Learners Monitoring Based on Traces in CSCL System

Abstract

In this paper, we propose to use traces in order to facilitate the process of learners monitoring in collaborative learning context, where the interactions between the human actors are very voluminous and very heterogeneous. These traces concern the entire pedagogical activities of the actors as well as the use of the system resources. Hence, first of all, these traces must be collected and filtered. Then, they must be analyzed to assist or support the tutors in their tasks. For facilitating their use, these traces are distributed into two main classes, which depend on the pedagogical activities (learning, assessment and collaboration) and on the system use. The collected traces are used by a collaborative learning system called SYCATA (SYstème pour la Collecte et lAnalyse des Traces dApprentissage collaboratif). For facilitating the learners monitoring process, a set of tools are offered to tutors (trace visualization, advice management, etc.). Furthermore, SYCATA has an assistant system of the tutor, which is based on the collected traces and a set of indicators calculated from theses traces. Our system was implemented and experimented with a sample of university students where good results have been obtained.