Word Sense disambiguation using Wordnet-based contextual representation of sentences

Abstract

In this study, we perform word sense identification of words with polysemy in English sentences. Although recent developments in transducer technology have greatly advanced word sense disambiguation systems in English sentences, more semantic context information should be required to further improve the accuracy of the system. We investigate the improvement of word sense disambiguation accuracy by providing Wordnet example sentences as sentence-level contextual information for ambiguous words, for which existing methods have accuracy problems.