Research Of Web User Access Patterns In Fuzzy And Rough Environments

Web log is investigated to find user interesting access patterns in web sequential analysis and clustering analysis. Knowledge extracted from web logs can be used to improve design of web sites, analyze system performance as well as network communication, understand user reaction and motivation, and build adaptive web sites. Thorough researches are made to understand the characteristic of user access patterns by web sequential analysis and clustering analysis. Major works are listed as follows:The algorithm based on Frequent Link and Access Tree (FLaAT) is proposed to mine frequent user access patterns. FLaAT stores all user access information and considers the merge of same routes with different prefixes to make the integration of mined information during search of frequent user access patterns.Another algorithm is proposed to mine fuzzy preferred access patterns with minimum support and preference in fuzzy environment, in which time durations on a web page are characterized as a fuzzy linguistic variable. Fuzzy preference patterns with fuzzy time durations more deeply reveal user interest and preference. In addition, linguistic inputs and outputs are more natural and more similar to people reasoning.An efficient algorithm is provided to mine user preferred access patterns based on the concept, (Fuzzy Web page preference function), which is along with support to disclose user interest and preference. It considers all possible factors, weight of web page, relative access frequency of web page, and time durations on web page. Thus it more deeply reveals user interest and preference.A fuzzy rough approximation-based approach is proposed to cluster user access patterns from web logs. In the process of clustering user access patterns, each user access pattern is denoted by a fuzzy vector representing visited web pages and time durations during a surfing. Finally rough approximation approach is adopted to cluster user access patterns denoted by fuzzy vectors.A rough k-means clustering algorithm based on properties of rough variable is developed to group web access patterns. Each web access pattern from web logs is transformed as corresponding fuzzy web access pattern. Finally rough k-means clustering algorithm is adopted to cluster user access patterns denoted...