Data-Centric Solutions – A Case Study for Road Data Based on Knowledge Graphs

Abstract. This study demonstrates how a data-centric architecture can be implemented with a knowledge graph based on generic datasets to enable a road maintenance tools. Two datasets were used: a 3D city model (CityGML/3CIM) and a national road database. Road polygons (CityGML/3CIM) were generated from a municipality basemap and knowledge graphs were created for both datasets. The data-centric tool was implemented in the open-source GIS software QGIS and evaluated to find how well the generic datasets fulfill the data requirement of a road maintenance tool. Much of the data required for road maintenance could be obtained from the two generic datasets; however, some information is missing. To overcome this, the information models for the generic datasets need to be extended or a dedicated information model could be created. For the latter, the information model should only contain the required road maintenance attributes and links on object level to the CityGML/3CIM objects to avoid separate descriptions of the road geometries; such an approach would avoid the data-silo structure that is common within municipalities today.