Formalization and implementation of Eliminating and Optimizing Selection (EOS) approach

The main challenge of e-learning systems is to provide different courses to different students with different learning abilities. Such systems must also be efficient and adaptive. However, adaptivity can be accomplished by improving the ability to select dynamically an appropriate learning object for a specific learner. a framework for individualized learning object selection, called Eliminating and Optimizing Selection (EOS) was proposed by Liu and Greer (2004). In this paper, the EOS framework is further analyzed, implemented and experimented. As a result, a formalization for this framework has been suggested. The computational results of this approach have been compared to the selection results done by other specialists. Comparisons have shown its superiority in terms of producing more optimized selection of learning objects. Moreover, this approach has demonstrated its competitiveness in terms of the selected sequences of learning objects for different learners with different needs.