Chain-of-programming (CoP): empowering large language models for geospatial code generation task

Chain-of-programming (CoP): empowering large language models for geospatial code generation task

Large Language Models (LLMs) have driven the development of geospatial code generation but face challenges such as incomplete user requirements and insufficient knowledge of platform-specific syntax, often leading to ‘code hallucination.’ To address this issue, this paper proposes the Chain-of-Progr...

Full description

Saved in:
Bibliographic Details
Main Authors: Shuyang Hou, Haoyue Jiao, Zhangxiao Shen, Jianyuan Liang, Anqi Zhao, Xiaopu Zhang, Jianxun Wang, Huayi Wu
Format: Article
Language:English
Published: Taylor & Francis Group 2025-08-01
Series:International Journal of Digital Earth
Subjects:
Geospatial code generation
Large language models
Code hallucination
Geographic information science
Google earth engine
Online Access:https://www.tandfonline.com/doi/10.1080/17538947.2025.2509812
Tags: Add Tag
No Tags, Be the first to tag this record!

Internet

https://www.tandfonline.com/doi/10.1080/17538947.2025.2509812

Similar Items