|
Conceptual Association for Compound Noun Analysis
Mark Lauer
Microsoft Institute
When: Wednesday, 8th February 1994
Time: 1:00pm
Where: Lab 1, Microsoft Institute
Abstract:
This talk describes research toward the automatic interpretation of compound nouns using corpus statistics. An initial study aimed at syntactic disambiguation is presented. Corpus derived lexical associations have proven successful for prepositional phrase attachment (Hindle and Rooth, 1993) suggesting that a similar approach may prove useful for compound noun analysis. The approach presented bases associations upon thesaurus categories rather than individual words, a technique described elsewhere as conceptual association (Resnik and Hearst, 1993). Association data is gathered from unambiguous cases extracted from a corpus and is then applied to the analysis of ambiguous compound nouns. While the work presented is still in progress, a first attempt to syntactically analyse a test set of 244 examples shows 75% correctness. Future work is aimed at improving this accuracy and extending the technique to assign semantic role information, thus producing a complete interpretation.
Enquiries: Maria Milosavljevic 9850 6345 mariam@mpce.mq.edu.au
| Last modified: July 1997 |