Towards a new approach to reduce the dropout rate in e-learning

MOOCs are currently facing a major problem, namely the decrease in the rate of certified at the end of a MOOC. The main cause of this problem is due to the difficulty of detecting user behaviors in the platform which does not help tutors in their follow-up and their educational supervision of learners as well as to interact easily with them. In this work, to solve this problem we will propose an information system capable of automatically analyzing and detecting the behavior of users who follow a distance training and subsequently, to classify their profiles according to predefined parameters to facilitate teachers choose courses and activities based on the behavior of each user.