Review of Large Language Models for Geospatial Query Understanding and Generation
Deepali Ahir · International Journal for Research in Applied Science and Engineering Technology · 2025
This paper examines Large Language Models (LLMs) for the task of interpreting and generating natural language queries, with a particular focus on the geographic data domain. This paper examines three primary functions: Text-to-SQL, which converts natural language queries to SQL queries for relational databases, Text-to-OverpassQL and Corpus query language, which enters the language analysis. This review examines various approaches used to improve the performance of Large Language Models (LLMs) in these areas, including rapid development, optimization, and data augmentation techniques. It also addresses issues such as the ambiguity of natural language, the complexity of geographic data, the need for specialized knowledge, and the visual problem. This paper describes in more detail the underlying data and measurement techniques used in this area. By tying together existing research, this paper highlights the potential of LLM to provide independent access to geographic data and identifies avenues for future research, such as improving security, integrating external information, and improving more comprehensive measurement