Research On Personalization Service Based On Combined Mining Of Web Contents And Web Server Logs

With the popularity of the Internet and the rapid development of information technology, the traditional way of acquiring Web information can not satisfy visitors who have different interests and different navigation purposes. It has become an urgent issue to get personalized knowledge effectively from mass Web information. Personalized service has become more important and has been paid more attention. Personalized service is based on the collection and analysis of user navigation information and provides active information recommendation.This paper focus on Web mining technology which is a very important technology in the area of personalized service. And a research is presented on a Web mining method which combines Web content mining and Web usage mining. In order to verify the performance of this Web mining method, this paper gives an integrated Web mining strategy. The mining strategy is used to automatic classify user navigation pattern. And it collects page theme information through textual terms of Web pages, combines Web server log files to establish user navigation pattern profiles, classifies user navigation patterns, and predicts users' future requests. Use the results of classification and prediction tasks to evaluate the performance of the integrated mining method through two measurements, classification accuracy A (C) and prediction accuracy A (P). The results show that, compared to separate Web mining methods, the classification accuracy and prediction accuracy of the integrated mining method are improved. The integrated mining method can better promote Web Personalized Recommendation.According to the project "Online combat system" of Dalian Public Security Bureau, and basing on combined mining of Web page contents and Web server logs, design an online personalized criminal information recommendation system. This paper describes the personalized information recommendation process which provides personalization service to different users. The system is designed to improve the efficiency of obtaining criminal investigation information through automatic recommendation, and plays a role of assisting rapid analysis and obtaining useful detection information.