Logo: to the web site of Uppsala University

This article presents a new, monthly updated dataset on organized violence—the Uppsala Conflict Data Program Candidate Events Dataset. It contains recent observations of candidate events, a majority of which are eventually included in the Uppsala Conflict Data Program Georeferenced Event Dataset as part of its annual update after a careful vetting process. We describe the definitions, sources and procedures employed to code the candidate events, and a set of issues that emerge when coding data on organized violence in near-real time. Together, the Uppsala Conflict Data Program Candidate and Georeferenced Event Datasets minimize an inherent trade-off between update speed and quality control. Having monthly updated conflict data is advantageous for users needing near-real time monitoring of violent situations and aiming to anticipate future developments. To demonstrate this, we show that including them in a conflict forecasting system yields distinct improvements in terms of predictive performance: Average precision increases by 20–40% relative to using the Uppsala Conflict Data Program Georeferenced Event Dataset only. We also show that to ensure quality and consistency, revisiting the initial coding making use of sources that become available later is absolutely necessary.

The proliferation of large-scale, geographically disaggregated data on armed conflicts, protests, and similar events has opened new avenues of research, but has also introduced significant data quality challenges. A notable yet often overlooked issue involves observations with “known geographic imprecision” (KGI), where event locations are unknown and instead arbitrarily assigned by dataset authors. Although this issue is widely recognized and accounts for up to a quarter of observations in datasets like UCDP GED, it is rarely addressed by users. This paper presents a stochastic method derived from the multiple-imputation literature, employing spatio-temporal Gaussian processes and leveraging latent actor-conflict features in the data to enhance location accuracy. Extensive Monte-Carlo simulations demonstrate that this approach substantially enhances the accuracy of these observations and improves predictive performance beyond the state-of-the-art when applied out-of-sample. Additionally, an adapted version of the UCDP GED dataset that employs this new procedure is provided, showcasing the practical application and benefits of the methodology.

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.

What explains conflict escalation during civil war? This article explores whether provocative attacks on religious sites and public transport constitute a precursor to a surge of violence. One argument pertains that the symbolic and doctrinal importance of places of worship means that attacks on these will affect individuals and the community emotionally and thereby increase the risk of escalation. However, it can also be suggested that the everyday societal importance of a public space is similar for religious sites and public transport hubs. We test these arguments using novel new global event data on these forms of selective targeting for 1989-2015, and find that the risk of conflict escalation increase in the aftermath of either attacks on places of worship or public transport, suggesting that community behavior is more affected to disruptions of societal everyday life than to the importance of symbols.

This study advances the field of conflict forecasting by using text-based actor embeddings with transformer models to predict dynamic changes in violent conflict patterns at the actor level. More specifically, we combine newswire texts with structured conflict event data and leverage recent advances in Natural Language Processing (NLP) techniques to forecast escalations and de-escalations among conflicting actors, such as governments, militias, separatist movements, and terrorists. This new approach accurately and promptly captures the inherently volatile patterns of violent conflicts, which existing methods have not been able to achieve. To create this framework, we began by curating and annotating a vast international newswire corpus, leveraging hand-labeled event data from the Uppsala Conflict Data Program. By using this hybrid dataset,  our models can incorporate the textual context of news sources along with the precision and detail of structured event data. This combination enables us to make both dynamic and granular predictions about conflict developments. We validate our approach through rigorous back-testing against historical events, demonstrating superior out-of-sample predictive power. We find that our approach is quite effective in identifying and predicting phases of conflict escalation and de-escalation, surpassing the capabilities of traditional models. By focusing on actor interactions, our explicit goal is to provide actionable insights to policymakers, humanitarian organizations, and peacekeeping operations in order to enable targeted and effective intervention strategies.

Temporally and spatial disaggregated datasets are commonly used to study political violence. Researchers are increasingly studying the data generation process itself to understand the selection processes by which conflict events are included in conflict datasets. This work has focused on conflict fatalities. In this research note, we explore how non-fatal conflict events are reported upon and enter into datasets of armed conflict. To do so, we compare reported non-fatal conflict events with the population of events in two direct observation datasets, collected using a boots-on-the-ground strategy: mass abductions in Nepal (1996-2006) and troop movements in Darfur. We show that at the appropriate level of aggregation media reporting on abductions in Nepal largely mirrors the "true" population of abductions, but at more disaggregated levels of temporal or spatial analysis, the match is poor. We also show that there is no overlap between a media-driven conflict dataset and directly-observed data on troop movements in Sudan. These empirics indicate that non-fatal data can suffer from serious underreporting and that this is particularly the case for events lacking elements of coercion. These findings are indicative of selection problems in regards to the reporting on non-fatal conflict events.