Classification and predictive analysis of educational data to improve the quality of distance learning courses

Educational Data Mining (EDM) is the research area concerned with the use of data mining methods and its development in order to explore data sets collected in educational platforms. EDM has a considerable potential to improve the education quality. In this paper, we first model the Moodle variables of the Brazilian National School of Public Administration (Enap) database. Next we apply EDM to understand the standard Moodle variables influence and the proposed variables in the attendees performance and dropout rates. Our framework includes the following data mining approaches: decision tree, Adaboost based decision tree, random forest and k-means. The obtained Adaboost based decision tree outperforms the other approaches and the state-of-the-art framework with 91 % of Precision, 89 % of Recall and 90 % of F1 score in terms of the attendees performance evaluation.