Análise e predição de desempenho de alunos da graduação a partir da inserção de material de apoio em ambientes virtuais de aprendizagem

Abstract

Virtual learning environments (VLEs) are increasingly being used as they enable education to reach more and more people. Its use is not limited only to distance education, but also assists in face-to-face education as a more advanced form of learning. However, managing the learning process in VLEs with quality integration and monitoring requires more and more of the teacher who needs to use updated materials and at the same time, monitor and measure students' learning through assessments. In this regard, this work presents a predictive model of data mining in a VLE, from the insertion of didactic material by the teachers in the classroom teaching. The objective was to verify the impact that the non-insertion of material has on the performance of classroom students using VLE as an extension of the classroom. Toward this, experiments were carried out with different data sets, where data mining techniques Bayesian networks and Decision Tree were applied, and the performance of each one was compared in order to obtain a more accurate model. As falows, it was possible to demonstrate, through the analysis of the data generated, that the insertion of didactic material in the VLE contributes to the better performance of the undergraduate students using VLE as an extension of the classroom.