Recommender systems, inferring users' preferences from their historicalactivities and personal profiles, have been an enormous success in the lastseveral years. Most of the existing works are based on the similarities ofusers, objects or both that derived from their purchases records in the onlineshopping platforms. Such approaches, however, are facing bottlenecks when theknown information is limited. The extreme case is how to recommend products tonew users, namely the so-called cold-start problem. The rise of the onlinesocial networks gives us a chance to break the glass ceiling. Birds of afeather flock together. Close friends may have similar hidden pattern ofselecting products and the advices from friends are more trustworthy. In this paper, we integrate the individual's social relationships intorecommender systems and propose a new method, called Social Mass Diffusion(SMD), based on a mass diffusion process in the combined network of users'social network and user-item bipartite network. The results show that the SMDalgorithm can achieve higher recommendation accuracy than the Mass Diffusion(MD) purely on the bipartite network. Especially, the improvement is strikingfor small degree users. Moreover, SMD provides a good solution to thecold-start problem. The recommendation accuracy for new users significantlyhigher than that of the conventional popularity-based algorithm. These resultsmay shed some light on the new designs of better personalized recommendersystems and information services.
展开▼
