User behaviour-driven group formation through case-based reasoning and clustering

Group formation for collaborative learning activities is a complex and time consuming task. Different criteria have been proposed for grouping learners in computer-based systems, such as performance and social characteristics. User behaviour is, however, rarely considered when groups are formed. This paper proposes an approach based on user behaviour that complements the current research on group formation based on different criteria. For this purpose, we propose a synergetic approach based on case-based reasoning and clustering to form groups on the basis of user behaviour. Case based reasoning is used to model user behaviour, while clustering uses the output of the CBR mechanism as criteria for placing learners in relevant clusters. The proposed approach is illustrated using an exploratory learning environment for mathematical generalisation called eXpresser.