Research On Personalized Recommendation Methodology Based On WUM

As everyone knows, the behavior of customer visiting web site may manifest his preference, the performance of web site, the pattern of navigation and service. So, we can obtain useful pattern from web visit log through Web Usage Mining(WUM). The research of personalized service based on WUM is one of main research direction of WUM. The personalized recommendation is come into being for.the problem that visitor who encounter abundant information may be at loose ends.This paper study the theory, method and process of WUM at first. The process of WUM may include Data Preprocessing, Pattern Discovery, Pattern Analysis , Pattern applications. Pattern Discovery is the kernel of WUM, it can include statistical analysis, path analysis, association rules, sequence pattern, classified, clustering etc.And then we research the Personalized Recommendation through Recommender Systems, Recommendation method and the hotspot about research, and we also analyze the status quo of research and application. Recommender Systems can be divided into E-commerce and search engine. The recommendation method may have seven kinds and the hotspot of recommender mainly have five question.In this paper, we analyze the research of ACO and the theory of rough set. And then we design personalized recommendation method based on ACO, and we also design personalized recommendation method on the theory of rough set. And we combine this two recommendation and give Collaborative Filtering Recommendation.