Mineração de dados no Moodle: análise de prazos de entrega de atividades

Abstract

Virtual Learning Environments became common practice as a course tool for both distance and presence learning courses, as they support the communication among the parties involved. This study describes research carried out on data that were generated by the interaction with the Moodle VLE within an educational institution, with focus on the analysis of due dates and actual submission dates for assignments in the course environment. The objective of this study is to obtain relevant information about how course assignments are posted in the learning environment, to guide actions supporting the reduction of submissions after the due date or close to the deadline, and to propose a transparent and automatic approach to integrating KDD activities to the Moodle environment, where the data mining stage is restricted to the algorithms selected within this study and the results are presented in a simplified manner within the user interface in the Moodle environment. The study considers the time the assignment remained open for posting, the course to which the assignment was proposed and the actual time when the assignment was posted into the environment. It was carried out following the steps of the knowledge discovery process in databases, using the Weka tool. As a result from the KDD process performed in our database, the number of postings that were closer to the final expiry date were higher for assignments longer than 15 days, graduate courses tended to have longer assignments than undergraduate courses, and they also presented a higher number of postings after the due date or close to the expiry date of the assignments. In this context, shorter assignments are recommended, in order to increase postings soon after the opening of assignments and to enable teachers to obtain faster feedback from the learning process undergone by the student. That makes possible to take corrective actions in shorter time in order to avoid student failure or dismissal. The implementation of the KDD process within Moodle enables the experimentation by users in an automatic and simplified manner.