A GEOSPATIAL KNOWLEDGE GRAPH PROTOTYPE FOR NATIONAL TOPOGRAPHIC MAPPING

Abstract. Knowledge graphs are a form of database representation and handling that show the potential to better meet challenges data interoperability, semi-automated information reasoning, retrieval. Geospatial knowledge (GKG) have at their core specialized forms applied ontology provide coherent spatial context domain including non-spatial attributes. This paper discusses research toward development prototype GKG based on national topographic databases geospatial feature instances, attributes, properties, metadata, annotations. The capture represent geographic semantics inherent in source data, align such graph models with standards where possible, test logical computations, visualize using cartographic user interface. Data integration from outside sources was tested through SPARQL GeoSPARQL queries. Called MapKB, approaches this use number software components build system architecture aligned those objectives composed entirely free open-source software. design were validated reasoning competency questions. Technical aspects succeeded, but customization found be needed for user-based design.