Educational processes’ guidance based on evolving context prediction in intelligent tutoring systems

Abstract

Considerable attention has been paid to content adaptation in ITS. However, process-oriented adaptation has been neglected and none of ITS addressed the correlation between the learning and the teaching process. Indeed, uncertainty coming from the dynamic preferences of learners and tutors, the evolutionary cognitive state of the learner and the choice of goals and educational strategies is not well considered in ITS. So, in this paper, a new proposal for guiding and adapting the construction of both learning and pedagogical processes named educational processes in intelligent tutoring system is described and detailed. The research strategy followed is based on a dynamic Bayesian network to predict the educational context, a strategic perspective for modeling educational processes, and guidance algorithms for their development and support. This aims to generate an individualized learning process for each learner by selecting the most appropriate pedagogical process according to the actual preferences of the tutor. Experimental results are given to illustrate the applicability of the proposed solution. The results show an improvement in guiding the learner and the teacher/tutor by considering the evolving and uncertain educational context. Learners will be therefore able to achieve the learning goal more efficiently when the pedagogical process is more adapted to their individual differences. The evaluation of knowledge improvement, the appropriateness of educational recommended context and the prediction effectiveness shows promising results.