria.fr/inria-00514368
Contributeur : Pascal Denis <>
Soumis le : jeudi 2 septembre 2010 - 09:09:56
Dernière modification le : mardi 11 octobre 2016 - 14:41:48

Identifiants

  • HAL Id : inria-00514368, version 1

Collections

Citation

Pascal Denis, Jason Baldridge. Specialized Models and Ranking for Coreference Resolution. The Association for Computational Linguistics. Empirical Methods on Natural Language Processing, 2008, Honolulu, Hawaï, United States. pp.660-669, 2008, Proceedings of the 2008 Conference on Empirical Methods in Natural Language Processing. 〈http://www.aclweb.org/anthology/D/D08/D08-1069.pdf〉. 〈inria-00514368〉

Partager

Métriques

Consultations de la notice

182