Trace-based computer supported cooperative work as support for learners group design

Group work, in certain circumstances, could encourage peer learning and provide the learners an opportunity to clarify and refine their knowledge. However, randomly grouping learners of different level of knowledge and activeness could decrease the effectiveness of a group. In this paper, we present a criterion for recommending the formation of groups of learners. This criterion is based on traces collected from interactive digital platforms. Traces are then processed with Bayes Classifier. We implemented this prototype using the MEMORAe approach.