Powered by Google

Associated Partner

Foiling fraudsters in just 10 keystrokes

Ground-breaking research in Newcastle aims to throw a spotlight on fraudsters and paedophiles using the internet to cover their tracks. John Hill reports.

Someone using a computer

WILL companies and governments soon be able to identify hackers from the way they use a keyboard?

Newcastle University is looking at applications for new technology which can pinpoint a typist’s sex, age and culture within 10 keystrokes – a breakthrough that could have multiple uses in law enforcement, corporate fraud prevention and protection of children.

Newcastle University associate professor Roy Maxion conducted his research at Pittsburgh’s Carnegie Mellon University, monitoring and timing students as they typed in an agreed password 400 times.

Prof Maxion’s work caught the imagination of former Northumbria Police detective chief inspector Phil Butler, who now works as director of external relations for Newcastle University’s Centre for CyberCrime and Computer Security.

Mr Butler said: “He takes 50 people at a time and hooks their fingers up to electronic sensors, then videos, monitors and records their typing patterns, speed and rhythms with a very accurate clock.

“He can now identify anyone using a keyboard within a 95% accuracy within 10 keystrokes. As soon as you type 10 numbers or letters he can work out your sex, your culture, your age and whether you have any hand injuries.

“In general, women’s typing tends to flow more and is a little quicker. You’d expect men’s typing to be a little more heavy-handed and apparently that’s the case.

“We’re looking at the application of the research, particularly in relation to internet grooming. If children are talking to each other on Windows Live or MSN Messenger, Microsoft might be able to see if there’s an adult on there.

“It can also be used with touch phones such as the iPhone and at ATM machines. Some corporations are thinking it might be useful to identify someone’s identity when they’re using a cash machine.

“We’re going to submit a proposal to the Engineering and Physical Sciences Research Council for a bid of probably about £1m to develop the research. We’d like to look at various adaptations of the research for use in law enforcement, forensics and for companies trying to avoid fraud.”

Share

Share