Knowledge Based Approaches to Nepali Word Sense Disambiguation

A word may have multiple senses and the challenge is to find out which particular sense is appropriate in a given context. Word sense disambiguation(WSD) resolves this ambiguity by finding out which particular sense of a word is appropriate in a given context. WSD is of critical importance in the areas of machine translation, information retrieval, speech processing etc. In this paper we present some approaches to Word sense disambiguation in Nepali using Nepali WordNet. These approaches are overlap based approach and conceptual distance and semantic graph based approach which falls under Knowledge based approach. Conceptual distance and semantic graph distance are used as a measures to score our WSD algorithm.