| |
| | AI Magazine: Corpus-based approaches to semantic interpretat... @ HighBeam Research (Site not responding. Last check: 2007-10-11) |
 | | In disambiguating the noun interest, the possible values of the verb-object feature are the verbs (such as sold) that stand in a verb-object syntactic relation with interest. |
 | | The goal is to be able to determine that the nouns tea, milk, wine, and so on, are most similar to the beverage sense of coffee without requiring that tea, milk, wine, and so on, be manually tagged with the correct sense in the training corpus. |
 | | Verbs are typically harder to disambiguate than nouns, and disambiguation accuracy for words chosen from a test corpus composed of a wide variety of genres and domains, such as the Brown corpus, is lower than the accuracy of words chosen from, say, business articles from the Wall Street Journal (Ng and Lee 1996). |
| www.highbeam.com /library/doc0.asp?DOCID=1G1:20475199&refid=holomed_1 (11012 words) |
|