基于粗糙集的特征选择方法的研究

文本自动分类是指将文本按照一定的策略归于一个或多个类别中的应用技术。文本分类是文本挖掘的基础,而特征选择又是文本分类中的核心。论文分析了以前特征选择方法中由于特征数目过多而造成分类时间和精度不高的缺点,提出了一种基于粗糙集的特征选择方法,其特点是以特征在文本分类中的重要性对特征进行选择。最后通过实验验证了该算法,证明该方法是可行的。

The Research of the Feature Selection Method Based on Rough Set

The technique of text automatic category is to classify texts into one or more classes according to some strategy.Text categorization is the foundation of text mining,and feature selection is the core of text categorization.This paper analyses feature selection methods disadvantages which causes the low of times and accuracy of text categorization because of many features.We propose a feature selection method based on rough set.It is the characteristic of the method that features based on the importance in the text categorization are selected.It is shown by our experiment that this method is feasible.