Machine Learning e ensino individualizado na Matemática: uma ferramenta para o professor

• Autor(a) principal: Hirai, Cíntia Yumi
• Data de Publicação: 2021
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da UFSCAR
• Texto Completo: https://repositorio.ufscar.br/handle/ufscar/14607
• topic: Ensino de Matemática; Taxonomia de Bloom; Machine Learning; Mathematics teaching; Bloom's Taxonomy; CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO; CIENCIAS EXATAS E DA TERRA::MATEMATICA; CIENCIAS HUMANAS::EDUCACAO::ENSINO-APRENDIZAGEM; CIENCIAS HUMANAS::EDUCACAO::ENSINO-APRENDIZAGEM::TECNOLOGIA EDUCACIONAL

This research aims to investigate how the use of Machine Learning can contribute to the teacher in the identification of the mathematical skills of the students of the three years of High School for individualized teaching proposals. The use of Machine Learning shows that we can program a machine to learn things and perform certain tasks on its own. The Orange Canvas software, which uses Machine Learning tools, identifies the student's profile according to Bloom's Taxonomy. Bloom's taxonomy or educational taxonomy aims to contribute to the teaching process as well as to guide the evaluation procedures. This is a practical tool for evaluating individual performance, which we used in the current research to identify the students' skills in three domains: cognitive, affective, and psychomotor. The research was exploratory and had a qualitative approach, in which the modality of case study was chosen, where we collected the data through online questionnaires made for a group of high school students, who were being tutored by the researcher. The data collection was carried out in two stages, the first one for the machine to start its calibration, and the second one for the machine to identify the students' skills and, thus, easily determine which instruments, methods, and techniques we should use for students, to classify educational behaviors and help with planning, organizing, and managing learning objectives. We observed that the use of Machine Learning on the Orange Canvas greatly helped the researcher to identify the needs of the students, providing her the opportunity to prepare activities suitable for the development of the skills of each student. Based on the current study, if the researcher had not used Machine Learning on the Orange Canvas, she would have to apply a diagnostic test to manually classify each student and discover their individual needs, concluding this technology can be a useful tool for teachers in the elaboration of activities for individualized teaching.