PoliLoom: Enriching Wikidata Politician Records with AI-Assisted Extraction

We are pleased to announce that PoliLoom is now available for community testing and feedback.

Overview

PoliLoom (https://loom.everypolitician.org/) is an AI-powered data extraction tool designed to identify and extract politician metadata from Wikipedia and other web sources to enrich Wikidata records.

The tool systematically extracts structured information including birth dates, birthplaces, political positions held, and tenure periods. It makes use of the existing Wikipedia / Wikidata authentication system to allow you to update Wikidata without most of the manual labor.

Community Feedback Requested

As we continue development, we are seeking input from the OpenSanctions and Wikidata communities on several key areas:

  • Extraction accuracy and data quality
  • User interface and verification workflow
  • Additional properties or data types that would be valuable to extract
  • Edge cases, errors, or limitations you encounter during testing

Get Involved

All you need is a Wiki account, if you don’t have one you can create one.
:link: Start evaluating: https://loom.everypolitician.org/

The project is open source and available for review and testing:
:link: GitHub Repository: GitHub - opensanctions/poliloom: Loom for weaving politician's data

We welcome contributions from data practitioners, Wikidata editors, and anyone interested in improving the quality and completeness of structured political data. Your feedback will directly inform ongoing development and help refine the tool’s capabilities.

Please share your experiences, technical observations, and suggestions in this thread or via the project’s GitHub issues.

2 Likes

Congratulations, @Monneyboi ! It’s such a nice tool, I really look forward to seeing people’s feedback :slight_smile: