Published Paper Details:

In natural language processing and understanding, semantic processing is an important task. In semantic processing some words has multiple senses (meanings) which are unrelated with each other. These multiple senses possess critical problems to linguists and they create ambiguity in sentence. Word sense disambiguation accepted this challenge. It is one of the central challenges in NLP and occurs in all the languages. Human can easily disambiguate the words but machine can not. WSD has numerous applications in machine translation, information retrieval, question-answering etc. The ambiguity can be lexical and semantic. In NLP, Word Sense Disambiguation(WSD) is the task of perfectly assigning the acceptable, correct sense (meaning) to the words having multiple senses in the given natural language text. WSD is categorized in three types wiz. Knowledge base, machine learning and Hybrid approach. The work carried out on Marathi language is limited. In the proposed work, we are resolving the ambiguity in Marathi words based on their senses and their context hybrid approach. Hybrid approach consist of modified Lesk with support Vector Machine.