Application of Knowledge Discovery in Data Bases Analysis to Predict the Academic Performance of University Students Based on Their Admissions Test

Abstract

In 2012, the Ecuadorian Higher Education System implemented a standardized test called “Ser Bachiller” as a compulsory requirement to be admitted on an undergraduate program at any public university of the country. Based on the test, applicants receive a score that allows them to apply to the academic program of their choice. On a meritocratic basis, students with the highest grades from the examination process have greater opportunities to be admitted. The present study focuses on the admission process at Universidad Nacional de Chimborazo (Unach), a public university located in Riobamba- Ecuador. Through the application of knowledge discovery in database (KDD), this analysis generated a data warehouse (DW) after collecting the information of the tests scores from the admitted students and the relation with their academic performance once they were enrolled in the leveling courses. In addition, through Data Mining (DM) processes, it was possible to identify patterns that provide information and knowledge about the relationship between the admission score and the academic performance of the students. In parallel, through Business Intelligence (BI) reports, higher education institutions can inform their decisions in order to generate strategies for strengthening the processes of admission and knowledge leveling. This can allow them to guarantee a high academic performance among the admitted students, which additionally supports the general development of institutions and their undergraduate programs.