中文文本分类中的特征选择算法研究

比较了文档频率、信息增益、互信息、X^2统计量、期望交叉熵、文本证据权以及几率比等7种常用于文本分类的特征选择算法。实验采用国家“八六三计划”中文文文本语料库和Rocchio分类器对以上的特征选择算法分别进行评估，测评结果表明，几率比法的性能优于其它特征选择算法。

Study on feature selection methods in Chinese text categorization

This paper presents a study of seven feature selection methods that are commonly used in text categorization: document frequency, information gain, mutual information, X\+2 statistic, expected cross entropy, weight of evidence for text, and odds ratio. In order to evaluate these methods, experiments have been carried out combined with Chinese texts set in national 863 project and Rocchio classifier. The results of measured indicate that odds ratio method is superior to other methods.