A team of researchers from seven different countries led by The University of Sheffield is developing a new system for Twitter that will help identify fake or hoax tweets and posts.
Dubbed Pheme the new system will use WebLyzard web intelligence platform and aims to categorise posts into four categories of online rumours including speculation, controversy, misinformation and disinformation.
Pheme will also categorise sources and determine their authority. Established news outlets, publishing houses and experts will be given more weight and bots known to send out spam will be ignored.
The system will also go and search through information from other sources and corroborate or deny the information being posted. Further, Pheme will also plot how conversations on social networks evolve and subsequently use all this information to determine whether the claim placed by a certain entity is true or false.
Team leader Dr Kalina Bontcheva said: “There was a suggestion after the 2011 riots that social networks should have been shut down, to prevent the rioters using them to organise. But social networks also provide useful information – the problem is that it all happens so fast and we can’t quickly sort truth from lies.”
Bontcheva revealed that because of the speed it is difficult to respond to rumours. Pheme will allow analysis, verification and tracking of information in real time and on top of that will allow to determine whether they are true or false.
