Introduction

Geospatial data visualization is a crucial methodology in Digital Humanities (DH) research. Visually representing spatial information — such as the distribution of historical events, locations of cultural heritage sites, and demographic transitions — can yield new insights.

This article introduces Kepler.gl, an open-source geospatial data visualization tool developed by Uber.

What is Kepler.gl?

Kepler.gl is a web-based tool designed for intuitive and visually stunning visualization of large-scale geospatial datasets. It was open-sourced in 2018 by Uber’s visualization team and is available under the MIT License, making it free to use and modify.

Its WebGL-based rendering engine enables smooth rendering of datasets containing millions of data points, which is one of its standout features.

Key Features

Heatmaps

Heatmaps represent data point density through color gradients, making them ideal for understanding distribution patterns in large datasets. For example, displaying the publication locations of historical documents as a heatmap can instantly reveal areas of concentrated cultural activity.

3D Maps

Data can also be represented using the height dimension. Hexagonal bins displaying regional quantity data as 3D columns and 3D maps reflecting building heights are both possible.

Time-Series Animation

Data with timestamps can be played back as animations. Dynamically visualizing data changes over time provides an intuitive understanding of how historical events unfolded.

Diverse Layer Types

Kepler.gl supports a variety of layer types including points, arcs, lines, and polygons. Arc layers are suitable for visualizing migration routes and trade paths, while polygon layers work well for displaying regional boundaries.

Supported Data Formats

Kepler.gl supports the following data formats:

  • CSV — tabular data containing latitude and longitude columns
  • GeoJSON — the standard format for geospatial data
  • JSON — custom data formats

Files can be loaded by simply dragging and dropping them onto the browser, making it accessible even without programming knowledge.

Use Cases in DH

Historical Map Creation

Historical map data and event locations can be plotted and animated over time. For instance, one could visualize the distribution of post towns along the Tokaido Road during the Edo period or the expansion of railway networks during Japan’s modernization.

Cultural Heritage Mapping

Geographic information from museum, library, and archive collections can be visualized to analyze the spatial distribution of cultural heritage. Workflows such as extracting coordinates from IIIF manifest metadata and displaying them in Kepler.gl are also viable.

Integration with Text Mining

Place names extracted from texts can be geocoded and visualized in Kepler.gl, enabling spatial analysis of the geographic relationships found in literary works and historical documents.

Getting Started

Kepler.gl offers a web application (kepler.gl) that runs directly in the browser. Data upload, visualization, and configuration export can all be completed within the browser.

For more advanced use, Kepler.gl can be embedded in your own applications as a React component. It is available as an npm package, and a Jupyter Notebook plugin also exists.

npm install kepler.gl

Conclusion

Kepler.gl is a powerful tool for intuitively visualizing large-scale geospatial data. It can be used directly in the browser without programming knowledge — simply load a CSV or GeoJSON file to create beautiful maps. Consider incorporating it as an entry point for spatial analysis in your DH research.