Content area

Abstract

Geographic Information Systems (GIS) have become indispensable tools in addressing complex spatial challenges across diverse domains, including urban planning, environmental conservation, and climate change. However, the technical expertise required to use GIS effectively, particularly proficiency in general-purpose programming languages (GPLs), remains a significant barrier to broader adoption. Domain-Specific Languages (DSLs) have proven to lower this barrier by offering tailored abstractions that simplify problem-solving in specific domains. Despite their potential, the GIS domain lacks a formal DSL framework that bridges the gap between the design of spatial analysis models and their practical implementation. This research addresses this gap by proposing Gaia, a declarative Domain-Specific Language designed to express geographic data models. Gaia decouples the specification of spatial operations (the “what”) from execution strategies (the “how”). The Gaia transpiler automates execution optimizations such as transformations, parallelization, and state management, thereby reducing the complexity of spatial modeling. Gaia is scoped to tackle predefined spatial analysis problems, enabling users to focus on model specification rather than implementation intricacies. Herein I elaborate the development lifecycle of Gaia from domain analysis to design and implementation following a Design Science Research (DSR) approach, where Gaia succeeded and where it came short. The results show that users with domain knowledge yet little programming skills perform much better in spatial analysis tasks using Gaia than a GPL, adding empirical evidence to one of the DSL community’s strongest claims. Most importantly, despite the successes resulting from GIS’s embrace of database management systems in the late 1980s and 1990s, the performance of Gaia is a strong indicator that the GIS community should not sleep on the recent advances in analytical databases. Finally, I reflect on the qualitative feedback from Gaia users and discuss unexpected and exciting open problems.

Details

1010268
Title
Declarative Design for Spatial Analysis Models
Number of pages
150
Publication year
2025
Degree date
2025
School code
0047
Source
DAI-A 87/3(E), Dissertation Abstracts International
ISBN
9798293812356
Committee member
Shabtai, Itamar; Uwaoma, Chinazunwa; Cáceres, Claudia
University/institution
The Claremont Graduate University
Department
Center for Information Systems and Technology
University location
United States -- California
Degree
Ph.D.
Source type
Dissertation or Thesis
Language
English
Document type
Dissertation/Thesis
Dissertation/thesis number
32173341
ProQuest document ID
3246819251
Document URL
https://www.proquest.com/dissertations-theses/declarative-design-spatial-analysis-models/docview/3246819251/se-2?accountid=208611
Copyright
Database copyright ProQuest LLC; ProQuest does not claim copyright in the individual underlying works.
Database
ProQuest One Academic