Movie Big Data Intelligent Recommendation System Based on Knowledge Graph

Abstract

The user-oriented recommendation system enables users to obtain their favorite movie programs from a large number of movies. Knowledge graph is a kind of heterogeneous network information, which provides rich structural knowledge for movie recommendation system. It helps to solve the problem of data sparseness and cold start. Based on this, we designed a movie recommendation model based on the knowledge graph. First, we embed the knowledge graph into the recommendation model, then use the interactive attention network represented by the user/item to update the vector representation, obtain rich neighborhood knowledge through multi-layer interactive information dissemination, and finally make predictions. We test on public data sets, and experiments show that our method is better than existing baseline methods.