Topic modeling for word sense induction

Abstract

In this paper, we present a novel approach to Word Sense Induction which is based on topic modeling. Key to our methodology is the use of word-topic distributions as a means to estimate sense distributions. We provide these distributions as input to a clustering algorithm in order to automatically distinguish between the senses of semantically ambiguous words. The results of our evaluation experiments indicate that the performance of our approach is comparable to state-of-the-art methods whose sense distinctions are not as easily interpretable. © 2013 Springer-Verlag.