M-learning activity-based context management

Abstract

It is usually difficult to formally define and efficiently reason with the various contextual elements of the learner’s context space. The latter is multi-dimensional, heterogeneous and of complex structure. It is also not well known how to use context to predict the dynamic learner behavior. In this paper an attempt is made to solve some of these challenges. In particular, the paper contribution is a context-aware model capable of recognizing and coordinating learner’s goals and activities in order to map these activities onto available learning services and resources. This is achieved through a structured and flexible ontology formulation of context, based on context fusion and cooperation, and driven by the learner’s goals and activities. The proposed context reasoning and recognition process relies on three ontologies where semantic structures, often crossing semantic levels, are used by a semantic matchmaker that can perform reasoning for context-oriented service discovery and access.