Combining Distributional and Morphological Information for Part of Speech Induction


Authors

Clark, Alexander

Year

2003

Abstract

In this paper we discuss algorithms for clustering words into classes from unlabelled text using unsupervised algorithms, based on distributional and morphological information. We show how the use of morphological information can improve the performance on rare words, and that this is robust across a wide range of languages.

link

local copy of pdf

citation in bibtex

comments powered by Disqus