Coordinated link sharing on Facebook.

Area of Science:

• Social Media Analysis
• Computational Social Science
• Network Science

Background:

• Coordinated posting is a common tactic used by malicious actors to manipulate social media platforms.
• Existing detection methods often rely on post-timing, which is easily manipulated and thus unreliable.
• There is a need for more robust and adaptable methods to detect coordinated inauthentic behavior online.

Purpose of the Study:

• To develop and validate a novel methodological advancement for detecting coordinated social media manipulation.
• To leverage statistical patterns in sharing speed and frequency as indicators of coordination.
• To create a framework for building interpretable and adaptable models of platform manipulation.

Main Methods:

• Analysis of 11.2 million link posts from 16,169 popular English-language Facebook pages.
• Focus on pages referencing top US politicians to capture significant user engagement.
• Development of a detection approach based on statistical patterns in sharing speed and frequency.

Main Results:

• The proposed method offers a significant advancement over timing-based detection techniques.
• The approach was validated on a large dataset of Facebook posts, demonstrating its effectiveness.
• The method's adaptability across different contexts, platforms, and time periods was highlighted.

Conclusions:

• Statistical patterns in sharing speed and frequency provide a more reliable signal for detecting coordinated social media manipulation.
• The developed approach enables the creation of valid, testable, and human-interpretable models for identifying platform manipulations.
• This research contributes to a better understanding and mitigation of inauthentic online behaviors.