Large Language Models and Generative AI for Conflict Research
November 13, 2025 (This event has concluded)
Workshop Program
| Morning Session |
|---|
| 9:00 – 9:30 am “Introduction to ConfliBERT with Applications to the Militarized Interstate Dispute Project” (Vito D’Orazio, West Virginia University) |
| 9:30 – 10:15 am “A Technical Overview of ConfliBERT” (Javier Osorio, University of Arizona) |
| 0:30 – 12:00 am “ConfliBERT Training tutorial” (Vito D’Orazio, Dagmar Heintze, Javier Osorio, Emmanuel Teitelbaum) Finetuning with binary classification |
| Afternoon Session 1 |
|---|
| 1:00-1:30 pm “Extracting Locations and Event Counts from Conflict Reports: The Sequence-to-Sequence Advantage” (Emmanuel Teitelbaum, George Washington University; Mohiddin Bacha Shaik, George Washington University) |
| 1:30 – 2:00 pm “(Semi)independent Human Rights Advocates: Extracting National Prevention Mechanism Empowerment Indicators Through LLMs” (Dagmar Heintze, UT Dallas) |
| 2:00 – 2:30 pm “Measuring Intensity Levels of Repression Using a Large Language Models and a Latent Variable Approach” (Rebecca Cordell, Pittsburgh; Chris Fariss, Michigan; Chad Clay, Georgia; Reed Wood, Essex; Thorin Wright, Arizona State) |
| Afternoon Session 2 |
|---|
| 2:45 – 3:15 pm “Networks of Influence and Support Dataset: Leveraging Large Language Models for Organization Classification in Peace Operations.” (Susanna Campbell, American University; Madeline Fleishman, University of Virginia; Cecilia Cavero-Sanchez, Penn State) |
| 3:15 – 3:45 pm “From Scarcity to Signal: Synthetic Data Enables Scalable Detection of Rare Violence Events in Text” (Simon Polichinel von der Maase, PRIO; Sonja Häffner, PRIO) |
| 3:45 – 4:15 pm “Political Event Coding as Text-to-Text Sequence Generation with Synthetic Training Data” (Ben Radford, UNC Charlotte; Yaoyaho Dai, UNC Charlotte; Andy Halterman, Michigan State) |
Sponsors
This workshop is sponsored by the Peace Science Society International Annual Meeting 2025 Peace Science Society (International). The organizers also count the support from the National Science Foundation (NSF) under award 2311142.