Common Group Correlation Analysis In The Application Research Of Personalized Service System

Abstract:With the rapid development of information technology and network technology, Internet applications and digital resources is more and more rich and perfect, for Internet applications, not only to provide information for users, but also need to provide individualized information service for different users, and improve the quality and level of application for Internet service, improve the utilization rate of resource information. Under this background, this thesis proposes a personalized service mechanism based on Web, and from the angle of system realization algorithm, were studied.This thesis briefly analyses the present situation of the research of personalized information service and the main recommendation algorithm, and then the problems of data sparsity and cold start traditional collaborative filtering algorithm in the proposed algorithm, collaborative filtering based on user interest spread, in this algorithm not only solved when the similarity of users interested in computational problems, and considering the effect of other neighbor user preference for the target user’s interest, to ensure the effectiveness of the proposed algorithm in sparse data situations, improves the performance of the system in a certain extent. At the same time, insufficient user social problems in the traditional system of personalized service, in order to solve this problem this thesis presents the analysis of common group association scheme based on user preference relation network construction, take the way of association data mining, on which the business database segmentation are common group, frequent itemsets mining then general group of community data obtained rules common group, finally the obtained rules in common group for recommendation service user personalization.Aiming at the problems in the personalized information service system, combining with the Internet application put forward the data mining technology, design to dig out the implementation of personalized information service system of user interest model, which adapt to the needs of the times, to meet the user’s current needs, improve the utilization of digital resources.