Principal Data Scientist · ACLED

Trey
Billing

Trey Billing is Principal Data Scientist at ACLED, where he applies machine learning and computational methods to the study of armed conflict and political instability. He holds a PhD in Government and Politics from the University of Maryland and previously held a postdoctoral fellowship at Ohio State's Mershon Center for International Security Studies. His research spans conflict prediction, food security, and computational social science, and has been published in outlets including the American Sociological Review, The Lancet Planetary Health, and the Journal of Conflict Resolution.

01

Selected Research

  1. 01
    LEMONADE: A large multilingual expert-annotated abstractive event dataset for the real world

    with Sina Semnani, Pingyue Zhang, Wanyue Zhai, Haozhuo Li, Ryan Beauchamp, Katayoun Kishi, Manling Li, and Monica Lam — Findings of the Association for Computational Linguistics, 2025

  2. 02
    How Radio Affects Violent Conflict: New Evidence from Rwanda

    with Hollie Nyseth Nzitatira and Jared Edgerton — American Sociological Review, 2024

    Co-winner, Best Paper — ASA Section on Sociology of Human Rights
  3. 03
  4. 04
  5. 05
02

Current Projects

Conflict Prediction

Developing forecasting models that leverage ACLED's conflict event data alongside external indicators to anticipate the onset, escalation, and spread of political violence. Work here connects machine learning methods to policy-relevant early warning.

AI, Text Modeling & Conflict

Applying large language models and NLP techniques to the structured extraction, classification, and analysis of conflict-relevant text. Projects include dataset construction, event coding automation, and leveraging agents in the analysis of conflict dynamics.