Defining a tutoring model is one of the most significant factors in order to get an effective e-learning course. Tutor’s activities are fundamental and, in case of courses with a large number of students, they must follow at least two different directions – which implies the coordination of two different professional figures:(1) the contents tutors;(2) the process tutors.In reference with the process tutor activities, the authors elaborated an experimental model in order to get an expert system, GUARDIAN TUTOR™ 1.0. The model is based on an artificial neural algorithm - GT Brain™ - able to learn from experience and define some management strategies to monitor the activities and intervene in the process.In the present paper we’ll examine the experimentation first results, while emphasizing next points:identification of the quality indicators structure: such indicators are the criteria to train the neuronal network;definition of the variables set, to determine the project evolution;description of the specific neuronal structure and the learning algorithm;definition of the network training activities;analysis and interpretation of the network training activities results;generalization of data and definition of tutoring activities management rules in the GT Proactive™ module.
(EN)
