Due to the complexity of the structure of knowledge points in online course videos, the relative difficulty coefficient of knowledge points will be different for learners. The recommendation algorithms came into being, and the collaborative filtering algorithms are the most widely used recommendation algorithms. However, the traditional collaborative filtering recommendation algorithms have limitations in the similarity calculation between learners. Therefore, we propose a new knowledge point

more »

... commendation algorithm. The algorithm fully considers the differences in learners' relative difficulty coefficient of knowledge points in different dimensions. Besides, we define the similarity between learners at an overall level by incorporating information from knowledge points learned by non-associated learners. We comprehensively consider different aspects to improve the performance and quality of the recommendation algorithm. Comparison experiments on the real dataset demonstrate that our proposed algorithm improves the overall performance of the recommended algorithm.