F. J. Tapia; C. A. Lopez; M. J. Galan; E. Rubio
Description:
In this paper, a Bayesian-Network-based model isproposed to optimize the Global Adaptive e-LearningProcess (GAeLP). This model determines the type ofpersonalization required for a learner according to his orher real needs, in which we have considered both objectsand objectives of personalization. Furthermore, cause-andeffectrelations among these objects and objectives with thelearning phases, the learner, and the Intelligent TutorialSystem (ITS) are accomplished. These cause-and-effectrelations were coded into a Bayesian Network (BN), suchthat it involves the entire GAeLP. Four fundamental phasesthat have a direct effect in the learner’s learning process areconsidered: Learner’s previous knowledge Phase, Learner’sProgress Knowledge Phase, Learner’s /Teacher’s Aims andGoals Phase, and Navigation Preferences and ExperiencesPhase. The efficacy of the Bayesian networks is proventhrough the first phase, in which learners of differentknowledge area were select. The main results in this workare: causal relations among objects and objectives ofpersonalization, knowledge phases, learner and electronicsystem. Personalization profiles set and their probabilities inthe first phase were obtained to diagnose the type ofpersonalization of the learner.