User Feedback Based Evaluation of a Product Recommendation System Using Rank Aggregation Method

The proliferation of the Internet has changed the daily life of a common man. There is a diverse effect of rapid growth of Internet in the daily life. The influence of Internet has changed the way we live and even the way we think. The use of the Internet for purchasing different products of the daily needs has increased exponentially in recent years. Now customers prefer online shopping for the acquisition of the various products. But the huge e-business portals and increasing online shopping sites make it difficult for the customers to go for a particular product. It is very common practice that a customer wishes to know the opinion of other consumers who already have acquired the same product. Therefore we tried to involve the human judgment in recommending the products to the users using implicit user feedback and applied a rank aggregation algorithm on these recommendations. In this paper we chose few products and their respective ranks arbitrarily taken from previous work. For obtaining user’s purchase activities a vector feedback is taken from the user and on the basis of their feedback, products are scored; hence they are again ranked which gives each user’s ranking. We propose a rank aggregation algorithm and apply it on individuals ranking to get an aggregated final users’ ranking. Finally we evaluate the system performance using false negative rates, false positive rates, and precision. These measures show the effectiveness of the proposed method.