An Efficient Cold Start Solution Based on Group Interests for Recommender Systems

This paper proposes an efficient solution for the cold start problem in recommender systems. This problem occurs with new users who do not have sufficient information in their records. This would cause the recommender system to fail in providing recommendations to these users. This problem is one of the common and important problems in recommender systems. Although some solutions have been proposed to solve it in the literature, these solutions would not work properly in some scenarios because they do not concentrate on finding the actual interests of the users and the hidden motives behind their behavior. Our proposed solution uses the hidden interests of the group to which the target user belongs to provide recommendations for that user. The experiments show that our proposed solution is efficient in terms of searching time and space consumption.