Elamur Research⌕
☰
home / papers / 2007 / Antecedent Selection Techniques for High-Recall Coreference Resolution

Antecedent Selection Techniques for High-Recall Coreference Resolution

Summary

The paper addresses coreferent bridging: definite noun phrases whose antecedent has a different lexical head (e.g., "the manor" referring back to "the house"). These cases limit recall because surface matching cannot find them. Versley studies antecedent selection on its own, assuming discourse-new definites are already identified. This yields an upper bound useful for feature design. The evaluation uses open-domain German newspaper text and compares several information sources:

  • hypernymy lookup in GermaNet
  • Web-mined shallow patterns
  • distributional similarity (Lin98, RFF)
  • association measures (Padó and Lapata's semantic space, Garera and Yarowsky's TheY)

The paper also separates lexical relation types and argues that relation-free association measures need semantic filtering. Otherwise they pick up non-coreferent associations such as house–door. A new distance-bounded similarity method nearly matches the precision of pattern search and beats Gasperin and Vieira's global similar-word lists. GermaNet and the Web patterns remain the most precise sources, but combining them with similarity- and association-based resolution substantially raises recall with only a modest precision cost.

Key points

  • Combining GermaNet and the Web-pattern approach with distance-bounded Lin similarity and a list of the 25 most similar words raised recall from 62% to 68% with little precision loss. Adding Lin's measure and TheY pushed recall to 69.7%, but at a larger precision cost.
  • The syntax-based distributional measures (Lin, 1998; Padó and Lapata, 2003) ranked candidates better than Garera and Yarowsky's TheY association measure, possibly because TheY suffers from sparse data or noise.
  • A hypothetical resolver combining these sources with perfect discourse-new detection would reach 88% precision and 83% recall on full NPs. This compares with 62% precision and 70% recall reported by Versley (2006) on the same data.
  • In preliminary full-system experiments, adding pattern-based information increased overall recall by 0.6%, which corresponds to 46% more resolved coreferent bridging cases. Distributional similarity gave no noticeable gain over GermaNet alone.
  • Sources covering different relations combine best, for example GermaNet for synonymy and hypernymy alongside patterns for instance relations. Precision matters especially when antecedent selection is integrated with discourse-new detection.

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: Discourse analysis · Coreference resolution

Cited by

8 papers of our library cite it (of the 140673 whose bibliographies we read); 4 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{versley2007antecedent,
  title = {{Antecedent Selection Techniques for High-Recall Coreference Resolution}},
  author = {Versley, Yannick},
  booktitle = {Proceedings of the 2007 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning (EMNLP-CoNLL)},
  year = {2007},
  url = {https://aclanthology.org/D07-1052/}
}

References

36 references, 7 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Asher, N. and Lascarides, A. (1998). Bridging. Journal of Semantics, 15(1):83–113. find
  2. Cardie, C. and Wagstaff, K. (1999). Noun phrase coreference as clustering. In Proceedings of the Joint Conference on Empirical Methods in Natural Language Processing and Very Large Corpora (EMNLP/VLC 1999), pages 82–89. find our card
  3. Clark, H. H. (1975). Bridging. In Schank, R. C. and Nash-Webber, B. L., editors, Proceedings of the 1975 workshop on Theoretical issues in natural language processing, pages 169–174, Cambridge, MA. Association for Computing Machinery. find
  4. Daumé III, H. and Marcu, D. (2005). A large-scale exploration of effective global features for a joint entity detection and tracking model. In HLT/EMNLP’05, pages 97–104. find
  5. Dunning, T. (1993). Accurate methods for the statistics of surprise and coincidence. Computational Linguistics, 19(1):61–74. find our card
  6. Foth, K. and Menzel, W. (2006). Hybrid parsing: Using probabilistic models as predictors for a symbolic parser. In ACL 2006. find
  7. Garera, N. and Yarowsky, D. (2006). Resolving and generating definite anaphora by modeling hypernymy using unlabeled corpora. In CoNLL 2006. find
  8. Gasperin, C., Salmon-Alt, S., and Vieira, R. (2004). How useful are similarity word lists for indirect anaphora resolution? In Proc. find
  9. Gasperin, C. and Vieira, R. (2004). Using word similarity lists for resolving indirect anaphora. In ACL’04 workshop on reference resolution and its applications. find our card
  10. Geffet, M. and Dagan, I. (2004). Feature vector quality and distributional similarity. In CoLing 2004. find
  11. Goecke, D., Stührenberg, M., and Wandmacher, T. (2006). Extraction and representation of semantic relations for resolving definite descriptions. In Workshop on Ontologies in Text Technology (OTT 2006). extended abstract. find
  12. Harabagiu, S., Bunescu, R., and Maiorano, S. (2001). Text and knowledge mining for coreference resolution. In Proceedings of the 2nd Meeting of the North American Chapter of the Association of Computational Linguistics (NAACL-2001). find
  13. Hinrichs, E., Kübler, S., and Naumann, K. (2005). A unified representation for morphological, syntactic, semantic and referential annotations. In ACL Workshop on Frontiers in Corpus Annotation II: Pie in the Sky, Ann Arbor. find
  14. Kunze, C. and Lemnitzer, L. (2002). Germanet – representation, visualization, application. In Proceedings of LREC 2002. find
  15. Lin, D. (1998). Automatic retrieval and clustering of similar words. In Proc. CoLing/ACL 1998. find
  16. Lund, K., Atchley, R. A., and Burgess, C. (1995). Semantic and associative priming in high-dimensional semantic space. In Proc. of the 17th Annual Conference of the Cognitive Science Society, pages 660–665. find
  17. Markert, K. and Nissim, M. (2005). Comparing knowledge sources for nominal anaphora resolution. Computational Linguistics, 31(3):367–402. find our card
  18. McCarthy, J. F. and Lehnert, W. G. (1995). Using decision trees for coreference resolution. In IJCAI 1995, pages 1050–1055. find
  19. Morton, T. S. (2000). Coreference for NLP applications. In ACL-2000. find
  20. Müller, F. H. and Ule, T. (2002). Annotating topological fields and chunks – and revising POS tags at the same time. In Proceedings of the Nineteenth International Conference on Computational Linguistics (COLING 2002). find
  21. Ng, V. (2007). Shallow semantics for coreference resolution. In IJCAI 2007, pages 1689–1694. find
  22. Padó, S. and Lapata, M. (2003). Constructing semantic space models from parsed corpora. In Proceedings of ACL 2003. find
  23. Padó, S. and Lapata, M. (2007). Dependency-based construction of semantic space models. Computational Linguistics, to appear. find our card
  24. Poesio, M., Alexandrov-Kabadjov, M., Vieira, R., Goulart, R., and Uryupina, O. (2005). Does discourse-new detection help definite description resolution? In Proceedings of the 6th International Workshop on Computational Semantics (IWCS-6). find
  25. Poesio, M., Schulte im Walde, S., and Brew, C. (1998). Lexical clustering and definite description interpretation. In AAAI Spring Symposium on Learning for Discourse. find
  26. Poesio, M., Vieira, R., and Teufel, S. (1997). Resolving bridging descriptions in unrestricted text. In ACL-97 Workshop on Operational Factors in Practical, Robust, Anaphora Resolution For Unrestricted Texts. find
  27. Ponzetto, S. P. and Strube, M. (2006). Exploiting semantic role labeling, wordnet and wikipedia for coreference resolution. In HLT-NAACL 2006. find our card
  28. Schmid, H., Fitschen, A., and Heid, U. (2004). SMOR: A german computational morphology covering derivation, composition and inflection. In Proceedings of LREC 2004. find our card
  29. Steinberger, J., Kabadjov, M., Poesio, M., and Sanchez-Graillet, O. (2005). Improving LSA-based summarization with anaphora resolution. In Proceedings of Human Language Technology Conference and Conference on Empirical Methods in Natural Language Processing, pages 1–8. find
  30. Uryupina, O. (2003). High-precision identification of discourse new and unique noun phrases. In Proceedings of the ACL Student Workshop. find
  31. Versley, Y. (2005). Parser evaluation across text types. In Proceedings of the Fourth Workshop on Treebanks and Linguistic Theories (TLT 2005). find
  32. Versley, Y. (2006). A constraint-based approach to noun phrase coreference resolution in German newspaper text. In Konferenz zur Verarbeitung Natürlicher Sprache (KONVENS 2006). find
  33. Versley, Y. (2007). Using the Web to resolve coreferent bridging in German newspaper text. In Proceedings of GLDV-Frühjahrstagung 2007, Tübingen. Narr. find
  34. Vieira, R. and Poesio, M. (2000). An empirically based system for processing definite descriptions. Computational Linguistics, 26(4):539–593. find
  35. Weeds, J. and Weir, D. (2005). Co-occurrence retrieval: A flexible framework for lexical distributional similarity. Computational Linguistics, 31(4):439–475. find
  36. Weeds, J., Weir, D., and McCarthy, D. (2004). Characterizing measures of lexical distributional similarity. In CoLing 2004. find

↑ back to top

This page in Markdown →