Research on collaborative filtering recommendation of learning resource based on knowledge association

Abstract

With the development of Internet communication equipment, educational resources are rapidly accumulating on the Internet. While learners enjoy the convenience of the information age, they often face problems such as “resource disorientation” and “learning theme drift”. As the most effective personalized recommendation technology, collaborative filtering is mainly based on the dual relationship between learners and resources. This paper utilizes the association of learners, knowledge points, and learning resources to construct associated matrix of the learners, knowledge points and learning resources. It introduces the related knowledge information into user similarity calculation and scoring prediction of traditional collaborative filtering algorithm, which can make the recommendation results conform to learners’ learning needs and improve the recommendation quality of personalized recommendation.