The performance of text categorization methods based on keywords is often not satisfactory in terms of the categorization precision, which is due to the existence of synonyms and polysemes. Fortunately a new concept-based text categorization method is gaining more and more attention in recent years. In this paper, we propose a new concept mapping method, which is core word, oriented. The main idea is to firstly extract core words of class, then use How-net to map keyword space to concept space based on these core words, and finally complete the text categorization process in the concept space. The experimental results show that the text categorization precision can be distinctively improved using this concept mapping method.
[1]
Yiming Yang,et al.
A Comparative Study on Feature Selection in Text Categorization
,
1997,
ICML.
[2]
Yiming Yang,et al.
A re-examination of text categorization methods
,
1999,
SIGIR '99.
[3]
Qin Jin.
Feature Extraction in Text Categorization
,
2003
.
[4]
Nir Friedman,et al.
Bayesian Network Classifiers
,
1997,
Machine Learning.
[5]
Gong Xiu.
AN ACTIVE BAYESIAN NETWORK CLASSIFIER
,
2002
.
[6]
Pat Langley,et al.
An Analysis of Bayesian Classifiers
,
1992,
AAAI.
[7]
Yu Fang.
A Feature Selection Method for NB-based Classifier
,
2004
.