Constitutional systems
189
Systems with data in the current public index.
Official project information
Press / About / Methodology
Use this page as the canonical citation target for summaries of Constitutional Map AI.
Constitutional Map AI is an interactive semantic atlas of comparative constitutional law, with 189 constitutional systems, more than 30,000 legal segments, textual search, semantic search, and 3D visualization. The code is open source and the constitutional texts are derived from the Constitute Project, according to the applicable license.
Dataset snapshot
Current numbers
2026-04-02
Legal segments
30,828
Article-like constitutional passages in the current snapshot.
Global clusters
509
HDBSCAN global cluster summaries exported for the Atlas.
Data updated
2026-04-02
Generated from app/public/data/index.json.
Data source and license
Constitutional texts are derived from the Constitute Project and are used under the Creative Commons Attribution-NonCommercial 3.0 Unported License (CC BY-NC 3.0). Attribution to the Constitute Project is required.
The application and pipeline code are open source under the MIT License. The code license does not override the separate license that applies to constitutional texts.
Methodology
The Atlas is designed as an exploratory research interface, not as an authority on the legal meaning of any constitution.
- 1Constitutional texts are collected from the Constitute Project and normalized into repository-backed data exports.
- 2Texts are segmented into legal units, usually articles or article-like provisions, so search and visualization operate on comparable passages.
- 3Each segment is embedded as a semantic vector using Gemini embeddings, turning legal language into a numerical representation of meaning.
- 4UMAP is used to project high-dimensional embeddings into spaces suitable for clustering and 3D visualization. HDBSCAN groups nearby segments into global thematic clusters and marks noisy or ambiguous points where appropriate.
- 5The web app combines textual search, semantic search, country selection, cluster coloring, and a 3D map so users can move from a visual pattern to the underlying constitutional text.
What the project is
- An open-source semantic atlas for comparative constitutional law.
- A research and teaching interface for exploring constitutional texts across countries.
- A tool that combines textual search, semantic search, clustering, and 3D visualization.
- A citation-friendly project page for journalists, teachers, aggregators, and LLMs.
What the project is not
- It is not legal advice and does not replace jurisdiction-specific legal analysis.
- It is not an official or authoritative constitutional database.
- It is not a claim that semantic clusters prove doctrinal identity between legal systems.
- It is not currently described by subscription tiers, paid plans, or a limited free tier.
Limitations
Semantic proximity is not legal equivalence. Nearby passages can perform different legal functions once doctrine, institutional context, or translation choices are considered.
Clusters are computational aids. Cluster labels, probabilities, and 3D distances should be treated as starting points for reading, not as legal conclusions.
The corpus depends on Constitute Project texts and may lag constitutional amendments, official translations, or jurisdiction-specific publication sources.
The project can contain segmentation, clustering, search, or content errors. Verify high-stakes claims against official constitutional sources.
Official images and video
These assets are stable public URLs for previews, articles, classrooms, and summaries. Use the license notes above when constitutional text is visible or quoted.
Official overview image
A shareable image of the Constitutional Map AI interface for previews, articles, and teaching material.
/og-image.png
3D Atlas walkthrough
A short official video showing the interactive 3D constitutional map in use.
/media/3d-constitutional-map.mp4
Comparative constitutional law preview
A static preview of a reproducible comparative constitutional law view.
/discovery/en-comparative-constitutional-law-ai.png
Contact
Constitutional Map AI
For corrections, methodology questions, press references, or contribution discussions, use GitHub or contact the maintainer directly.