Personalized Online Learning Resource Recommendation Based on Artificial Intelligence and Educational Psychology

The objective of the study is to explore an effective way for providing students with the appropriate learning resources in the remote education scenario. Artificial intelligence (AI) technology and educational psychology theory are applied for designing a personalized online learning resource recommendation scheme to improve students' learning outcomes. First, according to educational psychology, students' learning ability can be obtained by analyzing their learning behaviors. Their identities can be classified into three main groups. Then, features of learning resources such as difficulty degree are extracted, and a LinUCB-based learning resource recommendation algorithm is proposed. In this algorithm, a personalized exploration coefficient is carefully constructed according to student's ability and attention scores. It can adaptively adjust the ratio of exploration and exploitation during recommendation. Finally, experiments are conducted for evaluating the superior performance of the proposed scheme. The experimental results show that the proposed recommendation scheme can find appropriate learning resources which will match the student's ability and satisfy the student's personalized demands. Meanwhile, by comparing with existing state-of-the-art recommendation schemes, the proposed scheme can achieve accurate recommendations, so as to provide students with the most suitable online learning resources and reduce the risk brought by exploration. Therefore, the proposed scheme can not only control the difficulty degree of learning resources within the student's ability but also encourage their potential by providing suitable learning resources.