Twitter 'lie-detector' will combat false internet rumours

SCIENTISTS claim they are 18 months away from developing a "lie detector" for Twitter that could stop dangerous rumours from spreading across the social network.

The system, named Pheme after the Greek mythological figure known for rumour and gossip, will categorise sources to assess their authority and give more weight to established news outlets and experts, according to The Times.

It will identify "bots", computerised spammers, and look at the history of an account to spot if it has been set up to spread false information.

Subscribe to The Week

Escape your echo chamber. Get the facts behind the news, plus analysis from multiple perspectives.

SUBSCRIBE & SAVE

Sign up for The Week's Free Newsletters

From our morning news briefing to a weekly Good News Newsletter, get the best of The Week delivered directly to your inbox.

From our morning news briefing to a weekly Good News Newsletter, get the best of The Week delivered directly to your inbox.

Sign up

Pheme will also search for sources that can corroborate or deny a story in real time, as well as plot how the story is spreading to assess whether it is true or false. The results will be displayed to the user on screen, telling people if a false story is taking hold among the public.

The creators believe the system would have been valuable during events such as the London riots in August 2011, when Twitter-users diverted police attention from real events by spreading rumours that landmarks across the city had been set on fire and animals had escaped from London Zoo.

The program could also have been used to stop tweeters – including Sally Bercow and Alan Davies – from spreading rumours that falsely linked Lord McAlpine to child sex abuse. Both Bercow and Davies agreed financial settlements with the late peer as a result of the incorrect claims.

Pheme will classify online rumours into four types, says The Times: speculation, controversy, misinformation and disinformation.

The lead researcher on the project, Dr Kalina Bontcheva, from the University of Sheffield, says the system would enable governments, emergency services and health agencies to respond to untruths.

Bontcheva is hoping for a final version within 18 months, but believes that working prototypes may be available sooner. The project will cost about £3.5m to complete, met mostly through a grant from the European Union, with Warwick University and King's College London also taking part.