Tesis Doctorales de la Universidad de Alcalá
Autor/aOkewu , Emmanuel
DepartamentoCiencias de la Computación
Director/aFernández Sanz, Luís
Codirector/aMisra , Sanjay
Fecha de defensa06/10/2020
CalificaciónSobresaliente Cum Laude
ProgramaIngeniería de la Información y del Conocimiento (RD 99/2011)
Mención internacionalNo
ResumenSustainability implies meeting the needs of present generation while not jeopardizing the needs of future generation. Another term for it is Sustainable Development which means balancing socio-economic development with environmental protection. Hence, the three pillars of sustainability (or sustainable development) are social protection, economic development and environmental conservation. The protocols (principles or practices) that guide sustainability are referred to as sustainable development practices. Overtime, global sustainable development plans have been put in place to enrich lives and livelihoods such as the Millennium Development Goals (MDGs) (2000 - 2015) and the Sustainable Development Goals (SDGs) (2015 ¿ 2030). However, implementation of past plans such as MDGs have been below expectations. The less-than-impressive performances of past plans have been partly blamed on data inadequacy and lack of technologies for extracting hidden and useful patterns in existing data for informed decision making. As a solution, this PhD Thesis proposes the integration of model-driven engineering and predictive analytics into existing sustainable development protocols for enhanced implementation of present and future international sustainable development plans. We are motivated by the current trend of data explosion and the hidden treasure in big data. We show, using the SDGs, that software architecture could be used to demystify sustainable development concepts for improved citizens-stakeholders engagement that promotes transparency, accountability and good governance for optimal utilization of scarce public resources. We also show that neural network model could be used to elicit useful patterns in data and predict, with high accuracy, future events for improved decision making especially in emergencies. Our findings and proposed solution have been published in impact journals and proceedings of highly rated international conferences.