Dynamic Bayesian Belief Network for Modeling Student Knowledge Tracing

E-learning offers an important feature to conventional learning by providing optimized learning. For this, accurate representation of students current skills and adaptation to newly acquired skills are essential. In this, we show that the development of skills in children for those who are having difficulties in learning arithmetic skills is possible through graphical models by evaluation of data gathered from different children by conducting tests. The model consists of Dynamic Bayesian Network which incorporates domain knowledge and identifies the difficulties. The system posts appropriate tasks and exercise actions on the basis of estimated levels of gathered knowledge. The concepts that are believed to have been learned or not represent the evidence. Based on the evidence it is concluded that which concepts to be relearned and which not.