Blogger’s Interest Mining Based on Chinese Text Classification

Abstract

In this paper, a new blogger’s interest mining module is proposed, which is based on Chinese text classification. In fact, the problem of the interest mining is transformed into the problem of Chinese text categorization. Before the Chinese text categorization, the text is pre-processed for the text representation. The Chinese text is represented in vector space model and classified by support vector machine classification, while filter algorithm which filters the unrelated interest text is proposed. After the filtering, the text can get it’s interest category. Finally the new module has been made use of to carry out an interest mining experiment, and the other experiment which has not filter algorithm is also carried in order to compare with the new module. The two experimental results show that the support vector machine is a effective algorithm, and the comparing data of the two experiments shows that new module make the interest mining more effective.