Using Data Mining Techniques to Follow Students Trajectories in Secondary Schools of Uruguay

Abstract

It is possible to observe an enormous increase on the number of researches focused on automatically find patterns and factors that affect students behavior and performance during their learning process. The fields of Learning Analytics and Educational Data Mining are in constant growing, developing new and innovative tools. Furthermore, new methodologies are being created to follow and help students and professors inside the many different types of educational settings. At the same time, it is also possible to see that the majority of the existing works are still restricted to small and controlled experiments, conducted on samples of students data. The present work describes the first step of an international collaboration focused on implementing Learning Analytics on a national scale. Precisely, this work describes the methodology applied to find rules that can be used to follow students' trajectories in secondary schools in Uruguay. The results points out for the possibility of delivering rules by analyzing patterns of students clusters based on their success (or failure) in the school year. Among other findings, this work shows a strong relationship between students grades and their number of absences in the classes.