Data mining and statistical analysis of educational data

Abstract

The advancement of technologies related with the internet has enabled E-Learning to gain popularity as a way of transmitting knowledge. Universities and Companies, among others institutions, have been using E-Learning to disseminate educational content to remote locations, reaching out students, researchers and employees who are physically distant. The Moodle platform is an example of a Learning Management Systems (LMS). LMS provide on-line platforms where teachers and trainers can publish contents organized into activities, conduct assessments, and other tasks so that the students involved can learn and be assessed. In addition, LMS generate and store large amounts of data, named Educational Data, not only from the users activities but also functional data. This work presents data mining techniques applied to Educational Data. From the Moodle data repository of the University of E´vora, we apply supervised learning techniques with the aim of predict the students success from their interaction with Moodle. We will also show interesting conclusions when unsupervised learning techniques are applied.