Digital Publics and the Ukraine Dilemma: Topic Modelling of the Cumulative Twitter Discussion

In this study, we explore on what topics and to what extent digital publics - publics that exist on Twitter and share common topics - contributed to the global intensifying information warfare about Ukraine in the world. We use probabilistic topic modelling with time series for 3,676,245 unique tweets with the keyword 'Ukraine' or 100 political or regional hashtags in English or Russian written by 960,422 unique users for the period from 30 August 2021 to 24 February 2022. We reveal 38 politically significant topics (23 persistent topics and 15 event topics) and explore the scope of discussion and its dependency on political events. The application of SNP metric to tweets by topics allow us to carefully study and then describe the political ideas and arguments offered by influential ordinary Twitter users in online information confrontation. We demonstrate the process of cumulative formation of global clusters of digital publics on some important topics with regard to the approaching escalation of the Russian-Ukrainian conflict.