Peer-reviewed

This paper aims to report an experience of using an innovative CSCL tool to support real, collaborative learning by discussion. We based the success of on-line collaborative discussion on two crucial aspects: first, by providing an adequate dialogue structure and move types that promote meaningful contributions and achieve more effective interaction; second, by extracting relevant knowledge in order to provide learners and tutors with efficient awareness, feedback, and monitoring as regards learners¿ performance and collaboration. To this end, we employ a sociolinguistic model that builds a structure of the discussion through generic exchange types whereas it identifies a sufficient set of move types that users can use to contribute effectively to an on-line asynchronous discussion activity. Then these move types are used as indicators to analyze learners¿ interactions and let them and the ir tutors be aware of their behaviour and performance as well as of the particular skills exhibited during interaction. The ultimate goal is to achieve a more effective support and assessment of the discussion process and to enhance learning. To validate this model, a real virtual learning environment is employed to support collaborative activities based on asynchronous discussion.