Resumo-Com o objetivo de divulgar o potencial e a aptidão de Data Mining Educacional (EDM), como um instrumento de análise e de investigação, no apoio à gestão de instituições dedicadas ao ensino, apresenta-se, no presente artigo, uma sucinta descrição de alguns dos estudos mais relevantes da área. A análise efetuada permite evidenciar as inovações que o EDM tem vindo a promover, bem como as tendências de investigação atuais e futuras. Palavras Chave-data mining educacional, data mining, eficiência institucional.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use.
Resumo-No presente artigo apresenta-se uma metodologia desenvolvida com base no algoritmo random forest, para prever precocemente e de forma rigorosa o desempenho académico de graduação dos estudantes de uma instituição de ensino superior politécnico. A abordagem seguida permitiu isolar 11 variáveis explicativas, a partir de um conjunto inicial de cerca de meia centena, que garantem uma boa capacidade preditiva do modelo (R 2 =0.79). Estas variáveis revelam aspetos fundamentais para a definição de estratégias de gestão centradas na promoção do sucesso académico.
School dropout in higher education is an academic, economic, political and social problem, which has a great impact and is difficult to resolve. In order to mitigate this problem, this paper proposes a predictive model of classification, based on artificial neural networks, which allows the prediction, at the end of the first school year, of the propensity that the computer engineering students of a polytechnic institute in the interior of the country have for dropout. A differentiating aspect of this study is that it considers the classifications obtained in the course units of the first academic year as potential predictors of dropout. A new approach in the process of selecting the factors that foreshadow the dropout allowed isolating 12 explanatory variables, which guaranteed a good predictive capacity of the model (AUC=78.5%). These variables reveal fundamental aspects for the adoption of management strategies that may be more assertive in the combat to academic dropout.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.