A trust-based architectural framework for collaborative filtering recommender system

Recommender systems have been used to suggest the interesting items such as movies, books and songs according to the choice of users. These systems compute a user similarity among users and use it as a weight for the users' ratings. However, they have many weaknesses, such as sparseness, cold start and vulnerability to attacks. The traditional recommender system techniques are often ineffective and are not able to compute a user similarity weight for many of the users. The trust among two or more users in the web of trust increases the quality of recommendation in two ways. Firstly, the trust metrics reduce the computability of similarity assessment of users or items. Secondly, the reputation of users may be computed using trust propagation. In this paper, architecture of trust-based recommender systems is proposed. In which trust metrics and rating matrix are taken as input and neighbours are generated using trust metrics and user similarity respectively and importance of trust over collaborative filtering is described. In the proposed approach, trust-based issues are discussed to solve the problem of traditional recommender system such as, data sparsity, cold-start users, malicious attacks on recommender systems and centralised architectures.