A Tensor Decomposition Recommendation Method Using Multiple Contextual Information

In view of the weak semantics, high sparsity and cold start problems of user check-in data, many researches have been conducted to improve the performance of point-of-interest recommendation by combining context information. However, there are some problems such as low utilization of context information, difficulty in balancing the richness of context information and computational complexity. Therefore, this paper proposes a recommendation method based on tensor decomposition that integrates multiple contextual information. This method first gathers similar users into the same set, and then integrates time, location, and social information to build a tensor model. Experiments demonstrate that this method can overcome the above problems to a certain extent by considering multiple contexts.