Propaganda Detection in Russian Telegram Posts in the Scope of the Russian Invasion of Ukraine

Proceedings of Recent Advances in Natural Language Processing, pages 1162–1170 Varna, Sep 4–6, 2023 https://doi.org/10.26615/978-954-452-092-2_123 1162 Propaganda Detection in Russian Telegram Posts in the Scope of the Russian Invasion of Ukraine Natalia Vanetik 1 natalyav@sce.ac.il Marina Litvak 1 marinal@sce.ac.il Egor Reviakin 1 igorre@ac.sce.ac.il Margarita Tyamanova 1 margati@ac.sce.ac.il 1Shamoon College of Engineering, Beer-Sheva, Israel Abstract The emergence of social media has made it more difficult to recognize and analyze mis- information efforts. Popular messaging soft- ware Telegram (Durov, 2013) has developed into a medium for disseminating political mes- sages and misinformation, particularly in light of the conflict in Ukraine (Wikipedia contrib- utors, 2023). In this paper, we introduce a siz- able corpus of Telegram posts containing pro- Russian propaganda and benign political texts. We evaluate the corpus by applying natural lan- guage processing (NLP) techniques to the task of text classification in this corpus. Our find- ings indicate that, with an overall accuracy of over 96% for confirmed sources as propagan- dists and oppositions and 92% for unconfirmed sources, our method can successfully identify and categorize pro-Russian propaganda posts. We highlight the consequences of our research for comprehending political communications and propaganda on social media. 1 Introduction Because of social networks’ rising use in daily life, we increasingly rely on other people’s opinions when making both big and minor decisions, such as whether to vote for a new government or buy new products online. It is not surprising that by spread- ing propaganda, social media became a weapon of choice for manipulating public opinion. Social media is rife with fake content and propaganda, which needs to be identified and blocked or re- moved. Recent years have seen a major increase in the issue of information authenticity on social media, leading to significant research community efforts to address fake news (Pariser, 2011), click- bait (Chen et al., 2015b), fake reviews (Akoglu et al., 2013), rumors (Hamidian and Diab, 2016), and other types of misinformation. In this paper, we deal with Russian state-sponsored propaganda disseminated in Telegram. Telegram is one of the most widely used venues for information sharing in Russia, especially after blocking META Plat- forms. Therefore, Telegram draws much attention from organized groups that spread similar views through its channels and (most probably) funded by either state or related organizational sources. To influence the public to favor the war, the Russian government implemented new regulations that gave it control over traditional media channels (Geissler et al., 2022). The fundamentals of propaganda communication: persuasion using symbols, emo- tions, stereotypes, and pre-existing frameworks with the intention of swaying perceptions and influ- encing cognition and behavior in order to further the propagandist’s agenda (Alieva et al., 2022). Our work focuses on specific pro-Russian propaganda during the conflict between Russia and Ukraine. Several researchers have documented Russian pro- paganda during previous conflicts (Golovchenko, 2020; Geissler et al., 2022). This paper has two contributions: (1) it intro- duces a new dataset of posts about the Russia- Ukraine war in Russian, collected from Telegram channels and annotated …

Content truncated to 500 words