Method of Chinese Text Categorization Based on Variable Precision Rough Set

Text categorization is an important research direction of current information retrieval. The traditional text classification method use the support vector machine (SVM) and the Bayes classification algorithm (etc). On the basis of Rough Set on text categorization, this paper put forward the idea of variable precision rough set model for Chinese text categorization, which use the attribute reduct algorithm based on the importance of attributes as heuristic information to reduct the feature subset of the text, and analyses the influence of error classification rate on text classification. It can increase the flexibility of text categorization and improve the accuracy of text classification by setting different value to find the the best.