An Intelligent Agent for tutoring activities in e-Learning

Giving that to design an efficient e-learning environment, in addition to contents and technological infrastructure, it is necessary to focus on tutoring strategies; in this paper the analysis focus on the use of intelligent agents to support e-tutor activities. The idea is to allow on line tutoring activities, flexible and available 24 hours 24, through the use of an Intelligent Tutor Agent that permit a natural language communication between the learner and the agent. In particular, it is described an open source software, still in its developing phase, called TutorBot, that can be integrated in any e-elarning platform. The learner, in his own natural language, “converse” with the agent to retrieve educational contents and to be helped, just as it happend within real context when we interact with human tutors. Another main aspect is the integration of TutorBot within an Adaptive Search Engine (ASE) that makes possible to search for the answers, to be given to learners, on external resources, for example Google. Thanks to these adaptive features, ASE allows to customize the contents according to the user profile. In order to asses the effectiveness of the intelligent agent it was carried on a study on a sample of university students aimed to detect the strong or weak points of the system and its overall satisfaction.