Grouping Words Using Statistical Context

Abstract

This paper (cmp-lg/yymmnnn) has been accepted for publication in the student session of EACL-95. It outlines ongoing work using statistical and unsupervised neural network methods for clustering words in untagged corpora. Such approaches are of interest when attempting to understand the development of human intuitive categorization of language as well as for trying to improve computational methods in natural language understanding. Some preliminary results using a simple statistical approach are described, along with work using an unsupervised neural network to distinguish between the sense classes into which words fall.