Design And Implementation Of MOOC Course Hybrid Recommendation Method Based On Fuzzy Clustering

The MOOC platform for online education has developed faster and faster in recent years.The recommendation system based on the MOOC platform can not only help students match their learning needs with learning resources,but should also help students improve their learning efficiency and the utilization of learning resources.Therefore,under the application scenario of the MOOC platform,studying how to use the course information of high-quality courses and the personalized learning information of students to provide help for students’ independent learning has certain research significance and practical value.Facing the course recommendation needs of the MOOC platform,the main research work and results of this paper are as follows:(1)For MOOC data feature processing,the ML＿BPNN(Back propagation neural network with Medium logic)neural network is proposed,and the intermediate fuzzy theory is introduced to optimize the calculation method of the loss function.Using FCM(Fuzzy C-Means)algorithm,through ML＿BPNN neural network training,a hybrid score matrix containing three characteristics of student information,course grading information and course review information is obtained.(2)For similarity calculation,the PoiU-FHSM(Point wise User pair factored heterogeneous similarity model)hybrid similarity calculation model and calculation method are proposed,and the global similarity between students and students is introduced into the similarity calculation.Poi U-FHSM hybrid similarity is composed of predefined similarity and learning similarity,and learning similarity is obtained by BP neural network training based on the inner product of latent vectors.Experimental results show that the use of Poi U-FHSM hybrid similarity can improve the effect of course recommendation.(3)For course recommendation,a new hybrid recommendation method is proposed.Combine the PoiU-FHSM hybrid similarity calculation results with the course mapping relationship established in accordance with the courses arrangement plan of the syllabus to realize the simultaneous recommendation of Top K recommended courses and related pre-course and post-course courses.(4)Implemented and tested a MOOC course hybrid recommendation method based on fuzzy clustering,and visualized the recommendation results.