Efficient collaborative filtering using particle swarm optimization and K-harmonic means algorithm

Recommender systems have emerged in response to the problem of information overload and can help users to find their interest content. In order to alleviate the sparsity problem of recommender systems and increase the accuracy and diversity of recommendation results, we propose an effective collaborative filtering recommendation method. We adopt PSO-KHM algorithm which composes of Particle Swarm Optimization (PSO) and K-Harmonic means (KHM) to cluster users and then use improved collaborative filtering algorithm to compute the similarity between these clustered users. Further more, we take into account many influential factors in the process of similarity computation. The simulation results on two real-world datasets show that our algorithm achieves superior performance to existing methods. Copyright © 2015 American Scientific Publishers.