An Ontology-Based Framework for Geospatial Integration and Querying of Raster Data Cube using Virtual Knowledge Graphs
The combination of raster data cubes with other geospatial data forms, such as vector data, presents unique challenges in terms of effective management and representation using knowledge graphs. This integration holds significant potential for addressing the complexity of geospatial data and connecting raster data cubes with semantic technology standards. While various approaches have been explored in the past, they often lack formalization or focus solely on raster data cubes, missing the inclusion of semantic spatial entities and their spatial relationships. This limitation can hinder advanced geospatial queries and the enrichment of geospatial models.
In this GitHub project, we propose a comprehensive framework designed to facilitate semantic integration and advanced querying of raster data cubes, leveraging the concept of the virtual knowledge graph (VKG). Within this framework, we establish a semantic representation model for raster data cubes that extends the GeoSPARQL ontology. This model empowers us to merge the semantics of raster data cubes with feature-based models encompassing geometries, spatial relationships, and topological aspects. The result is the ability to formulate spatiotemporal queries using SPARQL naturally, employing ontological concepts at an appropriate level of abstraction.
To bring this framework to life, we provide an implementation based on a VKG system architecture. Furthermore, we conduct a thorough experimental evaluation, comparing our framework's performance and scalability with existing systems.