Today, Massive Open Online Courses (MOOCs) have the potential to enable free online education on an enormous scale. However, a concern often raised about MOOCs is the consistently high drop-out rate of MOOC learners. Although many thousands of learners enroll on these courses, a very small proportion actually complete the course. This work is at the heart of this issue. It is interested in contributing on multi-agents systems and ontologies to describe the learning preferences and adapt educational resources to learner profile in MOOCs platforms. The primary aim of this work is to exploit the potential of multi-agents systems and ontologies to improve learners' engagement and motivation in MOOCs platforms and therefore reduce the drop-out rates. As part of the contribution of this work, the paper proposes a model of Multi-Agent System (MAS), based on ontologies for adapting the learning resources proposed to a learner in a MOOCs platform according to his learning preferences. To model an adequate online course, the determination of learner's preferences is done through the analysis of learner behavior relying on his indicator MBTI (Myers Briggs Type Indicator). The proposed model integrates the main functionalities of an intelligent tutoring system: profiling, updating of the profile, selection, adaptation and presentation of adequate resources. The architecture of the proposed system is composed on two main agents, four ontologies and a set of modules implemented.
展开▼
