Improved Unsupervised POS Induction through Prototype Discovery
Cached
Download Links
by
Omri Abend
,
Roi Reichart
| Citations: | 3 - 0 self |
BibTeX
@MISC{Abend_improvedunsupervised,
author = {Omri Abend and Roi Reichart},
title = {Improved Unsupervised POS Induction through Prototype Discovery},
year = {}
}
OpenURL
Abstract
We present a novel fully unsupervised algorithm for POS induction from plain text, motivated by the cognitive notion of prototypes. The algorithm first identifies landmark clusters of words, serving as the cores of the induced POS categories. The rest of the words are subsequently mapped to these clusters. We utilize morphological and distributional representations computed in a fully unsupervised manner. We evaluate our algorithm on English and German, achieving the best reported results for this task. 1







