Coordinated link sharing on Facebook

Yang, Y., Paudel, R., McShan, J., Hindman, M., Huang, H. H., & Broniatowski, D. (2025). Coordinated link sharing on Facebook. Scientific Reports, 15, 15684. https://doi.org/10.1038/s41598-025-00233-w

View paper

Summary

This paper introduces a new method for detecting coordinated link-sharing behavior on Facebook that reduces reliance on post-timing signals, which manipulators can easily alter. The authors argue that while timing has been a mainstay of coordination detection, it is trivially manipulable and therefore vulnerable to evasion. Instead, they exploit statistical regularities in the speed and frequency of link sharing across accounts as more robust indicators of coordination. The approach is validated on a large corpus of 11.2 million Facebook link posts drawn from roughly 16,000 sources, positioning the work within computational social science and platform integrity research.

Key Contributions

• A methodological advance in coordination detection that reduces dependence on easily manipulated post-timing features.
• Introduction of speed- and frequency-based statistical signatures as detection signals.
• Empirical application and validation at scale using millions of Facebook link posts.

Methods

• Analysis of a large-scale Facebook dataset comprising 11.2 million link posts.
• Posts sourced from a curated list of roughly 16,000 accounts or domains.
• Statistical modeling of sharing speed and frequency distributions to identify coordinated activity.
• Empirical validation of the detection approach against this corpus.

Findings

• Link-sharing speed and frequency display consistent statistical regularities across accounts.
• These regularities can be leveraged to detect coordinated sharing behavior on Facebook.
• The proposed signals offer a more evasion-resistant alternative to timing-based methods, validated on a large empirical dataset.

Connections

This work sits squarely in the coordinated link-sharing detection tradition, most directly extending the coordinated link-sharing behavior (CLSB) framework developed across Giglietto2020-9d8acdd7, Giglietto2022-0e951ac5, and Giglietto2023-fa71a001, whose reliance on temporal co-sharing this paper critiques and seeks to improve. Its focus on evasion-resistant, network-based coordination signals connects it to broader methodological efforts in Luceri2025-tr and Minici2024-tf on robust detection of coordinated inauthentic behavior.