Word sense disambiguation using WordNet Lexical Categories

Abstract

In this paper a methodology for disambiguating the word senses of polysemous words using Lexical Categories present in WordNet is presented. WordNet is a commonly used English lexical database. The algorithm is applied to the data scraped from Wikipedia articles. The representative context used in the algorithm is extracted from the Wikipedia pages of the words belonging to the category. The lexical category of the given word is determined using the words in its neighbouring context. After finding the lexical category of the word, the correct sense is found using a modified version of Lesks Algorithm. The output word sense correspond to those available in WordNet. � 2016 IEEE.