Measurement centred evaluation in adaptive web-based learning systems based on CBR approach

Several Web-based learning systems are currently available. Nevertheless, most existing e-learning platforms lack efficient adaptivity to learner needs. Moreover, adaptive system development is quite costly. Evaluation of these systems and assessment of adaptation components may enable identifying failures and improving them, in one hand, and exchanging successful adaptive components, in the other hand. Adaptive behaviour is strongly determined by functional requirements available to the application domain. In Web-based system, granular functions are increasingly provided by Web services. First, in this paper we focus on measurement criteria related to both adaptive functions and related Web services. Second, Case-Based Reasoning are adopted in order to capitalize evaluation results, learn from past experiences and reuse successful patterns.