Blended learning has generated remarkable interest in educational research. This interest is being concentrated increasingly throughout the Covid-19’s duration. Indeed, this teaching method combines both, presential and distance learning. This teaching method’s philosophy is quite simple. it aims to exploit the qualities of the mentioned two teaching modes. However, the usage of Blended learning isn't sufficiently advanced as it causes genuine issues in the educational scripting of the instructive substance. In this research, it is an aim to concretize this philosophy in a fairly complicated course, as well as an advanced level: Python programming language for students of preparatory classes. To achieve this goal, a developed a personalized course model using Bayesian networks capable of helping educators in their teaching practices in addition to their students’ cognitive processes is presented, as well as actualizing this model in a course creation application. Moreover, the Python programming was tested with various understudies with various degree of aptitude in the subject educated. The results presented in this paper, shows the effectivity of this approach in teaching complex course especially during the COVID19 period, based on both the satisfaction rate and the success rate metrics that was chosen to asset the following approach.