Uma abordagem ontológica para modelação de informação espaciotemporal com aplicações em transportes

• Autor(a) principal: Alexey Seliverstov
• Data de Publicação: 2015
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: https://repositorio-aberto.up.pt/handle/10216/82891
• topic: Engenharia electrotécnica, electrónica e informática; Electrical engineering, Electronic engineering, Information engineering

This dissertation work is developed in the context the spatiotemporal data analysis, specifically transportation data. Nowadays there is a great amount of available data. This is because almost every electronic device that is around us generates data. Usage of this resource attracts researchers and companies to research and invest on ways to do so. Even roads are generating data through sensors. These sensors can be fixed, which is the case of inductive loops and video surveillance cameras, as well as mobile, such as the case of floating cars. Data readability, comprehension and usage are of vital importance for a proper traffic analysis. To be possible to analyse data from different data sources, or sensors in this specific case, it is necessary to integrate this data. Bearing this in mind, it is also known that this task is difficult if performed manually. So it is necessary to research new techniques which would allow us to simplify this task, and preferentially automate it. Fortunately there are methods based on ontologies which help tackling those problems. They allow us to integrate many data sources and maintain the original meaning and form of such data. Based on this methodology sensors and data will be approached on an ontology basis, allowing for their integration in one single representation. After that, such an integrated system will be shared through a service-oriented architecture. This will allow clients to easily access the data present in the system. GPS logs, OpenStreetMaps and inductive loop data will be used to populate this ontological model. Afterwards data from simulators such as SUMO will also be used to fill in possible gaps that might not be covered by real data. One expected result from this dissertation is that a scientific geographic repository can devised and implemented to be used for the transportation analysis by many clients with different needs and interests.