Hybrid Recommendation System using Particle Swarm Optimization and User Access Based Ranking
G. Sumathi, Selvaraju Sendhilkumar, G. S. Mahalakshmi · 2016
This paper introduces a novel architecture for a new user recommendation system which is based on Particle Swarm Optimization (PSO) algorithm and User Access Based Ranking (UABR) approach. The Recommendation System (RS) is an efficient tool for providing the relevant pages to the users. A vital issue for the RS that has enormously captured the attention of researchers is the cold-start problem. This issue is related to recommendations for new users. For new users, the system does not have data about their preferences in order to make recommendations. We proposed a technique with the swarm intelligence approach of Particle Swarm Optimization in combination with User access based ranking algorithm for the new user recommendation. The PSO algorithm is applied to user grouping. Using this approach, users with similar searching travels are gathered into the same cluster. Recommendations for new user are produced through the user access based ranking algorithm. The results of experiments exhibits that the proposed strategy can effectively enhance the quality of recommendation with the better precision, recall and F_Score values.