The discussion process plays an important social task in Computer-Supported Collaborative Learning (CSCL) where participants can discuss about the activity being performed, collaborate with each other through the exchange of ideas that may arise, propose new resolution mechanisms, and justify and refine their own contributions, and as a result acquire new knowledge. Indeed, learning by discussion when applied to collaborative learning scenarios can provide significant benefits for students in collaborative learning, and in education in general. As a result, current educational organizations incorporate in-class online discussions into web-based courses as part of the very rationale of their pedagogical models. However, online discussions as collaborative learning activities are usually greatly participated and contributed, which makes the monitoring and assessment tasks time-consuming, tedious and error-prone. Specially hard if not impossible by humans is to manually deal with the sequences of hundreds of contributions making up the discussion threads and the relations between these contributions. As a result, current assessment in online discussions restricts to offer evaluation results of the content quality of contributions after the completion of the collaborative learning task and neglects the essential issue of constantly assessing the knowledge building as a whole while it is still being generated. In this paper, we propose a multidimensional model based on data analysis from online collaborative discussion interaction that provides a first step towards an automatic assessment in (almost) real time. The context of this study is a real on-line discussion experience that took place at the Open University of Catalonia

Peer Reviewed