Date Added: Nov 2009
Web usage mining has recently emerged as a basis for extracting useful user access pattern information, such as user profiles, from enormous amounts of Web log data for web site personalization. A profile can consist of a set of URLs that are relevant to the sessions assigned to a given cluster. Once these profiles are discovered, the authors can be exploited as part of an automated personalization on the website, by treating them as summarized user models against which all future user sessions are compared. This paper focuses on obtaining user profiles based on intelligent rough clustering techniques.