Educational Data Mining: A Profile Analysis Of Brazilian Students

This paper presents an analysis of data referring to the profile of Brazilian students in the year 2016, according to the Higher Education Census. The information provided by this census is used to carry out various analyses related to the current situation of Brazilian education, as well as the profile of students and institutions. In this work, we analyze the modality of courses offered by educational institutions, investigating the profile and the quantitative of students who have scholarship, and how they enter the courses. We also investigate the distribution of students according to the informed skin color/race. The method used to conduct this research involved the application of the steps proposed by the CRISP-DM Reference Model and the use of the Apriori association algorithm to identify and analyze the data set. The results show that 30% of students choose not to report their skin color; Of those who reported, 37.4% are white and 6% are black. Approximately 82% of the students entered higher education through the traditional method (entrance examination), and 82.3% of the male and 77.2% of the female opted for face-to-face courses.