Modeling Learner's Emotions with PAD

摘要

Emotions have a direct influence on an individual's physical and cognitive behaviour, as well as their performance, a student with a positive emotional state will learn and perform better. This paper presents an agent framework that addresses the relationship between user's state of emotion during learning and the modification of learning pace and feedback-type in a virtual environment for learning effectiveness. The technique exploits the structure of emotion-evaluation from user's current interaction to dynamically regulate learning pace within a Virtual Learning Enviornment (VLE) via a proposed Emotion Regulation Agent (ERA) system to derive meaningful emotion information from user's interactions based on a simplified version of the Pleasure, Arousal and Dominance (PAD) emotion model. This information then regulates learning by adjusting the type of task and learning information to promote a beneficial and sustainable learning experience by optimizing learning emotion.