DIGITAL LIBRARY
DATA MINING FOR OPEN EDUCATIONAL GOVERNMENTAL DATA: THE CASE STUDY OF BRAZILIAN HIGHER EDUCATION
Federal University of Ceará (BRAZIL)
About this paper:
Appears in: INTED2017 Proceedings
Publication year: 2017
Pages: 9369-9374
ISBN: 978-84-617-8491-2
ISSN: 2340-1079
doi: 10.21125/inted.2017.2211
Conference name: 11th International Technology, Education and Development Conference
Dates: 6-8 March, 2017
Location: Valencia, Spain
Abstract:
The implementation of high standards of education necessarily requires the implementation of mechanisms for monitoring ongoing actions and policies for different reasons. In Brazil, the National Exam for the Assessment of Student Performance (ENADE) is the responsible to assess the undergraduate students’ performance and then infer about the quality of undergraduate programs. The ENADE is conducted by the National Institute for Educational Studies and Research "Anísio Teixeira" (INEP) and it is part of the National Higher Education Assessment System (SINAES), which was created in order to evaluate the quality of undergraduate courses and higher education institutions throughout Brazil. Basically, the ENADE generate a dataset composed by the results of tests applied for more than 500.000 undergraduate students, which evaluates the retained knowledge of students in the end of the undergraduate course. These data are provided by the Ministry of Education and can be openly accessed by researchers. The aim of this study is to perform an exploratory data analysis in order to extract relations among the Brazilian higher education institutions and compare its performance with international university rankings regarding to assess the quality of education and research of the institution. Some traditional data mining techniques, such as principal component analysis and analysis of variance, have been performed in order to understand the clusters formed by universities whose quality standards are similar. The main idea of data mining (DM) is to transform raw data into useful information for decision making on educational contexts. The techniques have proved that is possible to extract intrinsic and latent information of the data and this information can be used to classify the institutions based on its quality achievements. The application of DM into educational governmental data comes up with the possibility to better understand the relations among knowledge retention throughout the years of formation. The latent variables formed four clear clusters and these clusters are associated with educational quality, this association was made possible according to the quality indexes provided by international evaluations which are the responsible of ranking higher education institutions based on quality, publication and teaching skills, and were used in this study as a benchmark.
Keywords:
Higher Education, Educational Data Mining, Decision Making.