Can Large Language Models Integrate Spatial Data? Empirical Insights into Reasoning Strengths and Computational Weaknesses

2025-08-06

AI Summary

Existing rule-based approaches lack generalizability, while machine learning methods require extensive labeled data—both limiting scalability and robustness in urban spatial data integration, especially for human-informed spatial relationships (e.g., road–sidewalk associations). Method: This work investigates the applicability of large language models (LLMs) to integrate large-scale, heterogeneous, and noisy urban spatial data, proposing a dual-path optimization strategy: (1) augmenting LLMs with external spatial features to reduce reliance on intrinsic geometric reasoning, and (2) implementing a review-refinement mechanism for iterative error correction. Contribution/Results: Experiments reveal that although LLMs exhibit rudimentary spatial reasoning, they struggle to bridge high-level semantic understanding with low-level computational geometry tasks, often producing logically inconsistent outputs. The proposed framework significantly improves integration accuracy and delivers an interpretable, expert-controllable automation framework for domain practitioners.

Technology Category

• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Machine Learning: Large Multimodal Models (LMMs)
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

We explore the application of large language models (LLMs) to empower domain experts in integrating large, heterogeneous, and noisy urban spatial datasets. Traditional rule-based integration methods are unable to cover all edge cases, requiring manual verification and repair. Machine learning approaches require collecting and labeling of large numbers of task-specific samples. In this study, we investigate the potential of LLMs for spatial data integration. Our analysis first considers how LLMs reason about environmental spatial relationships mediated by human experience, such as between roads and sidewalks. We show that while LLMs exhibit spatial reasoning capabilities, they struggle to connect the macro-scale environment with the relevant computational geometry tasks, often producing logically incoherent responses. But when provided relevant features, thereby reducing dependence on spatial reasoning, LLMs are able to generate high-performing results. We then adapt a review-and-refine method, which proves remarkably effective in correcting erroneous initial responses while preserving accurate responses. We discuss practical implications of employing LLMs for spatial data integration in real-world contexts and outline future research directions, including post-training, multi-modal integration methods, and support for diverse data formats. Our findings position LLMs as a promising and flexible alternative to traditional rule-based heuristics, advancing the capabilities of adaptive spatial data integration.

Problem

• Can LLMs effectively integrate heterogeneous urban spatial datasets?
• Do LLMs struggle with macro-scale spatial reasoning tasks?
• Can LLMs replace rule-based methods for spatial data integration?

Innovation

• LLMs for integrating heterogeneous urban spatial datasets
• Review-and-refine method corrects erroneous LLM responses
• Multi-modal integration supports diverse data formats