Error Driven Adaptive Training Sequences

Teachers want to remedy the problem of heterogeneous levels of students before teaching a new concept. To deeply check a competency, we propose a system generating error driven adaptive training sequences. A test is created thanks to a bank of exercises. At each failure, the error detection mechanism is called and each detected misconception is taken into account using corrective exercises associated with explanations. Thus, each student´s path in the training sequence is different, adapted to his/her knowledge and tries to remedy his/her misconceptions.