Building Interactive Webmaps and Browser-based Historical GIS Tools with ChatGPT

Wright Kennedy, University of South Carolina

This paper details innovative use cases of ChatGPT for research and teaching in the field of historical mapping. ChatGPT and other large language models (LLMs) can reduce barriers to entry and lower resource requirements for creating interactive webmaps and online tools customized for historical GIS research and presentation. I leveraged ChatGPT to develop a browser-based JavaScript tool that joins tables and shapefiles downloaded from NHGIS and outputs a new GeoJSON file. The tool runs locally, ensuring that the dataset remains on the user’s computer, which improves privacy and eliminates any server costs for processing the data. Future work will focus on optimizing the tool for more complex geometries. Integrating ChatGPT into a digital history methods course accelerated student learning and allowed for an increased focus on advanced mapping concepts. Students learned browser-based workflows that combined ChatGPT with other free tools and resources, such as Python in Google Colab and hosting on GitHub Pages. Students with very limited coding experience built and deployed basic Leaflet webmaps within a one-hour class. While only scratching the surface, these experiments illustrate the potential of ChatGPT and other LLMs to streamline historical GIS processes and democratize access to advanced digital mapping tools, making them more accessible for educational and research purposes.

No extended abstract or paper available

 Presented in Session 167. New Methods in Historical Mapping: Deep Maps, ChatGPT, and New Media