Research On Movie Recommendation System Based On Hybrid Algorithm

The purpose of recommendation algorithms is to provide users with personalized and accurate recommendations,so that users can more conveniently and quickly obtain information or products that they are interested in.However,traditional recommendation algorithms such as collaborative filtering mainly rely on the historical behavior of users to predict the items that users may be interested in,so there will be a problem of performance degradation when facing new users or new projects.The knowledge graph has rich semantic information and complex entity relationships,which can provide a more comprehensive and in-depth description of users and items for the recommendation system.It can more accurately describe the characteristics and styles of items,analyze the suitable crowd,and improve the accuracy of recommendation.The thesis obtains the required data from two datasets,MovieLens and IMDb,and after data cleaning and processing,identifies and extracts entities and relationships to construct a collaborative knowledge graph in the film field.On this basis,proposing a movie recommendation algorithm combining knowledge graph and collaborative filterin.In the knowledge graph,the algorithm models the high-order connectivity of the graph using graph attention networks,recursively propagates information from neighbors,and distinguishes the importance of neighbors through attention mechanisms;The algorithm adds a time term function on the basis of the collaborative filtering algorithm to simulate the decline of user interest.Compared with other algorithms,the feasibility and progressiveness of the algorithm are verified.Finally,the thesis designs and develops a personalized movie recommendation system.Through front-end and back-end design and database table design,the system can provide personalized and accurate movie recommendation services for users based on their historical behavior and personal information,combined with recommendation algorithms and knowledge graph.