Research And Implementation Of Online Course Recommendation Based On Multi-feature Ranking Model

With the rapid development of Massive Open Online Courses,online education,as a new form of study,has been accepted by more and more people.Users can learn knowledge and skills from different domains through the Internet.But as the amount of online resources grows fast and the courses become more diverse,users will encounter selection problem when deciding which courses to learn.The use of recommendation algorithm can give advices for users' learning options.However,online course has its own limitation,such as the lack of text information,user behavior information and comments.Traditional recommendation algorithms cannot be applied directly to online courses.We need to innovate and improve the algorithm based on the unique scenario of online courses.In our proposal,after sufficiently analyzing the data of Cloud Class,we research and implement an online course recommendation algorithm based on multi-feature ranking model.The algorithm considers several features related with online courses and users,including the preference based on topic,the preference based on collaborative filtering,the popularity of courses and the influence of teacher.The algorithm combines the features with linear function by using the method of learning to rank.Then it calculates the matching degree of target user and online courses to make recommendation to the user.In order to verify the effectiveness of the proposed algorithm,we have done many experiments on the dataset of Cloud Class.The experiments prove that our algorithm can get great recommending result and outperform the compared algorithms.Furthermore,we design and implement the recommender system based on Cloud Class,which is mainly used on the personal learning page of users.The system works well and demonstrates the practicability of our algorithm.