Sub project 1: Digital disinformation networks – University of Copenhagen

Sub project 1: Digital disinformation networks


The task of tracking digital disinformation across a range of different platforms and in different languages requires robust, cross-lingual document similarity estimation, extraction of propositions and, potentially, fraud detection. All three are known problems in natural language processing (NLP), but we believe that bringing all three together, exploiting the synergies between them, will not only push computational social science research forward but also expand research horizons in the social sciences. After networks are identified, we map the network of professional disinformers (trolls) and investigate the use of nonhuman actors (bots) in disinformation.