ABSTRACT
The task of word sense disambiguation (WSD) is to choose, for a word with multiple senses in a particular context, which of its senses is the “correct” one for the context. In order to carry out the WSD tasks, a system must have sufficient background knowledge about various information sources. Although knowledge acquisition has been recognized as a very difficult problem in AI, it is undeniable that knowledge will be able to provide valuable help in language understanding. Up to now much research effort has been put on solving the WSD problem using a number of different approaches (e.g. Cottrell 1989, Hirst 1987, Stevenson 1990, Waltz 1985, Wu 1994a), and although some encouraging results can be found in the literature, progress is still considered limited.
