Utilizing popularity characteristics for product recommendation in internet shopping malls

Abstract This chat presents a novel approach to automated product recommendation that uses the popularity characteristics of products. Although they play a significant role in the consumer purchasing process, there has been little attention paid to the use of popularity in recommendation research so far. In order to use popularity features, this paper develops a threedimensional model of popularity which is used to develop popularity classes of products. A hybrid recommendation system is then developed for movie recommendation, using the MovieLens dataset, where the genre information of movies is used together with the popularity information. The performance of the suggested system is evaluated in comparison with the widely adopted collaborative filtering method. The experiments show positive results under data sparsity and cold-starting conditions. Many interesting issues for further research are also suggested based on the popularity concept.