Integrating User Suitability and Course Matching Degree for Online Course Recommendation Method

Online learning has become more and more popular for its convenience without time and space restrictions. How to choose a suitable course from thousands of online courses is a great challenge for online learners, so online course recommendations have emerged. But the existing course recommendation system still faces two main problems: 1)Different users have different learning abilities and needs. Therefore, the suitability of different courses for the target user should be carefully considered；otherwise, it may recommend too difficult courses for users. 2)Existing methods usually ignore the collocation relationship between the recommended course and the courses that the user has learned, which may lead to inappropriate recommendation. To address the above problems, in this paper the user’s learning characteristics, types and their study suitability for different courses are analyzed firstly; meanwhile, it explores the common selection frequency of courses to study the collocation relationship between different courses. Based on the above two aspects, the user-suitability and course-matching aware course recommendation model (SMCR for short) is proposed. The results of comparative experiments on the Canvas Network(CN for short) dataset and the China university MOOC(MOOC for short) dataset show that this method can achieve higher recommendation accuracy. Moreover, the SMCR model can recommend courses that are suitable for users’ learning ability as well as matching with the courses they have learned.