Integrating heterogeneous information via flexible regularization framework for recommendation

Recently, there is a surge of social recommendation, which leverages social relations among users to improve recommendation performance. However, in many applications, social relations are very sparse or absent. Meanwhile, the attribute information of users or items may be rich. It is a big challenge to exploit this attribute information for the improvement of recommendation performance. In this paper, we organize objects and relations in recommender system as a heterogeneous information network and introduce meta-path-based similarity measure to evaluate the similarity of users or items. Furthermore, a matrix factorization-based dual regularization framework SimMF is proposed to flexibly integrate different types of information through adopting users' and items' similarities as regularization on latent factors of users and items. Extensive experiments not only validate the effectiveness of SimMF but also reveal some interesting findings. We find that attribute information of users and items can significantly improve recommendation accuracy, and their contribution seems more important than that of social relations. The experiments also reveal that different regularization models have obviously different impacts on users and items.