Gradual Learners' Assessment in Massive Open Online Courses Based on ODALA Approach

This article proposes a gradual assessment process combined with an adequate learner modelling based on ODALA approach that can be an effective add-on for massive open online course (MOOC) platforms and engineering. The proposed learner modelling includes five dimensions: general information, disciplinary cognitive state, learning styles, preferences and behavior. This article focuses on the cognitive state dimension that is based on an assessment pyramid with four levels: closed-ended questions, half-open questions, open-ended questions and problem situations. The assessment pyramid is the backbone of the learning process since it needs a gradual progression with an adequate methodology where various computer aided or completely automated evaluation activities are proposed. The transition from a level to another is a conditional one since there are minimal threshold of disciplinary knowledge acquisition. An evaluation prototype was tested with the algorithmic discipline and developed so as to access the feasibility of our proposition.

The Combination of Bayesian Networks and Stereotypes to Initialize the Learner Model in Adaptive Educational Hypermedia Systems

This chapter aims to propose a new way to initialize a learner model in adaptive educational hypermedia systems. Learner modelling in adaptive systems contains several indicators. Even if there are several methods for initializing the learner model, they do not manage the side of uncertainty in the dynamic modeling of the learner. The main purpose of this chapter is the initialization of the learner model based on the combination of the Bayesian networks and the stereotypes method. In order to carry out a complete initialization of this model, the authors propose to use a combination of the stereotype method to process the content of the specific domain of information and the Bayesian networks to process the contents of the independent domain of information. The experiments and results presented in this work are arguments in favor of the hypothesis and can promote also reusing the modeling obtained through different systems and similar modeling situations.