A multi-agent architecture for learning paths-based personalized e-learning systems

Abstract

The application of artificial intelligence and semantic models for increasing learning and tutoring quality by personalization is an emerging research area. Мulti-agent frameworks facilitate the communication between the different components and ontological models can be used as knowledge sources for intelligent agents. In this research we analyse knowledge models and software architectures, outline trends in the personalized learning area and propose an agent-based architecture for e-learning systems that can conduct learning by generating and recommending personalized learning paths. Initial evaluation of prototype system is proposed and learning path generation scenarios are discussed.