After Donald Trump was ejected from Twitter he set up his own social network where he’d be free from censorship. This act of pique shed fresh light on the darker corners of the web where extremist attitudes dominate.
While we might think the former president’s presence would be the tip of a very large iceberg, research from the University of Exeter suggests that most of these fora are driven by a small, yet vocal, minority.
Tackling extremism
The study shows that a small group of posters can be responsible for around a third of all content. A larger group is then committed but generally more restrained. These are then followed by a group that contributes now and then, and the majority who rarely post at all.
Interestingly, the most active posters aren’t always those at the core of the network, with the study suggesting that most occupy the fringes instead. As a result, they tend to limit their contributions to a small number of threads. What’s more, these patterns seem to hold up regardless of the particular ideology, language, or even size of the community.
“Anecdotal evidence from existing case-studies already pointed to similarities in posting structure, so we anticipated to find it,” the researchers explain. “However, what we didn’t expect to see was such a low level of diversity—if you put all the graphs and metrics side by side, they are hardly distinguishable.”
“These results not only clarify how influence works in extremist online spaces, they also deliver actionable intelligence for government agencies tasked with tackling online extremism and radicalization. As the rapid growth and diversification of extremist online spaces causes serious challenges to security and law-enforcement practitioners, this type of knowledge—which offers both general and granular observations—is directly useful.”
The researchers hope that this better understanding of the nature of extremist communities can help develop better interventions to tackle them.
“Having this more fine-grained categorical structure of user behavior will allow for more focused intervention measures such as targeted undercover posting strategies or better allocation of scarce resources for investigating specific users,” they conclude.
