DMCA
Random-Walk Term Weighting for Improved Text Classification (2006)
Cached
Download Links
| Venue: | In Proceedings of TextGraphs: 2nd Workshop on Graph Based Methods for Natural Language Processing. ACL |
| Citations: | 14 - 0 self |
BibTeX
@INPROCEEDINGS{Hassan06random-walkterm,
author = {Samer Hassan and Carmen Banea},
title = {Random-Walk Term Weighting for Improved Text Classification},
booktitle = {In Proceedings of TextGraphs: 2nd Workshop on Graph Based Methods for Natural Language Processing. ACL},
year = {2006},
pages = {53--60}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
This paper describes a new approach for estimating term weights in a text classification task. The approach uses term cooccurrence as a measure of dependency between word features. A random walk model is applied on a graph encoding words and co-occurrence dependencies, resulting in scores that represent a quantification of how a particular word feature contributes to a given context. We argue that by modeling feature weights using these scores, as opposed to the traditional frequency-based scores, we can achieve better results in a text classification task. Experiments performed on four standard classification datasets show that the new random-walk based approach outperforms the traditional term frequency approach to feature weighting. 1
Keyphrases
improved text classification random-walk term weighting text classification task term weight co-occurrence dependency traditional frequency-based score word feature new random-walk random walk model term cooccurrence standard classification datasets new approach feature weight traditional term frequency approach particular word feature
