New Math Shows Which Social Media Links to Cut to Stop Viral Rumors

When false information goes viral, it spreads a lot like a cold. You see a post, you pass it on, and each new sharer infects more people. Researchers have long used this 'epidemic' math — the same models used to track flu outbreaks — to figure out who or what is driving the spread.

The big question they're trying to answer: if you could only cut a handful of connections, which ones would do the most good? Cutting the wrong links does nothing. Cutting the right ones can slow a rumor to a crawl. Earlier methods guessed at which links mattered, but their guesses were often too loose — like a smoke detector that goes off for burnt toast and a real fire alike.

This new paper, from Nguyen Hong Phuc, Duong Khanh Ly and Hoang Phi Dung, tightens that guess. They focused on a lesser-studied model that includes 'dormant' people — those who have seen a post but haven't shared it yet. That's actually how most of us behave online. By tracking both the dormant and active groups together, their formula pinpoints the influential connections more sharply. They tested it on a made-up random network and on Haslemere, a real dataset from a UK town, and it beat the older method at containing the spread.

The catch: this is a math paper, not a product. Nobody's platforms are using it yet, and actually cutting connections between real people raises obvious questions about censorship and fairness. Still, it's a step toward tools that could help platforms — or public health officials — target the handful of links that matter instead of blanket-blocking everything.