Bilateral Spatial Reasoning about Street Networks: Graph-based RAG with Qualitative Spatial Representations

2025/12/17 by Moratz, Reinhard, Daute, Niklas, Ondieki, James +3
Computer Science · Engineering · #Artificial Intelligence (cs.AI) #Constraint Satisfaction and Optimization #Data Management and Algorithms #FOS: Computer and information sciences #Spatial Cognition and Navigation

paper · doi:10.48550/arxiv.2512.15388

openalex publication_date 2025/12/17 · openalex created_date 2025/12/19 · openalex updated_date 2026/07/28

Abstract

This paper deals with improving the capabilities of Large Language Models (LLM) to provide route instructions for pedestrian wayfinders by means of qualitative spatial relations.

Citations

• From Local to Global: A Graph RAG Approach to Query-Focused Summarization

Related

• Algebraic Specifications of Wayfinding Using Cognitive Map
• Can LLMs Learn to Map the World from Local Descriptions?
• DistRAG: Towards Distance-Based Spatial Reasoning in LLMs
• RoadBench: Benchmarking MLLMs on Fine-Grained Spatial Understanding and Reasoning under Urban Road Scenarios
• An Evaluation of ChatGPT-4's Qualitative Spatial Reasoning Capabilities in RCC-8