Word Sense Disambiguation of Hindi Text using Fuzzified Semantic Relations and Fuzzy Hindi WordNet
Goonjan Jain, Daya Krishan Lobiyal · 2019
Word Sense Disambiguation (WSD) refers to the task of finding intended meaning of a word in the text. We have used Fuzzy Hindi WordNet (FHWN) which is an extension of Hindi WordNet with added fuzzy semantic relations between the synsets. However, fuzziness in semantic relations of Hindi WordNet was not considered in FHWN. Also, it is mentioned in literature that the sematic relations are fuzzy and some relations are more important than the others. In this work, we have amalgamated these two concepts and extended FHWN. We assigned membership values to the semantic relations of FHWN. With added fuzziness, resource becomes closer to human usage of language. Thereafter, we computed various centrality measures and the sense with highest value is deduced as the correct sense. The experimental results proved the proposed method leads to better performance than Fuzzy Hindi WordNet and proves the value for the research.