A product recommendation system based on hybrid approach

Product recommendation system are of great significance in e-commerce. Current recommendation approaches made great achievements, but do not consider different people's similar learning trajectories in the same major. This paper focuses on "book" as a product. In order to recommend books more accurately, this paper proposes a personalized book recommendation algorithm that is based on the time sequential collaborative filtering and content based collaborative filtering. In this algorithm we consider important factors: time sequence information of purchasing books of different users in database and content of the user profile. Here, results of both the algorithms are unified for obtaining the more accurate list of recommended books. This hybrid book recommendation system satisfies user by providing best and efficient books recommendations.