A Personalized Movie Recommendation System Based on Knowledge Graph

Abstract

In order to alleviate the problems of low accuracy, poor interpretability and data sparsity of movie recommendation system, this paper simultaneously considers movie attribute information and user-movie interaction information, and proposes a personalized movie recommendation algorithm based on knowledge graph, referred to as KGCFR (Knowledge Graph Collaborative Filtering Recommendation) algorithm. On the one hand, the movie knowledge graph is constructed by using the relationship between movies and movie attributes, and the relationship between movies is extracted. The user preference is calculated by the KGCFR model. On the other hand, the user preference is calculated by the collaborative filtering algorithm using the interaction information between movies and users. Finally, the recommendation list is obtained by combining the above two aspects, and the Top-K recommendation is performed according to the recommendation list. The experimental results show that the proposed method has a significant improvement in the accuracy of the recommendation effect, and has better interpretability and alleviates the data sparsity problem.