A group of researchers funded by the EU have been working on a type of social media lie detector. The project, named Pheme, is said to have grown out of a desire to properly respond to the 2011 London riots. Pheme is able to analyze information such as public facebooks posts and tweets to categorize the validity of the information in real-time.
Instead of analyzing the language in posts, Pheme tries to evaluate the quality of the information as well as it’s source. Pheme looks at information such as the source of the information (news outlets, alleged experts, eyewitnesses, automated bots, etc.) as well as the history of each social media account, to determine if the information is reliable or not. Pheme also looks for other sources to confirm or deny the information and plots how the topic evolves on social networks. Using this information, Pheme is able to categorize the information into 4 segments: speculation, controversy, misinformation & disinformation.
Intended for governments and emergency services, Pheme could certainly be useful for finding out real-time information on what’s actually going on in times of crisis. The ability to quickly provide resources towards areas of need (while avoiding false claims and ambiguity) could be extremely valuable. Pheme could also be useful for news outlets who want to report on real-time information while making sure the information they are providing is accurate and reliable. What other ways could you envision governments, organizations, or companies using Pheme?