Politicized Netizens on VKontakte: Factions and Their Influence

This study aims to map politicized online communities (groups) on VKontakte, a social networking service, and to identify their clusters – ideologically related and/or tightly connected factions – while measuring the online influence of these communities and factions. The authors propose a toolkit that enables a quantitative assessment of user engagement within these online communities, including individual’s openness to group opinions, willingness to change their ideas, or even take actions influenced by the group. The toolkit is grounded in the theory of self-organized criticality (SOC), which is an interdisciplinary concept developed within the framework of exact and natural sciences. The chronological scope of the research spans from May 31, 2021 to May 31, 2024. The study involves nearly 400 connected politicized communities, identified through an analysis of approximately 40,000 groups. A shared number of members among communities served as a criterion of establishing associations between them. The mapped network is presented as a graph, visualized using Gephi software with the ForceAtlas2 layout algorithm. All communities were categorized using high-quality coding that is continuously updated. This approach indicates the political orientation of each group, and provides insights into the assortment of political factions on the Internet. Sufficiently long time series were obtained for each community, reflecting user activity in disseminating messages posted by the community. Spectral analysis of these series revealed instances of pink noise – an attribute associated with SOC. It is possible to specify explanatory frameworks based on SOC theory, so that communities in a SOC state can be regarded as highly engaged. For such communities, an online influence index was calculated; for factions, a normalized index of generalized online influence (GOI) was determined for the past year (May 31, 2023 – May 31, 2024). The analysis established that the conglomerate of patriotic propresidential factions has the highest aggregate GOI index (50.85 out of 100 units). Next the conglomerate of nationalist factions (GOI=15.31) appeared, then liberal (GOI=8.89) and left factions (GOI=0.63) among others. Calculating the influence of factions and visualizing their relative position based on ideological attraction/repulsion enabled us to reconstruct the landscape of ideas within the politicized segment of the Internet.

Keywords: social networks; politicized communities; socio-political ideas; theory of selforganized criticality; pink noise; network activity; network analysis