Improved course recommendation algorithm based on collaborative filtering
Zheng Chen, Xueyue Liu, Li Shang
Abstract
As multidisciplinary educational interest increases, it is more and more important to support students course decision. This paper proposes a new novel recommended algorithm based on collaborative filtering for the course recommender to help student's decision. In this algorithm, the improved cosine similarity is used, according to the history of students' course selection records, and the better accuracy is obtained in the recommendation task, which meets the needs of users. In addition, both text vector and user behavior record are used to improve the calculation of course similarity. This paper evaluates 2022 students' 18457 records and 309 courses' real data. The experimental results show that the algorithm has a good effect on accuracy, recall rate and F1-score index.