Recommender system based on user information

One of the most successful recommender system, collaborative filtering (CF) still has problems: sparsity deteriorating the accuracy of recommendation and scalability making it difficult to expand data smoothly. In particular, sparsity can reduce the accuracy of recommendation, causing a serious problem in terms of reliability. In this paper, in order to reduce sparsity and raise the accuracy of recommendation, we propose a method that combines an item-based CF with user-based CF using weight of user information. The proposed method computes user similarity on the basis of weight of user information and thereby makes a prediction, once non-rated items pre-filled in the user-item rating matrix in the item-based CF. The result of the experiment shows that the proposed method can improve the extreme sparsity of rating data, and provide better recommendation results than traditional collaborative filtering.