Nadejda Komendantova from the Advanced Systems Analysis program is invited to give a keynote presentation on "Information technologies and their usage in social media" at the 8th Scientific Conference on Information Technologies for Intelligent Decision Making Support (ITIDS 2020). The conference is organized by the Ufa State Aviation Technical University, the North Caucasus Federal University, and the Ugra Research Institute of Information Technologies in cooperation with the Bashkir Branch of the Scientific Council of the Russian Academy of Sciences on the Methodology of Artificial Intelligence.
Information technologies and their usage in social media
Misinformation in social media is an actual and contested policy problem given its outreach and the variety of stakeholders involved. In particular, increased social media use makes the spread of misinformation almost ubiquitous. The presentation of Nadejda Komendantova on "Information technologies and their usage in social media" includes a framework for evaluating tools that detect misinformation using a preference elicitation approach as well as an integrated decision analytic process that will evaluate desirable features of systems to combat misinformation. The framework was tested in three countries (Austria, Greece, and Sweden) with three groups of stakeholders (policymakers, journalists, and citizens). Multi-criteria decision analysis (MCDA) is the methodological basis for this research. The results show that participants prioritized information regarding the actors behind the distribution of misinformation and tracing the lifecycle of misinformative posts. Another important criterion was whether someone intended to delude others, which shows a preference for trust, accountability, and quality in, for instance, journalism. It is also important how misinformation travels. However, all criteria that involved active contributions to dealing with misinformation were ranked low in importance, which shows that participants may not have felt personally involved enough in the subject or situation. The results also show that differences in preferences are explained by the cultural background of the participants, and that should perhaps be considered in the further development of tools.