State sponsored information operations, or SSIOs, are a growing problem across many of the information spaces we inhabit online. These instances of coordinated misinformation and propaganda have been perpetrated by over 80 state actors in the last decade, and have been used to exert influence on digital media consumption habits, discussions of contentious issues, and even national elections. Concern over the power that SSIOs wield is only growing as the proliferation of automated tools and services is making it easier than ever to launch large-scale manipulation campaigns. But what role do such automated agents play within the broader operations that they are deployed in? Are they even successful at making an impact in information spaces online? In this work, we address both of these questions through the use of a sequence-based clustering method and advanced linear modeling. Using these methods, we investigate the relationship between agent automation, role, and network characteristics and how much success those agent's achieve over the course of their lifetimes. We find that automated agents perform worse across every success metric compared to human agents, and that they play a smaller, supporting role to the primarily human SSIO workforce. What's more, we find that the extent to which agent's engage in amplifying- or producing-centric roles is by far the biggest determinant of how successful they will be, highlighting the importance of social-roles in the analysis of automated agents.

In January 2014, what started as peaceful protests in Kyiv, Ukraine escalated into an ousting of the Ukrainian president, then annexation of Crimea, and a protracted war between the new Ukrainian government and Russian-backed separatists. Meanwhile, Ukrainian citizens documented their experiences on Twitter, leading to the creation of multitudes of conversation spaces, or online publics, that evolved alongside the conflict they were discussing. We collected tweets in those conversations from December 2013 to August 2015. In this work, we use a novel, LDA-based topic tracking tool to analyze how the conversations evolved during the crisis in Ukraine, and the factors associated with this volatility. As existing tools often do not support researchers in pinpointing when online publics shift from one conversation to another, we discuss how our tool can be used to extract cleaner data from longitudinal crises. Finally, we discuss how our discursive analysis of Ukraine in 2014 provides insight on the current conflict in Ukraine.

Data visualizations can empower an audience to make informed decisions. At the same time, deceptive representations of data can lead to inaccurate interpretations while still providing an illusion of data-driven insights. Existing research on misleading visualizations primarily focuses on examples of charts and techniques previously reported to be deceptive. These approaches do not necessarily describe how charts mislead the general population in practice. We instead present an analysis of data visualizations found in a real-world discourse of a significant global event—Twitter posts with visualizations related to the COVID-19 pandemic. Our work shows that, contrary to conventional wisdom, violations of visualization design guidelines are not the dominant way people mislead with charts. Specifically, they do not disproportionately lead to reasoning errors in posters’ arguments. Through a series of examples, we present common reasoning errors and discuss how even faithfully plotted data visualizations can be used to support misinformation.

Unlike most other forms of coordinated, inauthentic behavior occurring online, the goals of state-sponsored information operations, or SSIOs, are often complex and multifaceted. These goals range from flooding conversations with a certain narrative, to increasing the public's engagement with news sources of questionable quality, to stoking tensions between ideologically opposed groups to weaken public trust. The prevailing theoretical framework for understanding SSIOs is to treat them as a social botnet: a behaviorally homogeneous cluster of coordinated activity. However, the social bot framework is both at odds with some of the behaviors observed in early SSIOs and more broadly with the wide swathe of goals these operations set out to accomplish. To examine the fit of the social bot framework in the SSIO context, we develop a novel bag-of-words based method for clustering and describing user activity traces. Applying this method to a comprehensive repository of SSIOs conducted on Twitter over the last decade, we find that SSIOs violate both the core assumption of the social bot framework, and how it is operationalized in practical work. Instead, we find that SSIOs exhibit a clear division of labor and propose cooperative work with social roles as a more effective theoretical framework for understanding SSIOs. Through applying this framework, we find that the roles that SSIO agents take on have become more stable and simple over time, which holds substantial implications for developing methods for detection of these operations in the wild.