A survey for trust-aware recommender systems: A deep learning perspective

A significant remaining challenge for existing recommender systems is that users may not trust either inaccurate recommendation or lack of explanation. Thus, it becomes critical to embrace a trustworthy system. This survey provides systematic summary three categories issues in systems: social-aware systems, which leverage users’ social relationships; robust filter untruthful information, noises and enhance attack resistance; explainable provide explanations the recommended items. We focus on work based deep learning techniques, an emerging area research.