Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
July 10, 2026ISPRS International Journal of Geo-InformationOpen Access

GeoIR-Compiler: A Geospatial Intermediate Representation and Compilation Framework for Chinese Urban Spatial Question Answering

View Full Paper
Ask AI
Bookmark
Share

Authors

CZC L ZhangJDJiqiu DengHZHui Zhang

Discussion

Loading...

Member takes

Overview

Randomized trial demonstrates enhanced accuracy in SQL generation for Chinese urban spatial queries, implying improved database access.

Key Points

  • This research aims to improve the accuracy of SQL generation for spatial queries in Chinese urban contexts by utilizing a specialized framework.
  • Developed a geospatial intermediate representation (GeoIR) for Chinese questions.
  • Constructed benchmarks, NJ-GeoIR-700 and WH-GeoIR-700, for evaluating SQL generation accuracy.
  • Compared performance of GeoIR-Full with Direct-SQL, DAIL-SQL, and Linking-SQL under controlled conditions.
  • GeoIR-Full achieved mean execution accuracies of 0.7271 on Nanjing and 0.7363 on Wuhan.
  • Outperformed other SQL generation methods under the fixed evaluation protocol.
  • Grounding was found to significantly enhance accuracy, although it may reduce answer reliability.

Cite This Study

Zhang et al. (2026) studied this question.

synapsesocial.com/papers/6a508d096eeac72a437a0be0https://doi.org/10.3390/ijgi15070310
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Empirical Post-Generation Correctness Assessment for Natural-Language Access to Urban PostGIS Databases2026
  2. 2Intent-Driven Hybrid Semantic–Spatial Retrieval–Augmented Generation for Intelligent Prospecting with GIS Visualization2026
  3. 3GeoSQL-Eval: First Evaluation of LLMs on PostGIS-Based NL2GeoSQL Queries2025
  4. 4CGD-QCSF: A Code Generation-Driven Query–Computation Separation Framework for Natural Language Geospatial Analysis2026
  5. 5Text‐to‐SpatialSQL: A LLM Based Method for Generating Spatial SQL Queries With Geo‐Knowledge Extracted From Software User Manual2026