Análise e predição de evasão dos alunos do curso de Sistemas de Informação da Universidade Federal de Santa Maria campus Frederico Westphalen por meio da mineração de dados educacionais

Abstract

The high school dropout rates are a reality present in several higher education courses offered in Brazil, which highlights the need to investigate this issue, which is responsible for economic losses in institutions and impacts on the global education scenario. This paper aims to develop a study on school dropout patterns in higher education, based on the analysis of data provided by the Information Systems course at the Federal University of Santa Maria (UFSM). These data went through a data processing system, in order to point out indicators related to factors that classify possible evasions. Results were generated through decision trees, which pointed to data regarding the student and his academic performance as important factors for school dropout in higher education.