Researchers developing system to detect hoax tweets and posts
21 Feb 2014
A new system for detecting fake or hoax tweets and posts is being developed for Twitter by a team of researchers. The team from seven different universities is being led by The University of Sheffield.
The new system called Pheme would use WebLyzard web intelligence platform and aims to categorise posts into four categories of online rumours including speculation, controversy, misinformation and disinformation.
The system would also categorise sources and determine their authority. More weight would be given to established news outlets, publishing houses and experts with bots known to send out spam to be ignored.
The system would also search through information from other sources and corroborate or deny the information being posted. Also, the system would plot how conversations on social networks evolved and subsequently use all this information to determine whether the claim placed by a certain entity was true or false.
Techienews quoted team leader Dr Kalina Bontcheva as saying there was a suggestion after the 2011 UK riots that social networks should have been shut down, to prevent the rioters using them to organise. He added, though that social networks also provided useful information, but "the problem was that it all happened so fast that one could not quickly sort truth from lies.''
According to commentators, the system could be useful for governments and emergency services.
The system is named after a Greek mythological figure who on learning about the private affairs of gods and men, repeated it with a dull whisper at first but louder with each successive repetition, until it became public knowledge.
The results would be focused on the information quality, unlike similar analytics tools that focused more on language.
The Pheme results would be displayed in a visual dashboard that would give some sense on the credibility or otherwise of the rumour.
"We can already handle many of the challenges involved [on the Internet], such as the sheer volume of information in social networks, the speed at which it appears and the variety of forms, from tweets, to videos, pictures and blog posts," Bontcheva, said in a statement. "But it's currently not possible to automatically analyse, in real time, whether a piece of information is true or false and this is what we've now set out to achieve."
(Red more: EU project to build lie detector for social media)
