Analysis of ENEM’s attendants between 2012 and 2017 using a clustering approach

Data analysis is increasingly being used as an unbiased and accurate way to evaluate many aspects of society and their evolution over the years. This article presents an analysis of student’s characteristics, between 2012 and 2017, in the most important exam for entry into higher education in Brazil, the Exame Nacional do Ensino Médio (ENEM). The intention is to gain insights of Brazilian regions, ENEM’s areas of knowledge, type of school and accessibility, using a clustering method (K-means). An extensive and careful cleaning of the database was made in order to homogenize it and avoid types of statistical bias. The results of this work are presented objectively in the article, so it may be useful and used as a numerical base in works of socio-educational disciplines or studies that are interested in better understanding the evolution of ENEM in recent years. Finally, some discussions and restrictions on grouping results were presented in a timely manner.