Project Gallery
Linked Urban Data: Leveraging the capabilities of data analytics by using Knowledge Graph
This thesis presents a framework for Linked Urban Data, aiming to harness data analytics capabilities through a knowledge graph.
Urban data comes in different formats and is fragmented across different platforms. It can include morphological data, such as building volume, heights, building program information, or mobility data, including bus records and traffic volume, among other types. Thus, in this thesis, I categorize data as static or dynamic. Both static and dynamic data significantly influence the decision-making processes of planners and designers, impacting aspects such as site selection, road-, pathway, and layout design, but also traffic flow and transportation network efficiency. Therefore, the lack of such data and their relation can lead to poor design choices.
To structure and connect all these fragmented data, this thesis focuses on the following developments: (1) the utilization of existing urban ontologies such as OpenStreetMap ontology and the development of an Urban Dynamic Data Ontology (UDO) as a schema specification for static and dynamic urban data, (2) a workflow that allows using the aforementioned ontologies, this step also includes the creation of a knowledge graph by utilizing static and dynamic data from Stuttgart, (3) the illustration of the power of cross-database queries through three use cases (urban planner, real estate developer, and municipality official), and (4) the development of a mockup web interface demonstrating how the developed methodology and workflow can be used.
The methods integrated three datasets: (1) urban static data retrieved from OSM, (2) traffic volume data obtained from UTD19, a multi-city traffic dataset, and (3) historical public transport data retrieved from the Stuttgart VVS website. However, the method does not impose restrictions on the integration of new datasets into the framework.
The findings of this thesis highlight that urban ontologies that cover static and dynamic data aid in data structuring, making data queries more accessible, and enabling the discovery of previously unknown relationships, such as the connection between morphology data and traffic volume data. In conclusion, I firmly believe that using knowledge graphs for linking urban data can significantly assist urban designers, municipalities, and other stakeholders in making well-informed decisions.
ITECH M.Sc. Thesis Project 2023: Linked Urban Data: Leveraging the capabilities of data analytics by using Knowledge Graph
Xi Peng
Thesis Advisers: Diellza Elshani, Hana Svatoš-Ražnjević
Thesis Supervisor: Prof. Thomas Wortmann
Second Supervisor: Prof. Achim Menges