Abstract: With the development of Internet and information technology, network information scale is explosively increasing. Among various type of information, the type of texts occupy a considerable proportion. Therefore, efficient and rapid classification and processing of text information in the network become a key issue. The paper analyzes and compares SVM algorithm, Bayes algorithm and KNN algorithm. By the experiments of the three algorithms in Chinese text classification, the results indicate SVM algorithm is superior than KNN algorithm and Bayes algorithm, SVM algorithm is an excellent Chinese text classification algorithm.

Key words: Chinese text classification, SVM, Bayes, KNN