Generative Modelling and Classification of Students' e-Learning and e-Assessment Results

In this paper we propose an intelligent modelling of the students' knowledge collected from the e-Learning and e-Assessment processes of a particular course. The paper is focused on proposing a methodology for extracting the students' knowledge from the e-Learning activities, which we refer to as Profiling, then modifying it in compliance with their e-Assessment results and eventually, using it to model the probability distributions of the students Profiles that have passed and of those that have failed the course. The probability distributions of the students Profiles are then applied in the Bayes' theorem to perform binary classification analysis, i.e., to classify the students, pass or fail. The purpose of the proposed methodology is to simulate a real teacher, more precisely, to observe the activities of the particular student during the whole course in order to derive a decision of his or hers overall success.