Uncovering Coordinated Networks on Social Media: Methods and Case Studies

The unsupervised bot detection approach typically requires calculating similarity of different accounts and clustering them into groups.

“This unsupervised approach typically requires calculating the similarity of different accounts and subsequently clustering them into groups (Pacheco et al., 2021).”

Popularity metrics can be gamed with inauthentic behavior, including bots, organized trolls, and fake-account networks.

“Popularity metrics can also be gamed with inauthentic behavior, including bots, organized trolls, and fake-account networks (Pacheco et al., 2021; Sen et al., 2018).”

Coordination in account activity does not necessarily coincide with automation.

“While coordination in account activity does not necessarily coincide with automation [36,37], accounts created on similar dates and following users successively are more likely to be automated [3,48].”

The BLOC model was compared to three baseline methods that make different assumptions about the behavioral traits shared among coordinated accounts: hashtag sequences (Hash), activity (Activity), and co-retweet (CoRT).

“We compared the BLOC model to three baseline methods, which make different assumptions about the behavioral traits that may be shared among coordinated accounts [6]: hashtag sequences (Hash), activity (Activity), and co-retweet (CoRT).”

Coordination detection methods typically extract account features designed to target specific suspicious behavioral patterns.

“These methods typically extract account features designed to target specific suspicious behavioral patterns [6].”

Coordination detection involves clustering suspiciously similar accounts into groups.

“While individual bot detection aims to separate individual human and bot-like accounts, coordination detection involves clustering suspiciously similar accounts into groups [6].”

Without explicit countermeasures, these systems will face the same coordinated inauthentic behavior that has plagued human social media.

“Our results suggest that without explicit countermeasures, these systems will face the same coordinated inauthentic behavior that has plagued human social media 17,23,24 .”

Key signals for unsupervised bot detection include temporal activities, common retweets, and URLs shared in tweets.

“Key signals include temporal activities (Chavoshi et al., 2016;Keller et al., 2017), common retweets (Nizzoli et al., 2021), and URLs shared in tweets (Pacheco et al., 2021;Giglietto et al., 2020).”