Elamur Research⌕
☰
home / papers / 1997 / A Frame-Semantic Approach to Semantic Annotation

A Frame-Semantic Approach to Semantic Annotation

Summary

Corpus semantic annotation needs a tagset. Generic thematic roles are disputed in number, and roles tailored to individual predicates threaten to multiply without limit. The authors work in the FrameNet project, then just starting at Berkeley (ICSI). They argue that frame semantics gives a principled way to choose and organize tags. Words are described as evoking frames, which are schematized scenarios such as a commercial transaction or a health-care event. Sentence constituents are then labeled with the frame elements they realize. Frames can form hierarchies, so a subframe (e.g., buying real estate with a loan) inherits elements from its parent, which keeps tagsets compact. Using a Health frame, the paper shows how corpus evidence and metaphor justify distinctions such as DISEASE versus WOUND. It also shows how combinations of elements (Frame Element Groups) characterize verb senses, and how omitted or doubly realized elements should be marked. No experimental results are reported. The contribution is a methodology and a plan for a hand-annotated starter lexicon that could later support automatic tagging and word sense disambiguation.

Key points

  • The project context is FrameNet ("Tools for lexicon building"), led by PI Charles J. Fillmore at ICSI Berkeley. It was funded by NSF grant IRI 96-18838 and started on March 1, 1997.
  • The CommercialTransaction frame has the elements BUYER, SELLER, PAYMENT and GOODS. The RealEstateTransaction frame inherits these and adds BORROWER, LOAN and LENDER, following inheritance ideas similar to the ACQUILEX Project.
  • The Health frame elements include HEALER, PATIENT, DISEASE, WOUND, BODYPART, SYMPTOM, TREATMENT and MEDICINE. DISEASE and WOUND stay separate because diseases are "cured" while wounds "heal." Metaphorical objects support the split: ills goes with cure, and rift, schism and breach go with heal.
  • The annotation records the status of missing elements, such as an implied disease in "The doctor cured me." A single constituent can carry several roles, as with leper or diabetic, tagged Pd (PATIENT plus DISEASE).
  • The planned starter lexicon covers about 5,000 lexical items, with token frequencies of FEGs per word sense. Later partial automation would draw on WordNet synonyms and hypernyms and on thesauri.

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: Word sense disambiguation · Semantic role labeling

Cited by

27 papers of our library cite it (of the 140673 whose bibliographies we read); 7 of them have a page here, 20 not yet.

Cite

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

@inproceedings{lowe1997frame,
  title = {{A Frame-Semantic Approach to Semantic Annotation}},
  author = {Lowe, John B.},
  booktitle = {Tagging Text with Lexical Semantics: Why, What, and How?},
  year = {1997},
  url = {https://aclanthology.org/W97-0204/}
}

References

14 references, 0 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Ted Briscoe, Valeria De Paiva, and Ann Copestake, editors. 1993. Inheritance, Defaults and the Lexicon. Studies in Natural Language Processing. Cambridge University Press, Cambridge, England. find
  2. Charles J. Fillmore and B.T.S. Atkins. 1992. Towards a frame-based lexicon: the semantics of risk and its neighbors. In A. Lehrer and E. F. Kittay, editors, Prames, Fields and Contrasts, pages 75-102. Lawrence Erlbaum Associates, Hillsdale, NJ. find
  3. Charles J. Fillmore and B.T.S. Atkins. 1994. Starting where the dictionaries stop: the challenge for computational lexicography. In B.T.S. Atkins and A. Zampolli, editors, Computational Approaches to the Lexicon. Oxford University Press, New York. find
  4. Charles J. Fillmore. 1968. The case for case. In Universals in linguistic theory, pages 1-90. Holt, Rinehart and Winston, New York. find
  5. Charles J. Fillmore. 1977a. The need for a frame semantics within linguistics, statistical Methods in Linguistics, pages 5-29. find
  6. Charles J. Fillmore. 1977b. Scenes-and-frames semantics. In Antonio Zampolli, editor, Linguistics Structures Processing, volume 59 of Fundamental Studies in Computer Science, pages 55-82. North-Holland Publishing. find
  7. Charles J. Fillmore. 1982. Frame semantics. In Linguistics in the morning calm, pages 111-137. Hanshin Publishing Co., Seoul, South Korea. find
  8. Ray S. Jackendoff. 1994. Patterns in the mind: language and human nature. Basic Books, New York. find
  9. Beth Levin. 1993. English Verb Classes and Alternations: A Preliminary Investigation. University of Chicago Press, Chicago. 1993. Selection and Information: A Class-Based Approach to Lexical relationships. find
  10. Philip Resnik. University of Pennsylvania dissertation. find
  11. Charles Ruhl. 1989. On monosemy : a study in lingusitic semantics. Albany, N.Y.: State University of New York Press. find
  12. Roger C. Schank. 1975. Conceptual information processing. North-Holland., New York. find
  13. John F. Sowa. 1984. Conceptual structures: information processing in mind and machine. Addison-Wesley systems programming series. Addison-Wesley, Reading, Mass. find
  14. Michael Sperberg-McQueen and Lou Burnard. (eds.) 1994. Guidefines for electronic text encoding and interchange (TEI P3). ACH, ACL, ALLC, Chicago. m mm m [] m m [] m [] find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/papers/lowe-1997-frame-semantic-approach-semantic-annotation/