A Unified Probabilistic Matrix Factorization Recommendation Fusing Dynamic Tag

Abstract

In existing social tagging systems, user interests changing is not considered. In order to solve this problem, in this paper we propose a unified probabilistic matrix factorization (TTUPMF) recommendation algorithm which combines social tags and time factor. In the proposed approach, A user-item rating matrix, a user-tag tagging matrix, an item-tag correlation matrix and a unified probabilistic matrix factorization are constructed to obtain the latent feature vectors of three matrices to be recommended to users by optimizing the training parameters. The experimental results demonstrate that the proposed model uses the tags semantics effectively and improves the recommendation quality.