Web Usage Mining Technology Research

Web usage mining is one of the most important research directions in the web mining research field. Through its Web log to the effective data mining, in to explore the rules and user access patterns hidden behind the log data, thus improving Web site design, and providing users personalized services.This study as follows:details of the pre-process Web log, made page views filtering algorithm to preprocess page data.In Web clustering, the focus on the user clustering and web pages clustering. The user clustering, user access to a sequence similarity with the pages of time to visit similarity metrics, use Web Fuzzy clustering methods, resolve user services cluster. The web pages clustering, there is proposed based on association rules with the support and confidence to construct Web pages clustering method.Users interested in frequent access paths pattern mining, with user access pages and sites of interest in the specific topology, based on the Markov model to the interest of navigation model was presented.This paper completed a page view filtering algorithm design and analysis of the Web clustering research, and improvement based on the Markov model found interest in navigation mode, each accompanied by a proof of the effectiveness, practicality of the algorithm,.