Policing online hate groups is like a never-ending game of whack-a-mole: moderators remove one neo-Nazi page on Facebook, only for another to appear hours later. It’s an approach that isn’t working, but researchers have suggested several alternative strategies that might.

The scale of the problem: The team from George Washington University examined the dynamics of “hate communities”—groups that organize individuals with similar views—on social media platforms Facebook and VKontakte (the equivalent in Russia) over a few months. They found that these networks are remarkably globally interconnected and resilient at the micro level when attacked, crossing social media platforms and jumping between countries, continents and languages. Real-world evidence of this interconnectedness can be seen in the way white extremist attackers in Norway, New Zealand and the US have explicitly drawn inspiration from each other.

The current approach is broken: Their mathematical model predicts that policing within a single platform, like Facebook, can actually make the spread of hate speech worse, and could eventually push it underground, where it’s even harder to study and combat. The team explained their findings in a paper in Nature this week.

What can be done? The researchers suggest policies which could be implemented by social media companies:

— Ban relatively small hate clusters, rather than the largest. These are easier to locate, and eliminating them can help to stop the larger clusters from forming in the first place.

— Ban a small number of users chosen randomly from online hate clusters. This avoids banning whole groups of users, which results in outrage and allegations of speech suppression.

— Platform administrators should promote the organization of clusters of “anti-hate” users to counteract hate clusters.

— Since many hate groups online have opposing views, platform administrators should introduce an artificial group of users to sow division between these groups. The researchers found these sorts of battles could bring down large hate clusters that have opposing views.

How likely is it? Some of the policies, especially the latter two, are pretty leftfield. But given current approaches are so profoundly ineffective, it’s surely worth social media companies trying them out.  Over to you, Mark Zuckerberg.

Sign up here for our daily newsletter The Download to get your dose of the latest must-read news from the world of emerging tech.

Let’s block ads! (Why?)

Source link