Knowledge-Based Approach for Word Sense Disambiguation Using Genetic Algorithm for Gujarati

This paper proposes a knowledge-based overlap approach, which uses Genetic algorithms (GAs) for Word Sense Disambiguation (WSD). Genetic Algorithms have been explored to solve search and optimization task in AI. WSD problem strives to resolve which the meaning of a polysemous word should be used in a surrounding context in a given text. Several approaches have been explored for WSD in English, Chinese, Spanish, and also for some Indian regional languages. Despite the extensive research in NLP for Indian Languages, research on WSD in Gujarati Language is very limited. Knowledge-based approach uses machine-readable knowledge source. We propose to use Indo-Aryan WordNet for Gujarati as a lexical database for WSD.