Research On Collaborative Filtering Algorithm Based On MOOC For Bipartite Graph Context Information

With the rapid development of Internet information technology applications,the way people are educated has changed,and the MOOC platform has become a new way for people to acquire knowledge.However,the quality of curriculum resources on the MOOC platform is uneven,and the phenomenon of redundancy of the same kind of resources has caused the high dropout rate and low completion rate of the MOOC platform.Currently,the recommended systems often have cold start problems due to the lack of historical data of new users,resulting in no recommended results.Based on these problems,the paper studies the personalized recommendation algorithm based on MOOC platform.For the cold start problem,it is proposed to solve the data sparse problem by using the bipartite graph network structure.In order to improve the accuracy of the recommendation algorithm,a collaborative filtering recommendation algorithm based on bipartite graph context information is proposed,The algorithm classifies all users in the original data set by using user context information;The user-based collaborative filtering recommendation algorithm is used to calculate the similarity between the target user and the similar user;Then,the nearest neighbor of the target user is obtained by using the similarity value of the similar user,and the neighbor user-resource score data is obtained from the original data set according to the nearest neighbor to implement data preprocessing;Obtaining a neighbor user node set,a resource node set,a user and resource bipartite graph network structure,and an interest matrix according to neighbor user-resource scoring data;Using the quantitative value calculation method based on the dynamic change of the target user's hobbies from the resource,the user and resource linkage quantitative value d is calculated according to the user and resource bipartite graph network structure and interest matrix;According to the size of the quantized value d,the TOP-N recommendation algorithm is used to obtain the resource recommendation set.Based on the Movielens data set and the MOOC platform data set,the paper conducts a comparison experiment based on the user collaborative filtering recommendation algorithm,the graph recommendation algorithm,the collaborative filtering recommendation algorithm based on the bipartite graph context information,and the algorithm running efficiency analysis.The results show that based on the bipartite graph context information collaborative filtering recommendation algorithm is the best.