TR#546: External Representation of Learning Process and Domain Knowledge: Affective State as a Determinate of its Structure and Function

Barry Kort, Rob Reilly, Rosalind W. Picard

Published in the Proceedings of Artificial Intelligence in Education Workshops (AI-ED 2001), San Antonio, Texas, May 2001


We present a model of the generic learning process and associated metacognitive processes that aid efficient learning. These models can be used as internal representations of a learners cognitiveemotive state while engaged in learning. They can also be used to present to the learner a representation of their progress in learning much like a coach or mentor might use to assist a student. These models and associated representations might be displayed alongside the primary external representation of the subject being learned or they might be integrated into it. The assessment of the metacognitive processes which shadow the primary learning process needs to be sensitive to the affective state of the learner. As the student proceeds through the learning journey, their affective state may cycle through a wide spectrum of emotions, which influence how efficiently, effectively, and enjoyably they succeed in the learning task. Based upon an understanding and application of our proposed model, the structure and function of external representations would reflect the learners cognitive-emotive state. For example, some representations would provide copious hints that would accelerate the learning at the expense of the playfulness of the exercise, while other representations might transform the le arning experience into an enthralling game. In particular these models enable the system designer to provide alternative intervention strategies for the learner who is laboring under a misconception, ranging from a no-nonsense remedial intervention to allowing the learner to play out their misconceptions in a free-wheeling simulation model that ultimately reveals the folly of their thinking.

PDF . Full list of tech reports