Qualitative representation and management of place in GIS: A multi-hierarchical framework

Satoti, Abdurauf 2026. Qualitative representation and management of place in GIS: A multi-hierarchical framework. PhD Thesis, Cardiff University. Item availability restricted.

	PDF (Abdurauf Satoti, PhD, Thesis) - Accepted Post-Print Version Restricted to Repository staff only until 28 August 2027 due to copyright restrictions. Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (38MB)
	PDF (Cardiff University Electronic Publication Form) - Supplemental Material Restricted to Repository staff only Download (1MB)

Abstract

Geographic Information Systems (GIS) predominantly represent places using geometric coordinates. This approach offers limited support for the qualitative descriptions humans naturally use to communicate about space. While prior work in qualitative spatial reasoning and semantic web technologies has explored place representation, existing approaches lack a unified framework for generating, storing, and querying qualitative place descriptions within operational GIS environments. This thesis addresses this gap by developing the Qualitative Place Model (QPM), a GIS-native framework that bridges human centric spatial descriptions with computational representation. The QPM formalises directional, containment, and proximity relationships as first-class entities that complement geometric representations. The QPM implements a homogeneous formalism in which spatial units and named places are treated uniformly, enabling deterministic generation of unique qualitative place descriptions. To address limitations of traditional spatial divisions, the framework integrates Discrete Global Grid Systems (DGGS), specifically H3 and S2, as globally consistent hierarchical representations. The framework enables qualitative place creation through spatial expressions (e.g., inside, next to), while a novel H3-based methodology transforms irregular, hierarchical spatial units into hexagonal representations for systematic computation of 17 distinct relationship types. Implementation uses ArcGIS Pro with optional RDF/GeoSPARQL export. Empirical validation using 52,830 place entities across traditional (e.g., Administrative) and DGGS-derived hierarchies demonstrates 100% uniqueness for all generated descriptions. Logical closure achieves 10.6-fold inference expansion for traditional hierarchies, while DGGS-based representations achieve 99.9999% graph coverage, demonstrating structural completeness. Cross hierarchy relationship computation shows 3.6-fold performance improvement over geometry-based methods. Evaluation via 26 competency questions confirms complete coverage (8/8) of spatial semantic requirements, exceeding the Ordnance Survey UK baseline of 6/8. The thesis contributes a formal model for qualitative place representation, a scalable inference methodology grounded in logical closure, and empirical validation demonstrating national-scale feasibility within standard GIS workflows.

Item Type:	Thesis (PhD)
Date Type:	Completion
Status:	Unpublished
Schools:	Schools > Computer Science & Informatics
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science Q Science > QA Mathematics > QA76 Computer software
Date of First Compliant Deposit:	28 August 2026
Date of Acceptance:	25 August 2026
Last Modified:	28 Aug 2026 14:31
URI:	https://orca.cardiff.ac.uk/id/eprint/189229

Actions (repository staff only)

Edit Item