June 10, 11.00–12.30
Room: Europa
Chaired by
Steven Wilson
Associate Professor
Brandeis University
Presenter
Therese Pettersson
Research Coordinator, Uppsala University
Title
Classifying fatalities: Challenges in Identifying Civilian and Combatant Victims in Organized Violence
Citation
Davies, Shawn, Therese Pettersson, Margareta Sollenberg & Magnus Öberg (2025). Organized violence 1989-2024, and the challenges of identifying civilian victims. Journal of Peace Research (Forthcoming)
The thematic section of the UCDP's forthcoming annual article at the Journal of Peace Research examines the challenges of classifying fatalities in organized violence. Drawing on historical and contemporary conflicts, the contributions highlight how information environments, data sources, and classification practices shape our understanding of victim categories. The article demonstrates that while accurate fatality classification is essential for understanding patterns of organized violence, it is often constrained by access limitations, conflict characteristics, and reporting infrastructures.
Presenter
Håvard Hegre
Professor, Uppsala University
Title
Forecasting Intensity of Armed Conflict with Uncertainty: The Role of Input Data
Presenter
Chandler Williams
Doctoral Researcher, The Peace Research Institute Oslo (PRIO)
Title
Predicting the Present: Nowcasting Wartime Fatalities to Improve Conflict Monitoring
Presenter
Mihai Croicu
Doctoral Researcher, Uppsala University
Title
Deep Active Learning for Data Mining from Conflict Text Corpora
High-resolution event data on armed conflict and related processes have revolutionized the study of political contention. However, most datasets of this type only collect spatio-temporal and conflict intensity data at that level of detail. Information on dynamics, such as targets, tactics, and purposes, is rarely collected due to the substantial effort of collecting data. This study proposes an inexpensive, high-performance approach to increase the feature richness of such datasets by leveraging active learning -- an iterative process of improving a machine learning model based on guided human input at each step of the learning process. Active learning is employed to then fine-tune (train in steps) a large, encoder-only language model fitted to the rich corpus of textual data underlying such datasets. This allows for the extraction of features related to conflict dynamics, such as electoral violence and attacks on religious targets. The approach achieves a performance comparable to the human (gold-standard) coding, while reducing the necessary human annotation by as much as 99 percent.
Demscore Conference 2025