A identificação de grupos de aprendizes no ensino presencial utilizando técnicas de clusterização
Resumo
Resumo: A utilização de pedagogias diferenciadas que considerem a heterogeneidade dos alunos de uma mesma sala-de-aula é, com certeza, uma ferramenta capaz de resgatar aprendizes condenados à não-aprendizagem e ao fracasso profissional. No entanto, aulas personalizadas no ensino presencial é algo ineficaz, além de impraticável. O "atendimento personalizado" em grupos homogêneos é uma possibilidade a se considerar. Este artigo descreve uma experiência de categorização de alunos utilizando técnicas de clusterização por aprendizado não-supervisionado com dados obtidos através de questionários e avaliações. Com isto, foi possível identificar grupos similares de aprendizes. Esta é mais uma técnica que poderá ser utilizada para criar e manter o modelo do estudante num Sistema Tutor Inteligente.
Abstract: The use of a different pedagogy that consider the students' heterogeneity in the same classroom is, certainty, a useful tool to rescue learners that was convicted to not learning and to a professional weakness. However, personalized lessons in classrooms are inefficacious and impracticable. The "personalized attendance" in homogeneous groups is a possibility to take into consideration. This paper describes an experience of students’ categorization using clustering techniques for unsupervised learning with data obtained from questionnaires and assessments. In this way, it was possible to identify similar groups of learners. This is one more technique that could be used to create and maintain the student model in an Intelligent Tutorial System.
Abstract: The use of a different pedagogy that consider the students' heterogeneity in the same classroom is, certainty, a useful tool to rescue learners that was convicted to not learning and to a professional weakness. However, personalized lessons in classrooms are inefficacious and impracticable. The "personalized attendance" in homogeneous groups is a possibility to take into consideration. This paper describes an experience of students’ categorization using clustering techniques for unsupervised learning with data obtained from questionnaires and assessments. In this way, it was possible to identify similar groups of learners. This is one more technique that could be used to create and maintain the student model in an Intelligent Tutorial System.
Texto completo:
PDFDOI: https://doi.org/10.5753/cbie.sbie.2003.495-504