Modeling learning group's communication in Intelligent Tutoring Systems

Collaborative learning is a method which can facilitate gaining knowledge. More than 20 years of academic research points that learning in groups is more effective than learning individually [1]. Many researches indicate that suitable strategy for creating collaborative learning groups is very important for group's efficiency. In this paper we analyze our original algorithm for creating heterogeneous groups [2] to assert a communication between their members. For this task Social Network Analysis is used. Such analysis delivers very interesting information about relationships in groups and allows to select less active group mates.