Interaction Analysis of the Virtual Learning Environment to Predict Evasion in Distance Learning Courses

• Autor(a) principal: Sonnenstrahl, Thiago Siqueira
• Data de Publicação: 2021
• Outros Autores: Bernardi, Giliane, Pertile, Solange
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: EAD em Foco
• Texto Completo: https://eademfoco.cecierj.edu.br/index.php/Revista/article/view/1463

This article aims, through Educational Data Mining (MDE), to analyze, through the interaction of students in the Virtual Learning Environment (AVA), possible dropouts in distance learning courses, providing strategic data for decision making by the institution's educational managers. In order to carry out the experiments, two sets of data were used containing the interactions of students in the AVA Moodle from two subsequent classes in distance learning mode. As a result, the hit rate for the first set of data was 93%, obtained with the Randon Forest algorithm, while for the second set, the hit rate was 85% with the model generated by the J48 algorithm. The mining results dispelled the visualized_task and the visualized_material as evasion indicators. Thus, it can be seen that students who have little interaction with the resources within the environment and do not see the materials and tasks available are more likely to drop out of the course. Keywords: Distance education. Evasion. Educational data mining.