Análise de desempenho acadêmico utilizando redes bayesianas: um estudo de caso

Abstract

Dropout and failure rates in the initial programming courses at the UFPB-Campus IV are considered high. Out of 759 students enrolled in these courses, from 2011.1 semester to 2014.1, 251 did not a success and 303 dropout. In this context, this paper presents a case study involving the analysis of the students performance in the introduction to Programming courses in Computer Science and Information Systems Degrees, at UFPB-Campus IV, using the Bayesian networks technique (RB) to extract implicit knowledge. This, in order to identify possible causes of abandonment and reproof. The networks were modeled manually, by domain expert, in the Netica tool and also automatically generated in the GeNie 2.1 software. Among the discoveries, it was possible to identify the dropout student profile and that 89.9% of students who do not obtain score more than 70% in the first unit work also get low score in the first test.