Personalized Web Recommendation Combining User-centered Collaborative Technique with URL Weighting
Web usage mining has become very popular in various business areas for learning more about the users' browsing behavior and recommending the perfect product in which the user is interested in. At present there are many systems that recommend for the users on web usage mining, but most of the systems suffer from inappropriate scalability, which would lead to very weak recommendations. In this paper, the authors proposed a new technique that gives emphasis on page view weighting based on transaction timing and building a session pattern graph for each session. This technique provides the scope for better scalability and also provides effective number of recommendations with remarkable accuracy.