Trust-aware recommender systems are programs that make use of trust information and user personal data in social networks to provide personalized recommendations. In this paper we propose a model to improve the accuracy of trust-aware recommender systems using reliability measurements. We use these measurements to assess the reliability of a prediction and reconstruction the neighborhood of users. In reconstruction phase, we remove the users that reliability of them is less than a threshold value from trust networks of users. An empirical evaluation on Epinions dataset shows that the proposed model is the most effective in term of accuracy while preserving a good coverage.
