Web Recommender System based on Consumer Behavior Modeling using Fuzzy Representation

Among various web related activities, web surfing, online purchases are more popular activities. Day by day World Wide Web proved to be the biggest source of information and contains a huge amount of data. But many times, most of the received or recommended data is irrelevant and inaccurate from users’ point of view. For users, it becomes necessary to use recommendation systems to discover and extract the desired information and resources. Web recommender systems predict the information needs of users. Recommender systems provide relevant recommendations to the users. This paper describes a Web recommender system that constructs a knowledge base using fuzzy temporal Web access patterns as input. The user's Web access habits and behavior is modeled using the knowledge base. The Knowledge base is used to generate association rules which are used to provide personalized recommendations to the user. The fuzzy representation is used to construct a knowledge base. Fuzzy logic is applied to requested resources of periodic patternbased Web access activities. These fuzzy patterns are used to generate association rules for Recommendation System. Experimental results on the session data of different Web sites showed the effectiveness of this approach.