Avaliação da estrutura do currículo do ensino superior com aprendizado de máquina

Abstract

The curricular structure of a bachelor degree is a very important matter for education as a whole, and it’s development usually revolves around qualitative factors. In the present work, we’ll use the synthetic control method, alongside the data from the students of Computer Science from Universidade Federal do Ceará, to provide a quantitative framework to evaluate such structure The curricular structure of a bachelor degree is of utmost importance for higher education, as it defines not only the contents of the degree itself, but also the chronological order of under which the students will study this content. A poorly built structure might compromise the degree, or even increase the drop out rates of the students, a frequent problem in brazilian higher education. This structure traditionally is made upon qualitative criteria, which then relies on the good sense of the pedagogycal professionals involved in the production of this structure, which will reflect what the educators believe should be present in the curricula. In the present work, we’ll use the Synthetic Control Method, a linear regression model which can make use of specialist domain knowledge, together with the data of students from Computer Science degree in Universidade Federal do Ceará, in order to provide a quantitative toolkit to assist in the evaluation of the performance of the students and of the curricular structure based on the relationship between the disciplines of the course.