Machine Learning for Peace:

Crisis Event Detection and Forecasting

PDRI-DevLab

University of Pennsylvania

November 12, 2023

Principal Investigator: Erik Wibbels

Director of Civic Research: Jeremy Springman

Data Scientists: Zung-Ru Lin, Hanling Su

Affiliates: Serkant Adiguzel, Mateo Villamizar Chaparro, Diego Romero, Rethis Togbedji Gansey, Jitender Swami

MLP: Approach


How can data contribute to the defense of human rights?

  1. Awareness: data on what’s happening very recently
    • Mass scraping online news + ML to track events
    • Interactive data dashboards

  2. Planning: predictive analytics for strategic decisions
    • Forecasting political events
    • Civic Space Early Warning System

Data Production

Input: Online news

  • 300+ news sources
  • 34 languages
  • approx 100 million articles

Data quality

  • Focus on reputable local sources
  • Much better coverage than other archives/aggregators (GDELT, LexisNexis, etc.)




Output: Monthly data

  • 56 countries
  • 2012 - last month

MLP: Digital Tools

MLP: Digital Tools

MLP: Digital Tools

Awareness: Civic event detection

Uganda: Arrests

Planning: Forecasting events

Planning: Forecasting events

Legal Changes:

  • Senegal (July)
  • Georgia (Feb)

Arrests:

  • Kosovo (May)
  • Nicaragua (Apr)

Security mobilization:

  • Philippines (Apr)

Legal Actions:

  • Uzbekistan (Jul)
  • Kosovo (Jul)
  • Zambia (Jun)

Protests:

  • India (Mar)

Non-lethal Violence:

  • Guatemala (Jun)

Applications & Extensions

Current applications

  • Directing Flexible Response Funds (INSPIRES Consortium)
  • Distribution to local civil society organizations
    • 1,800+ site users across 99 countries; mailing list of 500+

Extensions: Flexible data production infrastructure

  • Unique local news corpus to track + forecast new events
    • Climate-human interactions
    • Media polarization, threatening language

mlpeace.org