Work in Progress - A Distributed Approach to a Learning Management System using Multi-Agent Technology

The agent technology has in the autonomy of a single agent one of its most important features. The collaboration between agents, inference and learning capabilities are other interesting features. All these issues seem to be useful to represent students in an electronic system, because they permit the creation of virtual tutors that could follow the steps and the evolution of the students. It is a way to improve the interaction between students, teachers and professors. The agent technology could also be applied to create distributed Learning Management Systems (LMS). With some artificial intelligence methods it would be possible to build a distributed Intelligent Tutoring System (ITS). This paper presents two possible architectures of distributed LMS: one using a LMS to store the knowledge (such as a domain module of a ITS) and agents representing the tutor and student modules of an ITS. In the second approach, there is no LMS and all ITS functions are distributed between agents. This paper discusses the benefits of these approaches and how agents can provide help for the students in these architectures