This allows Aos Fatos to map out the flow of disinformation and make predictions.
Researchers feed linguistic clues into the algorithmic tools, and they surface potential instances of disinformation. These posts are rated numerically based on what signals they include which are related to known false narratives, with scores under 5 likely to be disinformation. This process generates data about terms associated with disinformation, networks of disinformation, trending narratives, and the use of and collaboration between different platforms.