Guidelines for the Application of Data Mining to the Problem of School Dropout

• Autor(a) principal: de Carvalho, Veronica Oliveira [UNESP]
• Data de Publicação: 2022
• Outros Autores: Penteado, Bruno Elias, de Sousa, Leandro Rondado [UNESP], Affonso, Frank José [UNESP]
• Tipo de documento: Artigo de conferência
• Idioma: eng
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://dx.doi.org/10.1007/978-3-031-14756-2_4; http://hdl.handle.net/11449/242213
• topic: Data mining; Revisited; School dropout; Systematic Literature Mapping

Dropout is a complex phenomenon based on interrelated factors such as personal, institutional, structural, sociocultural, among other ones. It represents a waste of resources for students, their families, schools and society, and continues to be a challenge for educational institutions. In the last decade, the growing amount of data from educational institutions and the emergence of data science have led to data mining methodologies to explore this problem empirically. In this work, we map the literature on how data mining has been addressed face-to-face dropout. We synthesize different aspects, all of them related to steps of a generic data mining process. Our findings reveal a low level of formalism in theories, methodologies and pre-processing steps, with most papers making comparisons of different algorithms and features on the data available in the institution’s information system. Finally, we present some guidelines that can be used to improve the research on this topic.