Contextual adaptation of learning resources

Pervasive learning systems must define new mechanisms to deliver the right resource, at the right time, at the right place to the right learner. This means that rich context information has to be considered: time, place, user knowledge, user activity, user environment and device capacity. This paper presents our approach which involves modeling of context and of situation. As context is based on numerous information which may change frequently (coming from a collection of captors), a more aggregate view is defined to work on more abstract objects: the situation. Context information and situation information have to be widespread into all the models of learning systems: context preferences have to be handled in the learner model, well-adapted situation and situation have to be in the learning resource model. The process of delivering and of adaptation is also enhanced by introducing the comparison between situations. A prototype validates this proposition.