An assistant agent for group formation in CSCL based on student learning styles

Abstract

In Computer Supported Collaborative Learning (CSCL) systems, students work in groups interacting by using computers. Each member of the team behaves in a particular way to collaborate with others, manifesting a particular learning style. In this paper we propose a new approach for automatically creating student groups in CSCL systems by considering their individual learning styles. Data mining techniques are applied to discover which existent combinations of learning styles lead to a better performance. The discovered knowledge will be used by a software agent to propose the creation of the most promising new groups. The approach also considers the creation and maintenance of a user model for each student and a group model for each team. The assistant agent will be implemented in an existing CSCL tool, and its performance will be validated with real students.