Which courses to choose? recommending courses to groups of students in online tutoring platforms

In the scope of this paper, we solve the course mismatch problem of recommending courses to a group of students on online tutoring platforms. Traditional recommendation methods are designed to simulate individual activities, but, they cannot meet all recommendation requirements for tasks with multiple participants. Thus, it is important to explore how to improve student participation by surveying background information. For this purpose, we develop a group recommendation model, -C2C(Course to Choose), considering the preferences of each user in the group. Specifically, we first obtain students’ free choice of classes formed by free enrolment based on the community detection algorithm. Then, we propose a single user rating method that takes into account the degree of student participation based on the amount of time spent watching each video. Finally, the service object of the group recommender system is further extended from a single user to a group member. The experimental results show that our method is more effective than all the baseline methods.