Social-Based Collaborative Recommendation: Bees Swarm Optimization Based Clustering Approach

摘要

This paper focuses on the recommendation of items in social networks, through which the social information is formalized and combined with the collaborative filtering algorithm using an optimized clustering method. In this approach, users are clustered from the views of both user similarity and trust relationships. A Bees Swarm optimization algorithm is designed to optimize the clustering process and therefore recommend the most appropriate items to a given user. Extensive experiments have been conducted, using the well-known Epinions dataset, to demonstrate the effectiveness of the proposed approach compared to the traditional recommendation algorithms.