Collaborative Topic Regression with social trust ensemble for recommendation in social media systems

摘要: Social media systems provide ever-growing huge volumes of information for dissemination and communication among communities users, while recommender aim to mitigate overload by filtering providing users the most attractive relevant items from information-sea. This paper aims at compound recommendation engine social systems, focuses on exploiting multi-sourced (e.g. networks, item contents user feedbacks) predict ratings make recommendations. For this, we suppose users' decisions adopting are affected both their tastes favors trusted friends, extend Collaborative Topic Regression jointly incorporates trust ensemble, topic modeling probabilistic matrix factorization. We propose corresponding approaches learning latent factors items, as well additional parameters be estimated. Empirical experiments Lastfm Delicious datasets show that our model is better more robust than state-of-the-art methods making recommendations in term accuracy. Experiments results also reveal some useful findings enlighten development media.