Abstract
The purpose of this article is to propose a learning optimization approach in a Massive Open Online Courses (MOOC) platform. To this end, we outline some research topics and we propose a multiagent system that represents human actors by intelligent agents. A coordinator agent is also implemented in order to assist the multicriteria decision-making process about the MOOC choice. This agent employs clustering algorithms and has an overview of the learning platform that makes it able to assist learners and to learn about its experiences and about other agents’ behaviour. Our aim is to increase the efficacy of knowledge acquisition by taking into account past experiences in order to design future online courses with better disengagement rates.