Social media platforms' growth has been nothing short of astounding, reshaping the way we communicate, share, and connect. The accessibility and user-friendly nature of platforms like YouTube, Facebook, Twitter, and others have enabled millions of individuals to express themselves, engage with content, and establish communities. However, alongside positive aspects, the rise of social media has also brought misinformation, which can spread like wildfire across these platforms. This research focuses on identifying and characterizing YouTube accounts that use automated or semi-automated means to boost their user engagement statistics and to amplify their misinformation. We assess statistics from 3542 channels, combining techniques including rolling window correlation analysis, anomaly detection, rule-based classification, and clustering. We found varying levels of inorganic activity across the channels, with some of those exhibiting high levels being suspended by YouTube. We also discovered that some channels that boost their content inorganically publish videos of a similar discourse.

PDF