Design and Empirical Research of Multi-role Intelligent Pedagogical Agent Oriented to Pre-class Learning

Intelligent pedagogical agent is an important way to improve the effectiveness of online learning, and further research is needed on how to build multi-role agents for pre-class learning and the differences between IPA with different roles. To this end, the article constructs a multi-role intelligent pedagogical agent model based on the Community of Inquiry, and further designs agent intervention strategies for both tutor and peer roles in pre-class learning. In order to verify the effectiveness of agents and explore the performance differences of agents in pre-class learning, a quasi-experimental study was conducted on 34 students who took " Database Principle " as an elective course in a university. The results show that it can improve students ' behavioral participation and learning performance. However, intelligent peer agent will reduce the completion rate of students ' pre-class collaborative learning activities.