A Systematic Review on Student Failure Prediction

Bruno Veloso, Maria Araújo Barbosa, Hugo Faria, Francisco Supino Marcondes, Dalila Durães, Paulo Novais

Abstract: Today, students tend to drop out of school more easily. It is necessary to find out what causes students to have such school failure in order to try to help them succeed in their school life. For this purpose, it is necessary to acquire data about students, and the area of Educational Data Mining (EDM) appears. EDM aims to develop methods for exploring data recovered from educational environments, thus allowing us to try to understand and predict student success [1]. Early prediction of school failure may be cornerstone on the effort of avoiding it. This paper presents a systematic review of school failure prediction systems in students up to high school. The goal is identify the main methods developed and tested, as well as the algorithms used in this task. For that intent, six papers were identified in the SCOPUS repository as relevant for include in the review.