MaaSDB: Spatial Databases in the Era of Large Language Models (Vision Paper)

TL;DR

This paper explores integrating large language models with spatial databases to enable intuitive, accessible spatial data retrieval from both structured and unstructured sources, enhancing usability for diverse users.

Contribution

It proposes a novel framework for combining LLMs with spatial databases, facilitating natural language interaction with complex spatial data.

Findings

Demonstrates potential for LLMs to understand and query spatial data.

Shows improved accessibility for non-expert users in spatial data retrieval.

Lays groundwork for future LLM-based spatial database systems.

Abstract

Large language models (LLMs) are advancing rapidly. Such models have demonstrated strong capabilities in learning from large-scale (unstructured) text data and answering user queries. Users do not need to be experts in structured query languages to interact with systems built upon such models. This provides great opportunities to reduce the barrier of information retrieval for the general public. By introducing LLMs into spatial data management, we envisage an LLM-based spatial database system to learn from both structured and unstructured spatial data. Such a system will offer seamless access to spatial knowledge for the users, thus benefiting individuals, business, and government policy makers alike.