Elamur Research⌕
☰
home / papers / 2010 / Metaphor Identification Using Verb and Noun Clustering

Metaphor Identification Using Verb and Noun Clustering

Summary

The paper addresses automatic detection of metaphorical language in unrestricted text, specifically verbs used figuratively with a subject or direct object. Earlier systems relied on hand-built knowledge such as WordNet, or were limited to one predefined domain. The authors instead start from 62 annotated seed metaphors drawn from the BNC and expand them using unsupervised clustering.

The intuition is that abstract nouns sharing a source domain behave alike in corpora even when their meanings differ. For example, marriage and political regime are both talked about as machines. Noun clustering groups such target concepts, and verb clustering builds a lexicon of source-domain verbs. The system then searches the RASP-parsed BNC for verb–object and verb–subject pairs that combine the two sets.

Human judges rated the output, and the system reached a precision of 0.79. A comparison with a WordNet-based baseline showed that the method finds metaphors well beyond synonyms of the seeds. The authors present it as the first unsupervised approach to metaphor identification, one that needs no hand-crafted resources beyond the seed set.

Key points

  • Seed set: 62 single-word verb metaphors from Shutova (2010), covering verb–object pairs (e.g. cast doubt, grasp theory) and subject–verb pairs (e.g. tension mounted).
  • Corpus: the whole BNC, using the RASP grammatical-relation output produced by Andersen et al. (2008).
  • Verb clustering:
    • The approach follows Sun and Korhonen (2009), which scored 80.4 F-measure on 204 verbs from 17 Levin classes.
    • Here it is scaled to 1610 VerbNet verbs, each seen more than 150 times in the VALEX source data.
    • Features are subcategorization frames (SCFs) parameterized by selectional preferences (SPs), extracted with the Preiss et al. (2007) system, which recognizes 168 SCF types.
  • Noun clustering:
    • Covers the 2000 most frequent BNC nouns.
    • Features are grammatical relations (subject, direct object, indirect object), normalized by the sum of feature values.
    • Both verbs and nouns are clustered with spectral clustering (MNCut) over a similarity matrix based on Jensen-Shannon divergence.
  • Motivation and result: a cited corpus study found metaphors in 241 of 761 BNC sentences, and in 164 of those phrases the metaphor came from the verb. The system's precision was 0.79 by human judgement.

Interesting ideas

  • I1. Treat distributional "errors" as signal: Distributional clustering tends to group abstract nouns by shared metaphorical source rather than by meaning. The authors exploit this side effect instead of fighting it, and the same reframing could help with other figurative or associative phenomena.
  • I2. Bootstrap from a few seeds by generalizing both slots: Clustering both the argument (target nouns) and the predicate (source verbs) turns a handful of examples into many new candidate pairs. This pattern suits low-resource languages where large annotated datasets are unavailable.
  • I3. Restrict search to parsed grammatical relations: Searching only within verb–subject and verb–object relations, rather than raw word co-occurrence, keeps candidates precise. The authors explicitly criticize bigram-based approaches for losing this information.
  • I4. Use a synonymy baseline to test generalization: Comparing against a WordNet baseline checks whether the method truly generalizes or merely rediscovers near-synonyms of the seeds. This is a useful sanity check for any seed-expansion method.
  • I5. Watch for class imbalance in reported results: The authors note that a related work's 95% accuracy barely beats a 92.9% majority-class baseline. Reporting such baselines is a simple safeguard against overstated results in imbalanced classification tasks.

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: Natural language processing

Cited by

95 papers of our library cite it (of the 140673 whose bibliographies we read); 6 of them have a page here, 89 not yet.

Cite

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

@inproceedings{shutova2010metaphor,
  title = {{Metaphor Identification Using Verb and Noun Clustering}},
  author = {Shutova, Ekaterina and Sun, Lin and Korhonen, Anna},
  booktitle = {Proceedings of the 23rd International Conference on Computational Linguistics (Coling 2010)},
  year = {2010},
  url = {https://aclanthology.org/C10-1113/}
}

References

36 references, 6 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Andersen, O. E., J. Nioche, E. Briscoe, and J. Carroll. 2008. The BNC parsed with RASP4UIMA. In Proceedings of LREC 2008, Marrakech, Morocco. find
  2. Bergsma, S., D. Lin, and R. Goebel. 2008. Discriminative learning of selectional preference from unlabeled text. In Proceedings of the EMNLP. find
  3. Birke, J. and A. Sarkar. 2006. A clustering approach for the nearly unsupervised recognition of nonliteral language. In In Proceedings of EACL-06, pages 329–336. find
  4. Brew, C. and S. Schulte im Walde. 2002. Spectral clustering for German verbs. In Proceedings of EMNLP. find
  5. Briscoe, E., J. Carroll, and R. Watson. 2006. The second release of the rasp system. In Proceedings of the COLING/ACL on Interactive presentation sessions, pages 77–80. find our card
  6. Burnard, L. 2007. Reference Guide for the British National Corpus (XML Edition). find
  7. Chen, J., D. Ji, C. Lim Tan, and Z. Niu. 2006. Unsupervised relation disambiguation using spectral clustering. In Proceedings of COLING/ACL. find
  8. Fass, D. 1991. met*: A method for discriminating metonymy and metaphor by computer. Computational Linguistics, 17(1):49–90. find
  9. Fellbaum, C., editor. 1998. WordNet: An Electronic Lexical Database (ISBN: 0-262-06197-X). MIT Press, first edition. find
  10. Fillmore, C. J., C. R. Johnson, and M. R. L. Petruck. 2003. Background to FrameNet. International Journal of Lexicography, 16(3):235–250. find
  11. Gedigan, M., J. Bryant, S. Narayanan, and B. Ciric. 2006. Catching metaphors. In In Proceedings of the 3rd Workshop on Scalable Natural Language Understanding, pages 41–48, New York. find
  12. Joanis, E., S. Stevenson, and D. James. 2008. A general feature space for automatic verb classification. Natural Language Engineering, 14(3):337–367. find
  13. Karov, Y. and S. Edelman. 1998. Similarity-based word sense disambiguation. Computational Linguistics, 24(1):41–59. find
  14. Kingsbury, P. and M. Palmer. 2002. From TreeBank to PropBank. In Proceedings of LREC-2002, Gran Canaria, Canary Islands, Spain. find our card
  15. Kipper, K., A. Korhonen, N. Ryant, and M. Palmer. 2006. Extensive classifications of English verbs. In Proceedings of the 12th EURALEX International Congress. find
  16. Korhonen, A., Y. Krymolowski, and Z. Marx. 2003. Clustering polysemic subcategorization frame distributions semantically. In Proceedings of ACL 2003, Sapporo,Japan. find
  17. Korhonen, A., Y. Krymolowski, and T. Briscoe. 2006. A large subcategorization lexicon for natural language processing applications. In Proceedings of LREC 2006. find
  18. Krishnakumaran, S. and X. Zhu. 2007. Hunting elusive metaphors using lexical resources. In Proceedings of the Workshop on Computational Approaches to Figurative Language, pages 13–20, Rochester, NY. find
  19. Lakoff, G. and M. Johnson. 1980. Metaphors We Live By. University of Chicago Press, Chicago. find
  20. Lakoff, G., J. Espenson, and A. Schwartz. 1991. The master metaphor list. Technical report, University of California at Berkeley. find
  21. Levin, B. 1993. English Verb Classes and Alternations. University of Chicago Press, Chicago. find
  22. Lin, D. 1998. Automatic retrieval and clustering of similar words. In Proceedings of the 17th international conference on Computational linguistics, pages 768–774. find our card
  23. Martin, J. H. 1988. Representing regularities in the metaphoric lexicon. In Proceedings of the 12th conference on Computational linguistics, pages 396– 401. find
  24. Meila, M. and J. Shi. 2001. A random walks view of spectral segmentation. In AISTATS. find
  25. Meila, M. 2001. The multicut lemma. Technical report, University of Washington. find
  26. Pantel, P. and D. Lin. 2002. Discovering word senses from text. In Proceedings of the eighth ACM SIGKDD international conference on Knowledge discovery and data mining, pages 613–619. ACM. find
  27. Pereira, F., N. Tishby, and L. Lee. 1993. Distributional clustering of English words. In Proceedings of ACL-93, pages 183–190, Morristown, NJ, USA. find our card
  28. Pragglejaz Group. 2007. MIP: A method for identifying metaphorically used words in discourse. Metaphor and Symbol, 22:1–39. find
  29. Preiss, J., T. Briscoe, and A. Korhonen. 2007. A system for large-scale acquisition of verbal, nominal and adjectival subcategorization frames from corpora. In Proceedings of ACL-2007, volume 45, page 912. find
  30. Resnik, P. 1993. Selection and Information: A Class-based Approach to Lexical Relationships. Ph.D. thesis, Philadelphia, PA, USA. find
  31. Resnik, P. 1997. Selectional preference and sense disambiguation. In ACL SIGLEX Workshop on Tagging Text with Lexical Semantics, Washington, D.C. find
  32. Rooth, M., S. Riezler, D. Prescher, G. Carroll, and F. Beil. 1999. Inducing a semantically annotated lexicon via EM-based clustering. In Proceedings of ACL 99, pages 104–111. find
  33. Schulte im Walde, S. 2006. Experiments on the automatic induction of German semantic verb classes. Computational Linguistics, 32(2):159–194. find our card
  34. Shutova, E. and S. Teufel. 2010. Metaphor corpus annotated for source - target domain mappings. In Proceedings of LREC 2010, Malta. find our card
  35. Shutova, E. 2010. Automatic metaphor interpretation as a paraphrasing task. In Proceedings of NAACL 2010, Los Angeles, USA. find
  36. Siegel, S. and N. J. Castellan. 1988. Nonparametric statistics for the behavioral sciences. McGraw-Hill Book Company, New York, USA. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/shutova-2010-metaphor-identification-verb-noun-clustering/