Have you ever told a fable on Twitter, perhaps for humor’s sake or as an accidental faux pas? Ever retweeted something, only to later discover it was false? Soon such tweets will be tested under the watchful eye of Pheme, a lie detector of sorts that aims to determine in real time whether a tweet contains the truth or is a lie.
The researchers behind the project chose “Pheme” as the name for their system based on the mythological Greek figure, and hope to provide a system that identifies whether a tweet is truthful by tracing its origins and testing the details against other information it gathers, determining whether it holds up under scrutiny or should be discarded.
By doing this, Pheme could help put a damper on Internet rumors that are false and other bits of less-than-truthful details that surface on the microblogging website. Such could be of a benefit to many, ranging from law enforcement that need to know if a tweet is true to journalists digging around for their next big story.
Under the system, information will fall under one of four categories: speculation, controversy, misinformation, and disinformation (a falsehood with malicious intent, essentially). The system could be available within the next 18 months, with prototypes possibly being made available sooner.