Elamur Research⌕
☰
home / papers / 2008 / Statistical Anaphora Resolution in Biomedical Texts

Statistical Anaphora Resolution in Biomedical Texts

Summary

The paper tackles anaphora resolution for non-pronominal noun phrases that mention biomedical entities. Pronouns are rare in this genre, while gene and protein names are frequent. Inspired by Ge et al.'s probabilistic pronoun resolver, the authors build a Bayesian model. It is derived with Bayes' theorem, the chain rule and selective independence assumptions, which makes it a richer variant of naive Bayes that suits sparse training data. For each anaphor–candidate pair it scores four outcomes: coreferent, associative "biotype" (for example, gene vs. its protein), associative "set-member", or no relation. If no candidate is linked, the NP is treated as discourse-new. Training uses a new corpus of full-text Drosophila papers, chosen over abstracts because anaphora unfolds across whole documents. The model reaches state-of-the-art accuracy on coreference and handles associative cases reasonably. It outperforms a baseline and a C4.5 decision tree that use the same features. Each decision also comes with a probability, which can support confidence-based uses such as active learning and paper-curation tools.

Key points

  • Training data: 5 manually annotated full-text Drosophila articles (about 33,300 words), released through the FlySlip project. The authors describe it as the first full-text biomedical corpus with anaphora annotation.
  • The model uses 11 features: NP form, head/modifier string matching, number agreement, syntactic relation, biotype matching, a domain-specific gene/product feature (gp), and sentence and markable distance.
  • Coreference results: roughly 56–69% precision and 54–67% recall. Proper names score best, at about 77.5 P / 71.9 R.
  • Comparison systems: a baseline (coreferent F 51.8) and a Weka C4.5 decision tree (coreferent F 53.5). Neither gives a confidence score.
  • Pronouns make up only about 3% of NPs in the corpus. This motivated the focus on non-pronominal anaphora.

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

17 papers of our library cite it (of the 140673 whose bibliographies we read); 8 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.

@inproceedings{gasperin2008statistical,
  title = {{Statistical Anaphora Resolution in Biomedical Texts}},
  author = {Gasperin, Caroline and Briscoe, Ted},
  booktitle = {Proceedings of the 22nd International Conference on Computational Linguistics (Coling 2008)},
  year = {2008},
  doi = {10.3115/1599081.1599114},
  url = {https://doi.org/10.3115/1599081.1599114}
}

References

18 references, 4 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Briscoe, Edward J., John Carroll, and Rebecca Watson. 2006. The second release of the RASP system. In Proceedings of ACL-COLING 06, Sydney, Australia. find
  2. Bunescu, Razvan. 2003. Associative anaphora resolution: A web-based approach. In Proceedings of EACL 2003 - Workshop on The Computational Treatment of Anaphora, Budapest. find
  3. Castaño, José, Jason Zhang, and James Pustejovsky. 2002. Anaphora resolution in biomedical literature. In Proceedings of International Symposium on Reference Resolution for NLP 2002, Alicante, Spain. find
  4. Cohen, K. Bretonnel, Lynne Fox, Philip Ogren, and Lawrence Hunter. 2005. Corpus design for biomedical natural language processsing. In Proceedings of the ACL-ISMB Workshop on Linking Biological Literature, Ontologies and Databases, Detroit. find
  5. Eilbeck, Karen and Suzanna E. Lewis. 2004. Sequence ontology annotation guide. Comparative and Functional Genomics, 5:642–647. find
  6. Gasperin, Caroline, Nikiforos Karamanis, and Ruth Seal. 2007. Annotation of anaphoric relations in Baseline R F 57.6 51.8 find
  7. C4.5 R 58.1 28.5 31.3 32.5 C4.5 relaxed P R F 52.7 61.6 56.8 22.9 29.9 26.0 30.4 33.3 31.8 48.5 32.5 38.9 P 49.6 21.7 28.5 48.5 F 53.5 24.6 29.8 38.9 10.7 15.6 30.2 33.4 biomedical full-text articles using a domain-relevant scheme. In Proceedings of DAARC 2007, Lagos, Portugal. find
  8. Gasperin, Caroline. 2006. Semi-supervised anaphora resolution in biomedical texts. In Proceedings of BioNLP’06, New York. find
  9. Ge, Niyu, John Hale, and Eugene Charniak. 1998. A statistical approach to anaphora resolution. In Proceedings of the Sixth Workshop on Very Large Corpora - COLING-ACL’98, Montreal, Canada. find our card
  10. Karamanis, Nikiforos, Ruth Seal, Ian Lewin, Peter McQuilton, Andreas Vlachos, Caroline Gasperin, Rachel Drysdale, and Ted Briscoe. 2008. Natural language processing in aid of flybase curators. BMC Bioinformatics, 9(193). find
  11. Ng, Vincent and Claire Cardie. 2002. Improving machine learning approaches to coreference resolution. In Proceedings of ACL 2002, Philadelphia. find our card
  12. Poesio, Massimo, Renata Vieira, and Simone Teufel. 1997. Resolving bridging descriptions in unrestricted texts. In Proceedings of the Workshop on Operational Factors In Practical, Robust Anaphora Resolution for Unrestricted Texts, Madrid. find
  13. Soon, Wee Meng, 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
  14. Strube, Michael, Stefan Rapp, and Christoph Müller. 2002. The influence of minimum edit distance on reference resolution. In Proceedings of the EMNLP 2002, Philadelphia. find
  15. Vieira, Renata and Massimo Poesio. 2000. An empirically-based system for processing definite descriptions. Computational Linguistics, 26(4):525– 579. find
  16. Vlachos, Andreas and Caroline Gasperin. 2006. Bootstrapping and evaluating named entity recognition in the biomedical domain. In Proceedings of BioNLP at HLT-NAACL 2006, pages 138–145, New York. find
  17. Witten, Ian H. and Eibe Frank. 2005. Data Mining: Practical machine learning tools and techniques, 2nd Edition. Morgan Kaufmann, San Francisco. find
  18. Yang, X., J. Su, G. Zhou, and C. L. Tan. 2004. An NP-cluster based approach to coreference resolution. In Proceedings of COLING 2004, Geneva, Switzerland. find our card

↑ back to top

This page in Markdown →