An emotion-based model for improving students' engagement using agent-based social simulator

Abstract

Students' emotional engagement in classrooms has always been regarded as a crucial factor influencing educational outcomes. Emotionally engaged students are highly motivated to attend classes and participate actively in discussions. Conversely, emotionally disengaged students are habitual truants and/or often misbehave in classrooms. It is time-consuming to conduct research using traditional methods such as data collection or questionnaire. Furthermore, as these experiments are conducted on humans, limited settings can be organized and difficult to control/monitor. Therefore, in this paper, we present our work-in-progress to address the problem. We propose a conceptual emotion-based model that is animated by an agent-based social simulator. We also propose an algorithmic approach that decides whether to maintain or update a proposed action plan for improving students' engagement in a classroom. The rudimentary theory of this proposal is based on three main elements: the engagement level of students, the emotional state of a lecturer and the emotional state of students in a classroom.