Automatic Profiling and Natural Language Retrieval of Geospatial Vector Data with Sub-10B LLMs: A GeoChatbot Approach to Spatial Analysis

Abstract excerpt

<title>Abstract</title> <p>Geographic Information Systems (GIS) rely on vector data (points, lines, polygons) for spatial analysis, but querying complex PostGIS datasets remains challenging for non-experts, limiting decision-making. This paper presents a framework using compact LLMs (<10B parameters) to enable natural language spatial analytics on vector GIS data. Central to the system is LD-PROBE, an LLM-powered...

Topics

Open a Topic to create a Post that cites this publication.

• Anaplastic Lymphoma Kinase
• Biological Ontologies
• Cell Compartmentation
• Data Accuracy
• Drosophila melanogaster
• Flight, Animal
• Glycolysis
• Mutation

Identifiers and source

Literature Corpus work

0931661b-8c43-57b6-9c24-8b0c70648b2d

DOI

10.21203/rs.3.rs-9436033/v1