Pinpointing Learning Moments; A finer grain P(J) model

Educational data mining and knowledge engineering methods have led to increasingly precise models of students’ knowledge as they use intelligent tutoring systems. The first stage in this progression was the development of Bayes Nets and Bayesian frameworks that could infer the probability that a student knew a specific skill at a specific time from their pattern of correct responses and non-correct responses (e.g. errors and hint requests) up until that time [cf. 2, 4, 5].