A PERFORMANCE OF SVM WITH MODIFIED LESK APPROACH FOR WORD SENSE DISAMBIGUATION IN HINDI LANGUAGE

Sandy Garg, Anand Kumar Mittal

Abstract: WSD is a Technique used to find the correct meaning of a given word in any human language. Each human language has a problem called ambiguity of a word. To finds the correct meaning of any ambiguous word is easy for human but for a machine it is great issues No of work has done on WSD but not enough in Hindi language. My objective is to provide the training to the system so that it can easily find the correct meaning of the any ambiguous word in Hindi language .For this purpose I simple used a one existing technique name modified lesk approach and give its output to the SVM to get the better result and show that SVM is better in compare to modified Lesk Approach, In this paper I simply take nine Hindi ambiguous words and three different databases to show the result.

Keywords: Support Vector Machine, NLP, Word Sense Disambiguation, Modified Lesk approach, Comparison

DOI: https://doi.org/10.15623/ijret.2015.0408006