Paper: A Practical Solution To The Problem Of Automatic Part-Of-Speech Induction From Text

ACL ID P05-3020
Title A Practical Solution To The Problem Of Automatic Part-Of-Speech Induction From Text
Venue Annual Meeting of the Association of Computational Linguistics
Session System Demonstration
Year 2005
Authors

The problem of part-of-speech induction from text involves two aspects: Firstly, a set of word classes is to be derived auto- matically. Secondly, each word of a vo- cabulary is to be assigned to one or sev- eral of these word classes. In this paper we present a method that solves both problems with good accuracy. Our ap- proach adopts a mixture of statistical me- thods that have been successfully applied in word sense induction. Its main advan- tage over previous attempts is that it re- duces the syntactic space to only the most important dimensions, thereby almost eli- minating the otherwise omnipresent prob- lem of data sparseness.