Social media can have an impressive impact on civic engagement and political discourse. Yet increasingly we find political actors using digital media and automated scripts for social control. Computational propaganda—through bots, botnets, and algorithms—has become one of the most concerning impacts of technology innovation. Unfortunately, bot identification and impact analysis are among the most difficult research challenges facing the social and computer sciences. COMPROP objectives are to advance a) rigorous social and computer science on bot use, b) critical theory on digital manipulation and political outcomes, c) our understanding of how social media propaganda impacts social movement organization and vitality. This project will innovate through i) “real-time” social and information science actively disseminated to journalists, researchers, policy experts and the interested public, ii) the first detailed data set of political bot activity, iii) a deepened regional expert network able to detect bots and their impact in Europe. COMPROP will achieve this through multi-method and reflexive work packages: 1) international qualitative fieldwork with teams of bot makers and computer scientists working to detect bots; 2a) construction of an original event data set of incidents of political bot use and 2b) treatment of the data set with fuzzy set and traditional statistics; 3) computational theory for detecting political bots and 4) a sustained dissemination strategy. This project will employ state-of-the-art “network ethnography” techniques, use the latest fuzzy set / qualitative comparative statistics, and advance computational theory on bot detection via cutting-edge algorithmic work enhanced by new crowd-sourcing techniques. Political bots are already being deployed over social networks in Europe. COMPROP will put the best methods in social and computer science to work on the size of the problem and the possible solutions.
Professor Philip Howard, University of Oxford
Funder: European Research Council
Proposal Number: 648311