Elamur Research⌕
☰
home / papers / 2004 / CorMet: A Computational, Corpus-Based Conventional Metaphor Extraction System

CorMet: A Computational, Corpus-Based Conventional Metaphor Extraction System

Summary

CorMet addresses the automatic discovery of conventional metaphors. These are systematic mappings that carry structure from a concrete source domain, such as liquids or warfare, into a target domain, such as money or medicine. Earlier computational metaphor models relied on small hand-built knowledge bases and covered only a few examples. CorMet uses only WordNet plus large web-mined corpora. For each user-specified domain, it builds a corpus from Google queries that combine random domain keywords. It then parses the pages, extracts verb case frames, and selects verbs that are unusually frequent compared with general English. Their domain-specific selectional preferences are learned with Resnik's algorithm followed by clustering. When a source-domain verb consistently takes a different concept class in the target domain, CorMet proposes an interconcept mapping. One example is pour, which takes funds in FINANCE but liquids in LAB. Clusters of such mappings indicate higher-level metaphors. Tested against a subset of the Master Metaphor List, the system recovers some listed metaphors with reasonable accuracy despite noisy components. This shows that a shallow corpus-based approach is viable, though not yet a practical NLP tool.

Key points

  • Data and corpora: Corpora are mined dynamically from the web. Each query returns up to 10,000 links, of which at most 3,000 documents are fetched. Across the 13 domains studied, about 75% of fetched documents were domain-relevant, based on a hand-checked sample of 100 per domain.
  • Pipeline: After HTML and scripts are stripped from the pages, the text is syntactically analyzed using Sekine and Grishman's (1995) Apple Pie Parser. Case frames are extracted using syntactic templates.
  • Characteristic verbs: Words are stemmed with the Porter stemmer. Each stem's frequency is compared with the BNC frequency list (Kilgarriff 2003), and the top 400 verb stems per domain are kept. Stems that score high in several domains, such as e-mail, are discarded.
  • Selectional preferences: These are computed with Resnik's (1993) relative-entropy measures, selectional-preference strength and selectional association, over WordNet nodes. Preferences are compared via cosine similarity.
  • Knowledge sources and scope: WordNet's noun hierarchy is the only knowledge base, and its verb information is deliberately ignored. The method works best for concrete, specific domains and common mappings. It does not detect every metaphor instance.

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

Topics: Idioms & multiword expressions

Institutions

🎓Brandeis UniversityWaltham · United States27 works · 4 999 citations

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.1162/089120104773633376

@article{mason2004cormet,
  title = {{CorMet: A Computational, Corpus-Based Conventional Metaphor Extraction System}},
  author = {Mason, Zachary J.},
  journal = {Computational Linguistics},
  year = {2004},
  doi = {10.1162/089120104773633376},
  url = {https://doi.org/10.1162/089120104773633376}
}

References

26 references, 2 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Carroll, J., and D. McCarthy. 2000. Word sense disambiguation using automatically acquired verbal preferences. Computers and the Humanities, 34(1–2). find
  2. Cho, See-Young. 1993. Metaphor and cultural coherence. In Proceedings of the 27th Conference on Cross-Language Studies and Contrastive Linguistics. find
  3. Fass, Dan. 1986. Collative semantics: An approach to coherence. Memorandum in Computer and Cognitive Science MCCS-86-56, New Mexico State University, New Mexico. find
  4. Fass, Dan. 1987. Collative semantics: An overview of the current meta5 program. Memorandum in Computer and Cognitive Science MCCS-87-112, New Mexico State University, NM. find
  5. Fass, Dan. 1988. Collative semantics: A semantics for natural language processing. Memorandum in Computer and Cognitive Science MCCS-88-118, New Mexico State University, NM. find
  6. Fass, Dan. 1991. Met: A method for discriminating metonymy and metaphor by computer. Computational Linguistics, 17(1):49–90. find our card
  7. Fellbaum, Christiane, editor. 1998. WordNet: An Electronic Lexical Database. MIT Press, Cambridge, MA. find
  8. Goatly, Andrew. 1997. The Language of Metaphors. Routledge, London. find
  9. Gruber, Jeffrey. 1976. Lexical Structures in Syntax and Semantics. Amsterdam, North-Holland. find
  10. Jain, Anil K., M. Narasimha Murty, and Patrick J. Flynn. 1999. Data clustering: A review. ACM Computing Surveys, 31(3):264–323. find
  11. Kilgarriff, Adam. 2003. BNC word frequency list. Available online at http://www.itri.brighton.ac.uk/Adam. Kilgarriff/bnc-readme.html. ↗
  12. Kucera, Henry. 1992. Brown corpus. In S. Shapiro, editor, Encyclopedia of Artificial Intelligence, volume 1. Wiley, New York, pages 128–130. find
  13. Lakoff, George. 1993. The contemporary theory of metaphor. In Andrew Ortony, editor, Metaphor and Thought. Cambridge University Press, Cambridge. find
  14. Lakoff, George, Jane Espenson, and Alan Schwartz. 1991. The master metaphor list. Draft 2nd ed. Technical Report, University of California at Berkeley. find
  15. Lenat, Douglas. 1995. Cyc: A large-scale investment in knowledge infrastructure. In Communications of the ACM, 38:11. find
  16. Li, Hang and Naoki Abe. 1998. Generalizing case frames using a thesaurus and the MDI principle. Computational Linguistics, 24(2):217–244. find
  17. Marcus, Mitchell P., Beatrice Santorini, and Mary Ann Marcinkiewicz. 1993. Building a large annotated corpus of English: The Penn Treebank. Computational Linguistics, 19:313–330. find our card
  18. Martin, James. 1990. A Computational Model of Metaphor Interpretation. Academic Press. find
  19. Martin, James. 1994. Metabank: A knowledge base of metaphoric language conventions. Computational Intelligence, 10(2):134–149. find
  20. Mason, Zachary. 2002. A Computational, Corpus-Based Metaphor Extraction System. Ph.D. thesis, Brandeis University. find
  21. Peters, Winn and Ivonne Peters. 2000. Lexicalised systematic polysemy in WordNet. In Proceedings of the Second International Conference on Language Resources and Evaluation, Athens. find
  22. Porter, Martin F. 1980. An algorithm for suffix stripping. Program, 14(3):130–137. find
  23. Resnik, Philip. 1993. Selection and Information: A Class Based Approach to Lexical Relationships. Ph.D. thesis, University of Pennsylvania. find
  24. Sekine, Satoshi and Ralph Grishman. 1995. A corpus-based probabilistic grammar with only two non-terminals. In Proceedings of the Fourth International Workshop on Parsing Technology, Prague, Czech Republic. find
  25. Wilks, Yorick. 1975. A preferential, pattern-seeking, semantics for natural language inference. Artificial Intelligence, 6:53–74. find
  26. Wilks, Yorick. 1978. Making preferences more active. Artificial Intelligence, 11(3):197–223. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/mason-2004-cormet-computational-corpus-based-conventional/