Predictive Research On Misinformation and narratives Propagation Trajectories

Fighting disinformation remains challenging, yet targeting misleading narratives offers hope. Our proposal advances the European Narratives Observatory's work, using innovative methods to counteract misinformation. We leverage advanced Large Language Models (LLM) to analyze vast textual data, pinpointing subtle misleading narratives overlooked by older tools. Our approach combines established techniques like dynamic network analysis with LLM's capabilities, promising efficient yet frugal deployment across varied sources and languages. We focus on key informational nodes, like media and Wikipedia, realizing their significant influence on public opinion. Deliverables include a refined language model, narrative monitoring system, and three in-depth case studies on the Ukraine conflict, LGBTQI+ narratives, and the 2024 European elections in France, Baltic Region, Italy and Romania. We'll produce 36 journalistic pieces showcasing our techniques and launch a pioneering MOOC for journalists.