User credit-based collaborative filtering

Memory-based collaborative filtering is the state-of-the-art method in recommender systems and has proven to be successful in various applications. In this paper we develop novel memory-based methods that incorporate the level of a user credit instead of using similarity between users. The user credit is the degree of one’s rating reliability that measures how adherently the user rates items as others do. Preliminary simulation results show that the proposed methods outperform the conventional memory-based ones. The methods are effective in a cold-starting problem.