Trollbusters: This algorithm can predict if you're going to get banned

Trolls can be spotted from just five comments

The way commenters write can be used to detect if they'll end up getting banned, according to US researchers working on troll-busting algorithms.

A team from Cornell and Stanford Universities scanned the comment threads on three news sites - CNN, Breitbart and IGN - over the course of a year and a half. That totalled up to 35 million comments, sent by almost two million users. 50,000 of those users went on to be banned from the sites.

The researchers found that those banned users wrote in a different way to others. Their comments were generally harder to read and used fewer words that indicated positive emotion. They also behaved slightly differently in how they moved around the site - spending more time focused in individual threads than users who weren't banned.

From that data, the researchers built a model that could guess with 80% accuracy whether or not a user would go on to be banned from the content of their first five posts. Examining the first ten comments raised the accuracy level further - by two percentage points.

The team hopes that the work can be used to develop moderation tools that automatically highlight users that may go on to be disruptive, saving moderators time. But they also warned that their findings showed overly harsh moderation tended to exacerbate antisocial behaviour, and that a light touch was more effective than throwing out the banhammer.

The research was published in Social and Information Networks (via The Guardian)

Duncan Geere
Duncan Geere is TechRadar's science writer. Every day he finds the most interesting science news and explains why you should care. You can read more of his stories here, and you can find him on Twitter under the handle @duncangeere.
Latest in Internet
A mobile phone showing the Signal logo in front of a screen showing the app
Signalgate explained: what is Signal, and how secure is the messaging app?
Quordle on a smartphone held in a hand
Quordle hints and answers for Friday, March 28 (game #1159)
NYT Strands homescreen on a mobile phone screen, on a light blue background
NYT Strands hints and answers for Friday, March 28 (game #390)
Google Maps on a phone being held in someone's hand
Google Maps is getting two key upgrades, for easier route planning and quicker access to Gemini AI
Quordle on a smartphone held in a hand
Quordle hints and answers for Thursday, March 27 (game #1158)
NYT Strands homescreen on a mobile phone screen, on a light blue background
NYT Strands hints and answers for Thursday, March 27 (game #389)
Latest in News
Power cables stretching out in front of the horizon
Solar grids could be hijacked and even potentially disabled by these security flaws
Lenovo | Thinkpad T14s Gen 6 Snapdragon
Windows 11’s latest patch declares war on BIOS updates for some Lenovo laptops, blocking them as a security risk in a bizarre turn of events
Samsung Galaxy Watch Ultra
Samsung confirms Galaxy Watches aren't tracking sleep properly – here's the fix if you're affected
Tomodachi Life: Living the Dream screenshot showing a Mii smelling some fresh flowers.
Tomodachi Life: Living the Dream is a sequel to my favorite 3DS game, and I think it's already packing the charm that inZOI lacks
Spam messages
Microsoft Stream classic domain hijacked, causing spam across SharePoint
ChatGPT logo
ChatGPT 4o just got better, although I’m yet to notice a difference