Framework de inteligência artificial para compreensão e predição de fatores que motivam a prática de atividades físicas

• Ano de defesa: 2023
• Autor(a) principal: Lima, Saulo Vinicius Stopa de
• Orientador(a): Araújo, Sidnei Alves de
• Banca de defesa: Araújo, Sidnei Alves de , Lucareli, Paulo Roberto Garcia , Librantz, Andre Felipe Henriques
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Nove de Julho
• Programa de Pós-Graduação: Programa de Pós-Graduação em Informática e Gestão do Conhecimento
• Departamento: Informática
• País: Brasil
• Palavras-chave em Português: atividades físicas; fatores motivacionais; inteligência artificial; árvores de decisão; mineração de dados; lógica fuzzy
• Palavras-chave em Inglês: physical activities; motivational factors; artificial intelligence; decision trees; data mining; fuzzy logic
• Área do conhecimento CNPq: CIENCIA DA COMPUTACAO::SISTEMAS DE COMPUTACAO
• Link de acesso: http://bibliotecatede.uninove.br/handle/tede/3521

Science has demonstrated that practicing physical activities improves health, contributing to reducing the risk of chronic diseases such as diabetes, hypertension, certain types of cancer, osteoporosis and depression. However, a common challenge for most people is to have the motivation to practice physical activities regularly, especially as they have less and less need to travel to carry out their daily activities, due to the various technological resources at their disposal and the increase in remote work practices, especially after the COVID 19 pandemic. In this scenario, it is important to study and develop tools that facilitate the understanding of the factors that motivate the practice of physical activities in order to assist health professionals in customizing of services. In this research, an Artificial Intelligence (AI) framework is proposed to help understand and predict the factors that motivate people to practice physical activities, based on their socioeconomic profiles. Initially, through Data Mining (DM), patterns described by IF...THEN rules are generated by Decision Trees (DT) and the Apriori algorithm. Subsequently, these rules are used to build a fuzzy inference mechanism (FIM), which forms a Recommendation System (RS) to indicate the factors that motivate a person and the most appropriate activities, based on their profile. In the experiments conducted, a database of 140 people was used, provided by a company in the field of developing applications for online training. The results obtained in the DM (accuracies ≥ 90% and Kappa indices ≥ 62%) showing the presence of consistent patterns and the outcomes achieved by the FIM, when coupled with the RS, show that the developed framework can be useful for education professionals physics in the guidance and development of personalized training for each person.