Tunneling model between adaptive e-learning and reputation-based social activities

We present the conceptual integration of two previously implemented prototypes. The result is twofold: the use of the student model in an adaptive e-learning course is extended to support choosing exercise activities, and the reputation and skills gained during social-collaborative activities is used to update the student model. LECOMPS is a web-based e-learning environment aiming at automating construction and delivery of adaptive learning paths. Individual student model is maintained during the taking of the personalized course, to support adaptivity. SOCIALX is a web-based system that supports social and cooperative aspects of e-learning, embedded in a reputation system. It provides social exchange through sharing and reuse of solutions to exercises, micro-forum discussions on exercises, and development of projects through "socio-collaborative group work". The proposed model designs a two-way communication between LECOMPS and SOCIALX, in such a way that 1) the student model, managed by the former system, can be used to determine which subset of the exercises is presently more suitable for the individual learner, and can therefore be proposed in the latter system; 2) the solution of exercises, and the associated learner's reputation gained in SOCIALX, can be used to update the student model maintained by LECOMPS. This two-way transfer of model parameters and structures, addresses the main issues, namely collaborative learning and support to proximal development, which make up the core of Vygotskij's educational theory. © 2010 IEEE.