In this paper we present an analysis of misinformation cross-platform dynamics by focusing on communications published by COVID19 negationists on Twitter and Telegram. Previous research shows the need for better explanations of the way misinformation travels across platforms. Here we pay attention to communities of users vulnerable to negationism, which refers to the tendency to revise history in order to omit something that actually happened. We start from searching specific key words previously identified by experts and used by negationists in Telegram channels. We retain only those public Telegram channels where those keywords are used. Then, we search on Twitter for those users who reference those Telegram channels. This way we obtain a list of potential Twitter negationist accounts and correct for false positives. We use the normalised compression distance (NCD) technique to reduce this error, while performing authorship attribution. We extract images and news domains shared by Twitter negationist accounts identified with NCD and by Telegram accounts initially identified; then we perform a reverse image search to identify other Twitter accounts that have used those images. We search in Telegram where those news domains appear and extract those Telegram channels and compare them with the original channels identified. The procedure is semi-automatic to ensure human supervision as required by The Assessment List on Trustworthy Artificial Intelligence (ALTAI). As discussed in the end, results are promising and motivate further research about the use of NCD to automate the identification of accounts spreading misinformation.
This research was funded by the projects (i) CYNAMON - Cybersecurity, Network Analysis and Monitoring for the Next Generation Internet” (funded by the Madrid Region under “Programas de Actividades de I+D entre grupos de investigación de la Comunidad de Madrid en tecnologías 2018” - P2018/TCS-4566; BOCM. No. 304; 21/12/2018) and (ii) XAI-DisInfodemics - eXplainable AI for disinformation and conspiracy detection during infodemics (Grant PLEC2021-007681 funded by MCIN/AEI/ 10.13039/501100011033 and by European Union NextGenerationEU/PRTR).