Feature optimization approach for improving the
collaborative filtering performance using particle
swarm optimization <Abstract> Collaborative filtering (CF) is currently one of the most popular and widely used recommendation techniques. It generates personalized predictions based on the assumption that users with similar tastes prefer similar items. It assumes that all features (users or items) have an equal importance in prediction formulation. However, if the importance of features is different, the later assumption will lead to inaccurate predictions. In this paper, a feature weighting method for cluster-based CF recommender systems is proposed. In this method, the particle swarm optimization (PSO) algorithm is utilized to estimate the features importance and allocate their weights accordingly. A prediction model, utilizing the spectral clustering technique in both user-based and item-based CF, is used to evaluate these weights and to predict the unknown ratings. In this work, the suggested prediction model utilizes the features weights to enhance the similarity measure and cluster formulation. The results of experiments demonstrate that the proposed method can effectively improve the quality of recommendation and eliminating main CF limitations. |