Paper Explores Change Point Detection and Multi-Level Analysis

Researchers Swetha Keertipati, Bastin Tony Roy Savarimuthu, Maryam Purvis, and Martin Purvis, all of the University of Otago in Dunedin, New Zealand, explore the application of change point detection for the automatic identification of major emergent situations captured in GDELT and propose a conceptual model of a multi-level analytic dashboard that would advance such indicators into actionable insights for stakeholders.  The authors propose two research questions regarding whether organizations can "use this almost real-time data as a reasonably reliable source to observe trends in global conflicts?" and "use the data obtained to identify and monitor for certain types of events (e.g. events that are significant)?", finding that "GDELT indeed captures the global trends as we could visualize and understand how the data narrates the story of Sri Lankan civil war and the Fijian coup from New Zealand’s perspective within a given time period" and "using change point analysis we are able to identify significant points along the course of conflicting events."

Read the Full Paper.