Large-scale Data To Send User Interest Based On A Composite Control Principle Adaptive Model

The fast development of WWW makes it being the most important information resource for data delivery and sharing worldwide. Huge information resource requires new technology and presents new research fields for traditional technology. How to manage the tremendous mount of information on WWW to meet the growing needs of personalize information is a new subject for our research.To meet the growing needs of users for personalize information requires the knowledge of user interests, which are described by user profiles. Therefore, the effectiveness of personalize information service provided by the system is determined by the fact whether or not the user profiles reflect user interests exactly.The method presented in this paper is based on the compound control principle combing feedback with feedforward and on the use of multiple interest vectors. User profiles are represented as multiple interest vectors, whose number, size, and elements may change adaptively with user access behavior; therefore, they can reflect the complex and changing interests of users. Although there has been significant research on user profiles, our approach is unique in that it can be tuned to make trade off between profile complexity and quality. This feature, together with its incremental nature increases its adaptability to the changes of user interests and makes our method suitable for use in large-scale information filtering applications such as push-based Web page dissemination.The main contributions of this paper can be summarized as follows. First, we introduce some issues relevant to our user modeling approach. Then, we propose a new incremental algorithm for constructing user profiles based on feedforward and user feedback after analysis of single profile vectors, multiple profile vectors and system based on pure feedback. We describe the approach for construction and maintenance of user profiles, and the frame and the functions of the compound control principle in detail. Finally, we discuss the test methodology and performance metrics of filtering system, and analyze our results to show the feasibility of our algorithm.