A method for recommending multi-interest learning resources based on capsule networks and learning styles

With the integration of Internet technology and the education field, online education has been developing rapidly. Learners tend to use online education platforms as a fast and efficient way to acquire knowledge. Nevertheless, it is not easy for users to get the learning resources they are interested in from massive data. Introducing recommendation algorithms commonly used in the e-commerce field into e-learning platforms is an efficient way. However, the current recommendation systems are not sensitive enough to the multiple-interest extraction of learners, which leads to the ineffectiveness of recommended learning resources. To solve this problem, this paper proposes a learning resource recommendation method based on capsule networks and learning styles for learners’ multiple interests. The Felder- Silverman learning style quantified table is introduced to model users’ learning style preferences. Experiments prove that the method proposed in this paper is more efficient in real datasets compared with existing recommended methods.