Recommendation System Based On Knowledge Graph

The objective of recommendation systems is to provide users with accurate project suggestions,addressing the information overload problem caused by massive amounts of data.These systems have been widely applied to websites such as movies,shopping,and news.However,traditional recommendation systems are limited by data sparsity and cold start issues,making it difficult to offer appropriate suggestions.Knowledge graphs can effectively express semantic relationships between entities,and their application to recommendation systems has become a research focus.This study constructs a knowledge graph for the movie domain and converts entities into vector representations through an improved knowledge representation learning framework.It combines the BRP neural collaborative filtering algorithm to more accurately recommend movies of interest to users.The main research content includes: firstly,describing the entities and relationships in the movie knowledge graph,expanding the attributes required for the knowledge graph based on the Movielens dataset,and determining the types of entities and relationships.Then,a graph database is established,and the entity and relationship triplet data are imported to construct the knowledge graph.Next,an improved Trans H-HC model is proposed based on the knowledge representation learning framework,using clustering algorithms to divide the entity set into multiple clusters,and performing negative triplet sampling in each cluster to obtain entity vector representations.Finally,the BRP algorithm is integrated into the neural collaborative filtering algorithm,combined with the vector representations extracted from knowledge representation learning,and the knowledge graph is embedded as auxiliary information in the recommendation system,thereby improving the accuracy and reliability of the recommendation system.In addition,this study also designs and implements a movie recommendation system based on the knowledge graph.Firstly,a requirement analysis is conducted,followed by a detailed design of the system architecture,functions,and database.Then,movie recommendation interfaces,movie detail interfaces,and movie management interfaces are implemented.Lastly,tests are carried out to ensure the stable operation of the system.This study provides theoretical support and practical guidance for the development of efficient and accurate movie recommendation systems.