A conceptual architecture with trust consensus to enhance group recommendations

Recommender Systems have been studied and developed as an indispensable technique of the Information Filtering field. A drawback of traditional user-item systems is that most recommenders ignore connections consistent with the real world recommendations. Furthermore, trust-based approaches ignore the group modeling and do not respect the users' individualities in a group recommendation set. In this paper, we propose a conceptual architecture which uses the social trust consensus from users to improve the accuracy of the trust-based recommender systems. It is based on an existent model and integrates user's trust relations and item's factors into a generic latent factor model. One advantage of our model is the possibility to bias the users' similarity computation according to a trust consensus that assists in the formation of groups, such as the group of individuals who share the same content. The proposal represents the first steps towards the development of a group recommender system model. We provide an evaluation of our method with the Epinions dataset and compare our approach against other state-of-the-art techniques.