A Genetic Algorithm Based Approach for Word Sense Disambiguation Using Fuzzy WordNet Graphs

Abstract

Due to the ever-evolving nature of human languages, the ambiguity in it needs to be dealt with by the researchers. Word sense disambiguation (WSD) is a classical problem of natural language processing which refers to identifying the most appropriate sense of a given word in the concerned context. WordNet graph based approaches are used by several state-of-art methods for performing WSD. This paper highlights a novel genetic algorithm based approach for performing WSD using fuzzy WordNet graph based approach. The fitness function is calculated using the fuzzy global measures of graph connectivity. For proposing this fitness function, a comparative study is performed for the global measures edge density, entropy and compactness. Also, an analytical insight is provided by presenting a visualization of the control terms for word sense disambiguation in the research papers from 2013 to 2018 present in Web of Science.