Unsupervised Parts-of-Speech Induction for Bengali

Abstract

We present a study of the word interaction networks of Bengali in the framework of complex networks. The topological properties of these networks reveal interesting insights into the morpho-syntax of the language, whereas clustering helps in the induction of the natural word classes leading to a principled way of designing POS tagsets. We compare different network construction techniques and clustering algorithms based on the cohesiveness of the word clusters. Cohesiveness is measured against two gold-standard tagsets by means of the novel metric of tag-entropy. The approach presented here is a generic one that can be easily extended to any language.

Resources

Details