EUVIMEDEUVIMED entdecken
Crossref

A Question‐Answering Framework for Geoscientific Operator Retrieval Based on Large Language Models and Knowledge Graphs

Mingda Zhang, Zhiwei Liang, Qian Yang, Rong Li, LiLong Wu

Transactions in GIS · 2026-11 · 30 (7)

DOI: 10.1111/tgis.70456 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Transactions in GIS, online erschienen am 06.11.2026. Verfasst von Mingda Zhang und 4 weiteren Autor:innen. Die Arbeit nennt 35 Literaturangaben. Gefördert von: National Key Research and Development Program of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The widespread use of GIS technology has led to the development of many open‐source geoinformation tools, such as QGIS and GRASS GIS. Users often find it hard to locate the right operators and access knowledge on different platforms. Large language models (LLMs) are good at natural language tasks but are less accurate in specialized domains. To address this, we propose a question‐answering framework for geoscientific operators. This framework combines LLMs with Graph‐based Retrieval‐Augmented Generation (GraphRAG), which increases the precision and relevance of operator retrieval. We organize geoscientific algorithms and operator details in a structured Geoscientific Operator Knowledge Graph (GeoOpKG). This structure enables efficient retrieval through multi‐relational context encoding with Relational Graph Convolutional Networks (R‐GCN) and Cross‐Encoder re‐ranking. LLMs improve fluency and reasoning, making question‐answering more reliable. Experiments show our approach achieves higher precision than other LLM‐based systems for known, similar, and unknown operator queries. The framework raises retrieval precision and answer reliability, helping users understand and access geoscientific operator knowledge.

Förderung

  • National Key Research and Development Program of China

Zitieren

APA

Zhang, M., Liang, Z., Yang, Q., Li, R., & Wu, L. (2026). A Question‐Answering Framework for Geoscientific Operator Retrieval Based on Large Language Models and Knowledge Graphs. Transactions in GIS, 30(7). https://doi.org/10.1111/tgis.70456

Vancouver

Zhang M, Liang Z, Yang Q, Li R, Wu L. A Question‐Answering Framework for Geoscientific Operator Retrieval Based on Large Language Models and Knowledge Graphs. Transactions in GIS. 2026;30(7). doi:10.1111/tgis.70456

BibTeX
@article{Zhang2026,
  author = {Zhang, Mingda and Liang, Zhiwei and Yang, Qian and Li, Rong and Wu, LiLong},
  title = {A Question‐Answering Framework for Geoscientific Operator Retrieval Based on Large Language Models and Knowledge Graphs},
  journal = {Transactions in GIS},
  year = {2026},
  volume = {30},
  number = {7},
  doi = {10.1111/tgis.70456}
}
Kostenlos starten →