Multi-agent architecture for intelligent E-learning

After the initial explosion of E-learning with tremendous hope, a more realistic era is starting where E-learning is assessed and new trends are followed. The paper presents a multi-agent architecture allowing the system not only to tutor and manage the user learning process but also to discover, by learning, the student model and evaluate his progress. For those purposes, agents count on different machine learning techniques such as functional-link neural networks, case-based reasoning and fuzzy logic.