DOCUMENT CLUSTERING USING WORD SENSE DISAMBIGUATION

摘要

In computational linguistics, word sense disambiguation (WSD) is the problem of determining in which sense a word having a number of distinct senses is used in a given sentence. This paper handles text document clustering as one of the major tasks of text processing. Document clustering is the process of finding out groups of information from the text documents and cluster these documents into the most relevant groups. Large document corpus suffers from ambiguity problems like synonyms, polysemous and other semantic relations. For this reason we perform WSD task for all terms in all documents to get the best sense to be used as document features in the clustering process. Our experimental results proved that the efficiency of document clustering using WSD increases linearly with the size of the documents dataset. Different part of speech (POS) taggers were tested to determine the best; also the effect of different window sizes on WSD task was compared.