Bayesian Inference of Students' Knowledge Based on a Potential Answer Tree Using STACK, A Mathematics e-Learning System
Abstract: We propose a procedure for inferring students' knowledge using the Bayesian network approach in the mathematics e-learning system STACK. If we appropriately construct a potential response tree (PRT) for STACK based on the knowledge structure and conditional probabilities among potential responses to questions - which are calculated based on a large amount of learning data - we can infer students' knowledge on the condition that a specific student's answer is obtained. This procedure is a well-known technique, and applying the Bayesian network to construct PRTs for STACK allows for the possibility of learning analytics through a mathematics e-learning system.
Presider: Rashid Khan, DCC- King Fahd University of Petroleum and Minerals, Dhahran Saudi Arabia