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

► Modelling user behaviour in an exploratory learning environment using casebased reasoning (CBR). ► A synergistic approach of CBR and array-based clustering is used for grouping for collaboration. ► Heterogeneous metrics and resemblance coefficients measure similarity between user behaviours. ► User behaviour is more informative than the typical single performance indicator. ► Homogeneous clusters are formed based on user behaviour; heterogeneous groups can be derived.