Discovering Web usage patterns for modeling learner's cognitive abilities

The study described in the present paper is part of a wider research that has developed an adaptive environment for Web-based learning - AdaptWeb/spl reg/. The paper focuses on the investigation of the learner's cognitive ability level (CAL) modeling based on the cognitive behavior during the learning process. It is based on high order thinking abilities according to Bloom's taxonomy. The investigation attempts to define the indicators of navigational behavior that could better discriminate the CAL classes defined. The learner's behavior assessment was performed by the analysis of the Web actions carried out in a free navigation content module. In order to accomplish our objective we have adopted an exploratory empirical and inductive research framework. We expect that the main benefit resulting from the adjustment of the instructional presentation to the learner's CAL characteristic is the possibility of designing pedagogical strategies and didactic tactics that provide the learner with an in-depth learning and more complex thinking.