As if it’s not bad enough that your bosses could find out about your anonymous blogging, the US Government could soon be using automated robots to track down what you write and attribute it to you.
Researchers at the University of Arizona have been working on their “Dark Web” project with the not-at-all-daunting mission to “systematically collect and analyse all terrorist-generated content on the Web”.
Their Writeprint system is supposed to be able to identify who wrote particular articles, with up to 95% accuracy, based on studying “thousands of lingual, structural and semantic features in online postings”.
If that weren’t enough, it “analyses which types of individuals might be more susceptible to recruitment by extremist groups, and which messages or rhetoric are more effective in radicalising people.”
So, you’d better tone down your unsigned rants about Steve Ballmer. Oh, no, wait. That’s me.
It’s an interesting project that isn’t without its risks. Terrorists are fighting back (wouldn’t you know it) by setting up booby-traps in online forums and tricking the automated bots into downloading viruses. Dirty rotten scoundrels, eh? Anyone would think they didn’t want to be caught.
I can’t help feeling that the terrorists might work their way around this. Remember that game where fold up a piece of paper and then take turns to draw different bits of a human body?
“Right, Osama, I’ll write the first two lines, then you write the next, then one of your body doubles can write the next…”
Good luck, Dark Web.