Africa remains one of the world’s most dynamic regions, offering vast economic opportunities while facing persistent risks of violent conflict, political instability and security threats. Early warning has moved from expert judgement to data-driven forecasting: they perform best where violence is already under way. Sudden escalation remains the hard case, where even sophisticated models are still caught by surprise. Combining conflict modelling, machine learning and media analysis, CYSAN starts from one research question: can we anticipate how severe a conflict will become, early enough for prevention to act?
CYSAN, a Canadian startup based in Gatineau, has developed a proof-of-concept early warning system that monitors conflict in Africa using Tableau dashboards and publicly available conflict event data, and is designed to adapt to other regions. CYSAN’s co-founders will present the prototype, then the conceptual framework of p-CSI (Predictive Conflict Severity Index). The framework aims to forecast the severity of conflict, not only the number of events, by combining four dimensions of severity with signals from the media environment to estimate, region by region, the probability of escalation.
The session aims to:
- introduce the prototype and the p-CSI framework to the security and conflict prevention research community and practitioners;
- invite researchers to critically assess the expected value of p-CSI, framed as hypotheses to be tested;
- gather feedback on empirical validation, data limitations and ethical issues;
- prepare a follow-up session demonstrating the p-CSI proof of concept.
Speakers
- Étienne Emboussi, Eng., M.Sc., President and Co-founder, CYSAN
- Stéphane Akaba, Vice-President and Co-founder, CYSAN; Part-time Professor and Doctoral Candidate, University of Ottawa
Comments are closed.