Onto Solar Flare - uma ontologia de domínio/aplicação sobre explosões solares sob escopo do método ECID

• Ano de defesa: 2020
• Autor(a) principal: França, Natália de Souza
• Orientador(a): Ribeiro, Marcela Xavier
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de São Carlos; Câmpus São Carlos
• Programa de Pós-Graduação: Programa de Pós-Graduação em Ciência da Computação - PPGCC
• Palavras-chave em Português: Ontologia; Explosões solares; ECID; Previsão
• Palavras-chave em Inglês: Ontology; Solar explosions; Prediction
• Área do conhecimento CNPq: CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: https://repositorio.ufscar.br/handle/20.500.14289/16233

Solar flares are phenomena capable of causing several damages in radio systems, communication satellites, global positioning systems (GPS), among others. The possible damage can vary according to the intensity of the explosion and these explosions are classified as A, B, C, M and X. Predicting this phenomenon is not an easy task due the sparse data and the unknowledge of relevante characteristics of the phenomenon. In this work we aim to improve the process of prediction of solar flares building the knowledge about the domain of the solar cycle structured in a domain ontology. The complexity of the data was abstracted by creating a domain Ontology in order to improve the prediction process. To apply the ontology, it was decided to incorporate it to the ECID solar explosion prediction method, which performs multi-class and multi-label prediction with treatment for imbalanced data. In order to contribute to research on solar flares, ECID was adapted to encompass the Onto Solar Flare ontology. The ontology was developed to semantically enrich the ECID algorithm, the combination of the two tools generated the ECID Onto. The results show a better distribution of classes, facilitating the final decision of the domain specialist.