A web-based system for taxonomy labeling and comparison of old visualization images.
You can find a live demo here with taxonomy labeling and taxonomy comparison. The functions related to the server are disabled in the live demo. To use the full-fledged version, please follow the instructions in How to Use.
This repository is structured as:
./apps/label: The interface for taxonomy labeling../apps/compare: The interface for taxonomy comparison../website: The GitHub Pages landing that links to both apps../packages/shared: Shared leaf modules (@image-taxonomy-labeler/shared) — API helpers and the visualization catalog../packages/ui: Shared Vue helpers and chrome (@image-taxonomy-labeler/ui) — label-task composables,VImage, dialogs, pagination, styles../scripts: The scripts for one-time data preprocessing../server: The server shared by the apps for serving image resource and computation services.
From the repo root: pnpm install, then pnpm start:label (or pnpm start:compare).
To preview the GitHub Pages site locally (landing + both apps): pnpm --filter ./website preview.
Warning
The following instructions for setting up the interfaces assume the annotated dataset is oldvis/dataset. If you want to use the interfaces to annotate other images datasets, you will need to
- ensure the image metadata matches the structure of visualizations.json and replace the metadata at
./packages/shared/assets/visualizations.json - store the images at
./server/static/images/ - setup the cache files (with
./scripts/setup_cache.py)
Use the startup script (assume Node.js and Python are installed):
pnpm start:label- Setup the server resources and launch the server (see details at ./server/README.md).
- Start only the client:
pnpm --filter ./apps/label dev(environment flags: ./apps/label/README.md).
Use the startup script (assume Node.js and Python are installed):
pnpm start:compare- Setup the server resources and launch the server (see details at ./server/README.md).
- Start only the client:
pnpm --filter ./apps/compare dev(environment flags: ./apps/compare/README.md).
If you use this repository in a scientific publication, we would appreciate citations to the following paper:
@Article{Zhang2025VisTaxa,
author = {Zhang, Yu and Chen, Xinyue and Zheng, Weili and Guo, Yuhan and Li, Guozheng and Chen, Siming and Yuan, Xiaoru},
title = {{VisTaxa}: Developing a Taxonomy of Historical Visualizations},
doi = {10.1109/TVCG.2025.3567132},
number = {6},
pages = {3850--3862},
volume = {31},
journal = {IEEE Transactions on Visualization and Computer Graphics},
publisher = {IEEE},
year = {2025},
}The steps to reproduce Figure 2 of the paper "VisTaxa: Developing a Taxonomy of Historical Visualizations" are available at graphics-replicability.md.