A Computational Analysis of Conflict Dynamics and Ceasefire Resilience in South Sudan: A Network and Event Data Approach

This original research employs computational methods to analyse the structural and temporal dynamics of conflict in South Sudan, with a focus on the resilience of ceasefire agreements. Using a novel dataset synthesising ACLED and UNMISS reports from 2018 to 2023, we construct dynamic actor networks and apply survival analysis to model ceasefire failure. Results indicate that conflict networks exhibit scale-free properties and that ceasefire longevity is significantly predicted by the centrality of signatory actors and the co-occurrence of communal violence. The discussion situates these findings within the literature on hybrid peacemaking, arguing for a data-informed approach to sequencing and monitoring peace agreements. The conclusion outlines implications for third-party mediation and the design of more robust, locally-embedded ceasefire mechanisms.