SIDE – INTELLIGENT ESCAPE DETECTION SYSTEM
APPLICATION IN A FEDERAL EDUCATION INSTITUTION
Abstract
Controlling school dropout requires integrated actions and a collective commitment to ensure that all students have access to quality education. Higher education institutions must take measures to prevent dropout, which may include offering academic and financial support programs, creating internship and job opportunities for students, raising awareness about the importance of higher education, among other measures. Machine learning is a valuable tool for preventing dropouts, as it allows data collection and analysis to identify at-risk students, customize support programs, monitor progress, and evaluate the effectiveness of prevention programs. In Brazil, according to the PNAD (Pesquisa Nacional por Amostra de Domicílios), in 2019, around 89.2% of the population, between 15 and 17 years old, was included in the schooling rate, demonstrating the long road to be traveled to reach the universalization proposed in the 1988 Constitution. Thus, this research aimed to develop a system of evasion using Data mining and Machine learning, as an efficient means of identifying evasion in educational institutions. Based on information from the students of the Federal Institute of Education, Science and Technology – IFSP campus Itapetininga, data were obtained such as place of residence, means of transport used to reach the institution, motivation for enrolling, among other factors that are significant in the analysis of the probability of a student dropping out. The program demonstrated who is most likely to drop out of the student body, helping managers to allocate resources and make the necessary decisions. The results obtained in the two models had an accuracy of 86% in a model for the analysis of dropouts and another model with an accuracy of up to 100% with small oscillations due to the lack of coherence of some of the data provided.
References
AGRUSTI, F.; BONAVOLONTÀ, G. University dropout prediction through educational data mining techniques: a systematic review. Journal of e-Learning and Knowledge Society, v. 15, n. 3, p. 45-60, 2019. Disponível em: https://doi.org/10.20368/1971-8829/1135017. Acesso em: 1 nov. 2024.
ANPEC - Associação Nacional dos Centros de Pós-Graduação em Economia. Retenção e evasão no ensino superior brasileiro: uma análise dos efeitos da bolsa permanência do PNAES. In: ENCONTRO DA ANPEC, 14., 2018, São Paulo. Anais […]. São Paulo: Anpec, 2018. Disponível em: https://link.ufms.br/osNnk. Acesso em: 20 ago. 2022.
ARARUNA, L. T.; SANTOS, T. A.; FIAMENGUI, A. H. R.; DECANINI, J. G. M. S. Sistema inteligente para identificação de grupos de risco relacionados à evasão escolar. Sinergia São Paulo, v. 16, n. 4, p. 263-268, 2015. Disponível em: https://link.ufms.br/ZBNrk. Acesso em: 1 nov. 2024.
ASHA, P. et al. Predicting university dropout through data analysis. In: INTERNATIONAL CONFERENCE ON TRENDS IN ELECTRONICS AND INFORMATICS, 3., 2020, Tamil Nadu. Anais […]. Tamil Nadu: ICOEI, 2020. Disponível em: https://link.ufms.br/hbX70. Acesso em: 1 nov. 2024.
CARRANO, D.; ALBERGARIA, E. T.; INFANTE, C.; ROCHA, L. Combinando Técnicas de Mineração de Dados para Melhorar a Detecção de Indicadores de Evasão Universitária. In: SIMPÓSIO BRASILEIRO DE INFORMÁTICA NA EDUCAÇÃO, 1., 2019, Rio de Janeiro. Anais […]. Rio de Janeiro: SBIE, 2019. Disponível em: https://link.ufms.br/gk75P. Acesso em: 1 maio 2022.
COSTA, J. B.; BORGES, A. P. F. Evasão escolar no curso de Eletrotécnica do Ensino Médio Integrado do IF Sertão Pernambucano. Semiárido de Visu, Petrolina, v. 3, n. 7, p. 366-370, 2019. Disponível em: https://link.ufms.br/jjbkt. Acesso em: 15 set. 2022.
ERNÁNDEZ-GARCÍA, A. J. et al. A real-life machine learning experience for predicting university dropout at different stages using academic data. IEEE Access, v. 9, p. 133076-133090, 2021. Disponível em: https://doi.org/10.1109/ACCESS.2021.3115851. Acesso em: 1 nov. 2024.
HENRIQUES, F. G. Evolução histórica, expansão e financiamento da educação superior no Brasil: das primeiras faculdades à expansão do início do século XXI. Comunicologia, Brasília, v. 11, n. 2, p. 124-145, 2018. Disponível em: https://doi.org/10.31501/comunicologia.v11i2.10058. Acesso em: 1 mar. 2022.
JORNAL NACIONAL. IBGE mede o problema nacional da evasão escolar. Jornal Nacional, São Paulo, 2020. Disponível em: https://link.ufms.br/Gq5W6. Acesso em: 30 jul. 2022.
MARIA, W.; DAMIANI, J. L.; PEREIRA, M. Rede Bayesiana para previsão de evasão escolar. In: V CONGRESSO BRASILEIRO DE INFORMÁTICA NA EDUCAÇÃO, 5., 2016, Uberlândia. Anais […]. Uberlândia: CBIE, 2016. Disponível em: https://link.ufms.br/xBJR8. Acesso em: 2 jul. 2022.
MEEDECH, P.; IAM-ON, N.; BOONGOEN, T. Prediction of student dropout using personal profile and data mining approach. In: LAVANGNANANDA, K., PHON-AMNUAISUK, S., ENGCHUAN, W., CHAN, J. (ed.). Intelligent and Evolutionary Systems. Proceedings in Adaptation, Learning and Optimization. Springer, Cham: 2015. v. 5. Disponível em: https://doi.org/10.1007/978-3-319-27000-5_12. Acesso em: 12 set. 2022.
MINGHELLI, M. Institutos Federais de Educação, Ciência e Tecnologia: um futuro incerto. Encontros Bibli: revista eletrônica de biblioteconomia e ciência da informação, v. 23, n. 51, p. 157-165, 2018. Disponível em: https://doi.org/10.5007/1518-2924.2018v23n51p157. Acesso em: 1 nov. 2024.
MINISTÉRIO DA TRANSPARÊNCIA E CONTROLADORIA-GERAL DA UNIÃO. Relatório de gestão exercício 2017. 2018. Disponível em: https://link.ufms.br/zil3h. Acesso em: 15 set. 2022.
ORTIZ-LOZANO, J. M. et al. University student retention: best time and data to identify undergraduate students at risk of dropout. Innovations in Education and Teaching International, v. 57, p. 74-85, 2020. Disponível em: https://doi.org/10.1080/14703297.2018.1502090. Acesso em: 1 nov. 2024.
PÉREZ, B; CASTELLANOS, C; CORREAL, D. Applying data mining techniques to predict student dropout: a case study. In: 1st COLOMBIAN CONFERENCE ON APPLICATIONS IN COMPUTATIONAL INTELLIGENCE, 15., 2018, São Paulo. Proceedings […]. Bogotá: IEEE, 2018. Disponível em: https://doi.org/10.1109/ColCACI.2018.8484847. Acesso em: 5 jul. 2022.
SILVA FILHO, R. B.; ARAÚJO, R. M. L. Evasão e abandono escolar na educação básica no Brasil: fatores, causas e possíveis consequências. Educação Por Escrito, [S. l.], v. 8, n. 1, p. 35, 2017. Disponível em: https://link.ufms.br/WpSem. Acesso em: 15 set. 2022.
SOSU, E. M.; PHEUNPHA, P. Trajectory of university dropout: investigating the cumulative effect of academic vulnerability and proximity to family support. Front. Educ., v. 12, p. 89-104, 2019. Disponível em: https://link.ufms.br/YbrX9. Acesso em: 1 nov. 2024.
UNIVERSIDADE DE LJUBLJANA. Orange. 2022. Disponível em: https://link.ufms.br/YklrB. Acesso em: 1 nov. 2024.
VILORIA, A. et al. Integration of data technology for analyzing university dropout. Procedia Computer Science, v. 155, p. 569-574, 2019. Disponível em: https://doi.org/10.1016/j.procs.2019.08.079. Acesso em: 1 nov. 2024.
Edutec uses the Creative Commons License - Attribution 4.0 International as a basis. It believes in the importance of the open access movement in scientific journals, such as the Open Archives Initiative.
By submitting a text to the journal, the authors agree to the following terms:
- The authors agree to license works accepted for publication under the Creative Commons License - Attribution 4.0 International.
- With CC-BY 4.0 licensing, authors maintain intellectual rights over the text and grant Edutec the right to first publication.
- The authors authorize the sending and indexing of texts written by them in databases and academic and scientific information portals.
- Authors are allowed and encouraged to publish and distribute their work online, in institutional repositories, on personal pages and academic social networks, after the editorial process, as long as the forms of licensing and presentation of the text are respected, in addition to the indication of the place of origin of the publication, in this case, the link to the Edutec.
- The authors of the works published in the Edutec are expressly responsible for their content.
- All works submitted to Edutec that have images, photographs, figures in which images of human beings are registered in their bodies must be accompanied by an Image Use Authorization Term by the member participating in the image and, in the case of children, family members of exposed children, with data and signature.
- In commitment to practices for the openness and popularization of science, Revista Edutec may create content (including with the participation of Generative Artificial Intelligence, always with editorial supervision) for scientific dissemination of the publication on social media, citing the authorship and the access link.











2.png)

