Leveraging the SPARQL SERVICE Clause for External Computation on Linked Building Data
Year: 2026
Type: conference-paper
Abstract: The documentation of buildings accumulates heterogeneous data over time, such as inspection reports, plans, Building Information Modeling (BIM) models, and photographs. Correlating these sources requires domain-specific computation, such as distance calculations or topological reasoning. Linked Data approaches can integrate heterogeneous sources in their native formats and make them accessible through the query language SPARQL. Domain-specific computations, however, often fall outside the native scope of SPARQL, and existing engine-level extensions depend on vendor adoption, a condition that solutions in the Architecture, Engineering, and Construction (AEC) domain rarely meet. This paper proposes Functional Satellites (FS), a pattern for lightweight, self-hostable services that provide domain-specific computation within a standard SPARQL workflow by receiving delegated subqueries through the SERVICE clause, requiring no modification of the underlying SPARQL engine. The pattern is demonstrated through a spatial querying satellite implementing distance operations equivalent to the GeoSPARQL standard and is evaluated across three triplestore implementations. The evaluation shows that portability depends on query design and identifies a query pattern that is portable across the tested systems. The pattern thereby shows a path towards lowering the barrier for the AEC domain to provide its own specialised computation services, such as spatial reasoning over building data.
Authors: Schulz, Oliver, Oraskari, Jyrki, Göbels, Anne, Werbrouck, Jeroen, Beetz, Jakob
Language: English
Topic: Semantic Web and Ontologies
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7213285500
FWCI:
Cites:
Cited by:
Related to: