Combined lesk-based method for words senses disambiguation

Abstract

Word sense disambiguation (WSD) is a linguistically-based mechanism for defining automatically the correct sense of a word in the context. Lesk-based methods use context information for WSD. In this paper, we propose a method of word sense disambiguation based on the combination of the original Lesk method and the simplified one with additional application of large lexical resources, like synonym dictionaries, ontologies, etc., for calculations of sense plausibility. We present experimental results that show that our method has better precision than the baseline Lesk-based methods. © 2006 IEEE.