Now showing 1 - 4 of 4
No Thumbnail Available
Publication

This Account Doesn't Exist: Tweet Decay and the Politics of Deletion in the Brexit Debate

2021-05-01, Bastos, Marco

Literature on influence operations has identified metrics that are indicative of social media manipulation, but few studies have explored the lifecycle of low-quality information. We contribute to this literature by reconstructing nearly 3M messages posted by 1M users in the last days of the Brexit referendum campaign. While previous studies have found that on average only 4% of tweets disappear, we found that 33% of the tweets leading up to the referendum vote are no longer available. Only about half of the most active accounts that tweeted the referendum continue to operate publicly and 20% of all accounts are no longer active. We tested whether partisan content was more likely to disappear and found more messages from the Leave campaign that disappeared than the entire universe of tweets affiliated with the Remain campaign. We compare these results with an assorted set of 45 hashtags posted in the same period and find that political campaigns present much higher ratios of user and tweet decay. These results are validated by inspecting 2M Brexit-related tweets posted over a period of nearly 4 years. The article concludes with an overview of these findings and recommendations for future research.

No Thumbnail Available
Publication

From Global Village to Identity Tribes: Context Collapse and the Darkest Timeline

2021-07-23, Bastos, Marco

In this article we trace the development of two narratives describing social media that informed much of internet scholarship. One draws from McLuhan’s axiom positing that communication networks would bring forth a ‘global village,’ a deliberate contradiction in terms to foreground the seamless integration of villages into a global community. Social media would shrink the world and reshape it into a village by moving information instantaneously from any location at any time. By leveraging network technology, it would further increase the density of connections within and across social communities, thereby integrating geographic and cultural areas into a village stretching across the globe. The second narrative comprises a set of metaphors equally inspired by geography but emphasizing instead identity and tribalism as opposed to integration and cooperation. Both narratives are spatially inspired and foreground real-world consequences, either by supporting cooperation or by ripping apart the fabric of society. They nonetheless offer opposing accounts of communication networks: the first is centered on communication and collaboration, and the second highlights polarization and division. The article traces the theoretical and technological developments driving these competing narratives and argues that a digitally enabled global village society may in fact reinforce intergroup boundaries and outgroup stereotyping typical of geographically situated communities.

No Thumbnail Available
Publication

The IMPED Model: Detecting Low-Quality Information in Social Media

2021-05-01, Bastos, Marco, Walker, Shawn, Simeone, Michael

This paper introduces a model for detecting low-quality information we refer to as the Index of Measured-diversity, Partisan-certainty, Ephemerality, and Domain (IMPED). The model purports that low-quality information is characterized by ephemerality, as opposed to quality content that is designed for permanence. The IMPED model leverages linguistic and temporal patterns in the content of social media messages and linked webpages to estimate a parametric survival model and the likelihood the content will be removed from the Internet. We review the limitations of current approaches for the detection of problematic content, including misinformation and false news, which are largely based on fact-checking and machine learning, and detail the requirements for a successful implementation of the IMPED model. The paper concludes with a review of examples taken from the 2018 election cycle and the performance of the model in identifying low-quality information as a proxy for problematic content.

No Thumbnail Available
Publication

Guy Next Door and Implausibly Attractive Young Women: The Visual Frames of Social Media Propaganda

2023-08, Bastos, Marco, Mercea, Dan, Goveia, Fábio

This study introduces a mixed-methods approach to classifying the visual frames of state-sponsored social media propaganda. We relied on Twitter’s Election Integrity data to sample five key propaganda targets of the Internet Research Agency (IRA), including Russian and American partisan groups. We manually coded profile images and subsequently applied qualitative and quantitative processing to the images. The visual motifs identified in IRA Twitter profiles allowed us to explore how their operations deviated from canonical state propaganda marked by symbols of national identify and heroic masculinity. Indeed, the results show that the visual frames employed by the Internet Research Agency are designed to embody the vox populi with relatable, familiar, or attractive faces of ordinary people. The results also show that IRA influence operations displayed cultural acuity and familiarity with the social identity of their targets, and that the visual narrative it crafted trafficked primarily in the tropes of regular guys or implausibly attractive young women. We discuss these findings and argue that state propaganda has effectively attuned to both subcultural and visual affordances of social platforms.