Temporal Dynamics of Coordinated Online Behavior: Stability, Archetypes, and Influence

Abstract

Large-scale online campaigns, malicious or otherwise, require a significant degree of coordination among participants, which sparked interest in the study of coordinated online behavior. State-of-the-art methods for detecting coordinated behavior perform static analyses, disregarding the temporal dynamics of coordination. Here, we carry out the first dynamic analysis of coordinated behavior. To reach our goal we build a multiplex temporal network and we perform dynamic community detection to identify groups of users that exhibited coordinated behaviors in time. Thanks to our novel approach we find that: (i) coordinated communities feature variable degrees of temporal instability; (ii) dynamic analyses are needed to account for such instability, and results of static analyses can be unreliable and scarcely representative of unstable communities; (iii) some users exhibit distinct archetypal behaviors that have important practical implications; (iv) content and network characteristics contribute to explaining why users leave and join coordinated communities. Our results demonstrate the advantages of dynamic analyses and open up new directions of research on the unfolding of online debates, on the strategies of coordinated communities, and on the patterns of online influence.

Coordinated Behavior in Information Operations on Twitter

“…Similarly, as anticipated in Section VI-A, also the choice of the action with which to compute user similarities (e.g., co-retweets, as in our case) can deeply influence the shape and structure of coordination networks [3]. Here, we based the selection of the parameters of our method on the current best practices and on the latest results in the field [11], [64], [85]. In addition, in Section V-D we carried out extensive sensitivity analyses to assess the robustness of our results to small variations of the parameters in our datasets and method.…”

“…On the contrary, works that explicitly consider all instances of coordination, including spontaneous coordination by independent users, tend to use long time windows (e.g., some days or weeks) [9], [43]. Here, we set the time window length = 7 days, which allows evaluating both medium-and longterm interactions [64], as well as to model the activity of both malicious and genuine users in our networks. A second methodological choice regards the adoption of overlapping or non-overlapping time windows.…”

“…For example, this can be achieved by studying coordination with multi-layer networks, where each layer models one activity type [44]. Therefore, a promising direction for future work involves experimenting with rich and multidimensional user representations, such as multi-layer networks [44], [64], [72] or multivariate time series [73], to jointly model the multifaceted behavior of online users, thus surpassing one of the limitations of existing coordination detection methods that almost exclusively rely on unidimensional networks.…”

Tracking Fringe and Coordinated Activity on Twitter Leading Up to the US Capitol Attack

“…Second, our analyses focus on common manipulation tactics, but novel and more sophisticated strategies might have been employed (Luceri et al 2024;Cinelli et al 2022;Tardelli et al 2023). Similarly, we do not consider the activity of software-controlled accounts (i.e., bots) and statesponsored human operators (Luceri, Giordano, and Ferrara 2020;Ezzeddine et al 2023), who might have played a role in the amplification of fringe content (Ferrara et al 2020).…”

Clock against Chaos: Dynamic Assessment and Temporal Intervention in Reducing Misinformation Propagation

“…By capturing higher-order relations and scale-free characteristics, THINK outperformed state-of-the-art methods across various tasks. Tardelli et al (2023) presented the first dynamic analysis of coordinated online behavior by building a multiplex temporal network and employing dynamic community detection. In that direction, (Hristakieva et al 2022) investigated the interplay between propaganda and coordinated behavior in online debates, specifically focusing on the 2019 UK general election on Twitter.…”