Um modelo computacional para a classificação automática do tipo de personalidade de estudantes do ensino a distância

Abstract

In the last decades, the role of personality traits has gained relevance for understanding, ex- plaining and more recently, the behavioral patterns that people exhibit when they experience different situations, including everyday situations such as: making decisions, using interpersonal relationships, dealing with finance, career and work management issues, etc. Research has demonstrated the importance and applicability of the concept of personality in the educational context; correlations were identified between the students’ personality with various subjects of interest to the learning process. For example, previous studies have identified a strong correlation between students’ personality and the scores obtained by those students in the learning assessment tests; it is possible, when knowing in advance the personality type of a student, to estimate with a good safety margin whether he will pass his course evaluations. However, the diagnosis of a student’s personality traits is not a simple activity and is usually carried out through interviews and observations, or through the use of questionnaires. In a distance learning environment, the application of both approaches proves to be even more complex, so research has sought new ways to make the diagnosis of personality traits viable. One of the alternatives is to carry out the diagnosis of the personality with the aid of mathematical models that are applied to analyze the behaviors and interaction patterns of students carried out in the distance learning environment. The objective of this work was to investigate the possibility of developing and evaluating a computational model that proposes to identify the personality traits of students through the analysis of the interactions carried out in the virtual learning environment.For this, an experimental research was carried out in the context of distance undergraduate courses at the Federal University of Alagoas. In the experiment, a personality questionnaire was applied to a sample of students and, subsequently, a set of classification algorithms was used to analyze the use of the learning environment, in order to obtain possible correlations between students’ personality traits. and their interactions in the virtual learning environment. As a general result, it was observed that none of the classification algorithms obtained satisfactory results.