Using of context vectors in text classification

A. S. Eprev

Omsk F. M. Dostoevsky State University

Received September 27, 2010

 

Abstract. This paper presents an automatic text classification method based on k–nearest neighbors classifier and word sense disambiguation method based on WordNet and context vectors. Experimental results confirm the effectiveness of the approach.

 

Keywords: text classification, word sense disambiguation, context vector, WordNet.