Ferramenta de recomendação híbrida de objetos de aprendizagem com predição das necessidades personalizadas de cada estudante

With a large amount of data available, it is increasingly difficult to identify information that will contribute to the student's learning process. In this scenario, the present study seeks to apply a hybrid model framework to recommend learning objects based on the preferences and needs of eac...

Descripción completa

Detalles Bibliográficos
Autor: Schrammel, Lucca Alexandre
Tipo de recurso: tesis de maestría
Estado:Versión publicada
Fecha de publicación:2023
País:Brasil
Institución:Universidade Federal de Santa Maria (UFSM)
Repositorio:Manancial - Repositório Digital da UFSM
Idioma:portugués
OAI Identifier:oai:repositorio.ufsm.br:1/30895
Acceso en línea:http://repositorio.ufsm.br/handle/1/30895
Access Level:acceso abierto
Palabra clave:Recomendação
Objetos de aprendizagem
Predição
Desempenho
Recommendation
Learning object
Predict
Performance
CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
Descripción
Sumario:With a large amount of data available, it is increasingly difficult to identify information that will contribute to the student's learning process. In this scenario, the present study seeks to apply a hybrid model framework to recommend learning objects based on the preferences and needs of each student. To identify students' needs, the framework carries out a performance prediction process to identify any future difficulties that a student may present. In this way, the recommendations generated precede a possible need for more specific learning objects for the needs of each student. Recommendations are generated from collaborative, content, and knowledge-based filtering methods. With the framework implemented, the validation of the algorithms made a significant contribution to the recommendations generated, considering that it was evident that the framework was capable of integrating filtering methods and data prediction algorithms, for recommending learning objects. The framework generated recommendations with accuracy above 80% in all test scenarios. Furthermore, it generated a low number of recommendations, which highlights that even if several possibilities of proposals were considered, only the learning objects with the highest probability of acceptance by students were recommended.