Semantic Web Spatial Rule Language (SWSRL)

This foundational research addresses the limitations of OWL (Web Ontology Language) for representing and reasoning with geographic ontologies. While OWL provides a rich form of reasoning through its description logic underpinnings, it cannot represent spatial datatypes, compute information using spatial operators, or support the complex property composition needed for qualitative spatial reasoning.

The work proposes SWSRL (Semantic Web Spatial Rule Language), a new language based on the Description Logic Programs paradigm (DLP) and Logic Program semantics. SWSRL supports the expression of geospatial ontological axioms together with geospatial integrity and deduction rules, providing a foundation for building and maintaining consistent geo-ontologies on the Semantic Web.

Research Challenges

• OWL Limitations: OWL is not an integrity checking language due to its non-unique name and open world assumptions
• No Spatial Support: OWL cannot represent spatial datatypes, compute spatial operators, or use spatial indexes
• Property Composition: OWL does not support the complex property composition needed for qualitative spatial reasoning
• Geo-ontology Evolution: Maintaining consistency when integrating new data sets from different sources

Key Contributions

• Hybrid Framework: Integrates qualitative symbolic information in SWSRL with quantitative geometric information using spatial datatypes in a spatial database
• Prioritised Default Logic: Allows expression of default integrity rules and their exceptions
• Interleaved Inference: On-the-fly computation of qualitative or quantitative spatial relations
• OGC Compliance: Supports OGC-compliant spatial syntax
• Rule Metadata: Standardised definition of rule metadata for construction, description, identification and categorisation within large rule sets

Technical Approach

The SWSRL framework bridges the gap between symbolic qualitative reasoning and computational geometric operations. The language enables:

• Expression and enforcement of spatial integrity constraints on geo-ontologies
• Deduction of implicit spatial relationships from explicit ones
• Integration with spatial databases for quantitative computation
• Scalable reasoning over large geographic datasets

Evaluation

The language and engine were evaluated using both synthetic and real geographic datasets in the context of developing geographic ontologies for information retrieval on the Semantic Web. Empirical experiments tested the scalability and applicability of the developed framework.

Impact and Legacy

This foundational work established the basis for subsequent research in qualitative spatial reasoning and place modelling, including:

• Qualitative Place Models on the Linked Data Web (Almuzaini, 2017)
• GIS-Native Framework for Qualitative Place Models (Satoti, ongoing)
• Global Place Knowledge Graphs using DLIG (Muhajab, ongoing)