Supersense Tagging of Unknown Nouns using Semantic Similarity

What it is about

Curran classifies unknown common nouns into the 26 WordNet lexicographer supersenses without any annotated examples. The method extracts about 50 distributionally similar nouns for the target from a shallow-parsed corpus of 2 billion words, and each of these synonyms votes with its WordNet 1.6 supersenses, weighted by similarity. The best configuration reaches 68% on the WordNet 1.6 test set and 63% on the harder WordNet 1.7.1 test set. The baseline is 21% and 28%, and the supervised perceptron of Ciaramita and Johnson reaches 53% on both sets, so the unsupervised method is significantly better. Proper names are not included, which makes locations hard, and multiword terms (23% of the 1.7.1 test set) use the synonyms of their last word. The fallback for nouns without any synonym is 37% accurate on the 1.7.1 set.

Ideas

• I1. onto/src/wsd.rs: a synonym vote for unknown nouns. Take the nouns that co-occur with the target in a large open corpus (parsed with en), let each WordNet neighbour vote with its lexicographer file, and keep the top 50 as the explanation. Measure accuracy on the unknown nouns in an open treebank, and check the first two runs are the same.
• I2. Multiword fallback in en/src/lexgen.rs: when an unknown noun is a multiword term, look up the head (last) noun's supersense and count how often that is right on a list of multiword nouns from onto. The paper reports 23% of its test set are multiword.
• I3. Proper-name routing in en/src/ptag.rs and prag: send capitalised names to a separate place or person decision, so that the vote is not used for them. Measure the location errors that the paper reports, on EWT names.

For Mova

The touch points are onto/src/wsd.rs (the vote and its explanation), en/src/lexgen.rs (multiword heads) and en/src/ptag.rs (names). The chapter is 11. Nouns, with 19. The lexicon: vocabulary, multiword expressions and register for multiword terms. The 2-billion-word corpus is not open and is not used in the build; the method can run on an open corpus instead, and the WordNet 1.7.1 test is for tests only. Measure first the head-noun fallback on multiword unknown nouns.

Cite

The paper itself, not this card. BibTeX from the library's record of the work. DOI 10.3115/1219840.1219844

@inproceedings{curran2005supersense,
  title = {{Supersense Tagging of Unknown Nouns using Semantic Similarity}},
  author = {Curran, James},
  booktitle = {Proceedings of the 43rd Annual Meeting of the Association for Computational Linguistics (ACL'05)},
  year = {2005},
  doi = {10.3115/1219840.1219844},
  url = {https://doi.org/10.3115/1219840.1219844}
}

References

27 references, 3 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. L. Douglas Baker and Andrew McCallum. 1998. Distributional clustering of words for text classification. In Proceedings of the 21st annual international ACM SIGIR conference on Research and Development in Information Retrieval, pages 96–103, Melbourne, Australia. find
2. Doug Beeferman. 1998. Lexical discovery with an enriched semantic network. In Proceedings of the Workshop on Usage of WordNet in Natural Language Processing Systems, pages 358–364, Montréal, Québec, Canada. find
3. Thorsten Brants. 2000. TnT - a statistical part-of-speech tagger. In Proceedings of the 6th Applied Natural Language Processing Conference, pages 224–231, Seattle, WA USA. find our card
4. Anita Burgun and Olivier Bodenreider. 2001. Comparing terms, concepts and semantic classes in WordNet and the Unified Medical Language System. In Proceedings of the Workshop on WordNet and Other Lexical Resources: Applications, Extensions and Customizations, pages 77–82, Pittsburgh, PA USA. find
5. Sharon A. Caraballo and Eugene Charniak. 1999. Determining the specificity of nouns from text. In Proceedings of the Joint ACL SIGDAT Conference on Empirical Methods in Natural Language Processing and Very Large Corpora, pages 63–70, College Park, MD USA. find
6. Massimiliano Ciaramita and Mark Johnson. 2003. Supersense tagging of unknown nouns in WordNet. In Proceedings of the 2003 Conference on Empirical Methods in Natural Language Processing, pages 168–175, Sapporo, Japan. find our card
7. Massimiliano Ciaramita, Thomas Hofmann, and Mark Johnson. 2003. Hierarchical semantic classification: Word sense disambiguation with world knowledge. In Proceedings of the 18th International Joint Conference on Artificial Intelligence, Acapulco, Mexico. find
8. Massimiliano Ciaramita. 2002. Boosting automatic lexical acquisition with morphological information. In Proceedings of the Workshop on Unsupervised Lexical Acquisition, pages 17–25, Philadelphia, PA, USA. find
9. Stephen Clark and David Weir. 2002. Class-based probability estimation using a semantic hierarchy. Computational Linguistics, 28(2):187–206, June. find
10. Koby Crammer and Yoram Singer. 2001. Ultraconservative online algorithms for multiclass problems. In Proceedings of the 14th annual Conference on Computational Learning Theory and 5th European Conference on Computational Learning Theory, pages 99–115, Amsterdam, The Netherlands. find
11. James R. Curran and Stephen Clark. 2003. Investigating GIS and smoothing for maximum entropy taggers. In Proceedings of the 10th Conference of the European Chapter of the Association for Computational Linguistics, pages 91–98, Budapest, Hungary. find
12. James R. Curran and Marc Moens. 2002a. Improvements in automatic thesaurus extraction. In Proceedings of the Workshop on Unsupervised Lexical Acquisition, pages 59– 66, Philadelphia, PA, USA. find
13. James R. Curran and Marc Moens. 2002b. Scaling context space. In Proceedings of the 40th annual meeting of the Association for Computational Linguistics, pages 231–238, Philadelphia, PA, USA. find
14. Christiane Fellbaum, editor. 1998. WordNet: An Electronic Lexical Database. MIT Press, Cambridge, MA USA. find
15. Gregory Grefenstette. 1994. Explorations in Automatic Thesaurus Discovery. Kluwer Academic Publishers, Boston, MA USA. find
16. Marti A. Hearst and Hinrich Schütze. 1993. Customizing a lexicon to better suit a computational task. In Proceedings of the Workshop on Acquisition of Lexical Knowledge from Text, pages 55–69, Columbus, OH USA. find
17. Rob Koeling. 2000. Chunking with maximum entropy models. In Proceedings of the 4th Conference on Computational Natural Language Learning and of the 2nd Learning Language in Logic Workshop, pages 139–141, Lisbon, Portugal. find
18. Mitchell P. Marcus, Beatrice Santorini, and Mary Ann Marcinkiewicz. 1994. Building a large annotated corpus of English: the Penn Treebank. Computational Linguistics, 19(2):313–330. find our card
19. Guido Minnen, John Carroll, and Darren Pearce. 2001. Applied morphological processing of English. Natural Language Engineering, 7(3):207–223. find
20. Tom Morton. 2002. Grok tokenizer. Grok OpenNLP toolkit. find
21. Marius Pasca and Sanda M. Harabagiu. 2001. The informative role of WordNet in open-domain question answering. In Proceedings of the Workshop on WordNet and Other Lexical Resources: Applications, Extensions and Customizations, pages 138–143, Pittsburgh, PA USA. find
22. Darren Pearce. 2001. Synonymy in collocation extraction. In Proceedings of the Workshop on WordNet and Other Lexical Resources: Applications, Extensions and Customizations, pages 41–46, Pittsburgh, PA USA. find
23. Philip Resnik. 1995. Using information content to evaluate semantic similarity. In Proceedings of the 14th International Joint Conference on Artificial Intelligence, pages 448–453, Montreal, Canada. find
24. Jeffrey C. Reynar and Adwait Ratnaparkhi. 1997. A maximum entropy approach to identifying sentence boundaries. In Proceedings of the Fifth Conference on Applied Natural Language Processing, pages 16–19, Washington, D.C. USA. find
25. Hinrich Schütze. 1992. Context space. In Intelligent Probabilistic Approaches to Natural Language, number FS-92-04 in Fall Symposium Series, pages 113–120, Stanford University, CA USA. find
26. Dominic Widdows. 2003. Unsupervised methods for developing taxonomies by combining syntactic and statistical information. In Proceedings of the Human Language Technology Conference of the North American Chapter of the Association for Computational Linguistics, pages 276–283, Edmonton, Alberta Canada… find
27. David Yarowsky. 1992. Word-sense disambiguation using statistical models of Roget’s categories trained on large corpora. In Proceedings of the 14th international conference on Computational Linguistics, pages 454–460, Nantes, France. find

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/papers/aclbulk-2005-curran-supersense-tagging-unknown-nouns-semantic/