Abstract:

Abstract Lexical ambiguity in machine translation and information retrieval is the challenge. Lexical ambiguity is caused by polysemous words where the word has multiple meanings. In Natural Language Processing before processing human commands the disambiguation of ambiguous commands should be done. The existing disambiguation methodologies disambiguate ambiguous sentences with available context information. The main identified problem is what if an ambiguous sentence doesn’t have enough information for disambiguation. The proposed model elaborates an adaptive sentence semantic similarity based complex network approach for identification of ambiguity and resolving it using semantic information. The discussed model represents the sentences of ambiguous documents as a vertex. The weighted complex network is constructed with respect to semantic similarities. The complex network is further processed for the ambiguous sentences having lack of context information. The main goal of this model is to provide an adaptive solution to lexical ambiguity of the paragraph or large document.

Keywords: