A group of researchers are creating a false tweets lie detector. The goal of the detector is to stop fast spreading, and potentially dangerous tweets, that are inaccurate and unwarranted.
Researchers point to the 2011 London riots and vote-rigging accusations that were said to have occurred during the Kenyan elections.
Researchers at the University of Sheffield are not only tying the program to Twitter. They also say it can analyze data from other sites. The product is being funded by the European Union and is titled “Pheme.”
When complete the product will filter rumors into the following categories: Speculation, controversy, misinformation and disinformation.
After sorting through sources the Pheme platform will then evaluate sources for authority ranking. For example, a New York Times post would hold more authority and trust than a source from an unknown person or business.
As Pheme continues to follow certain conversations it will determine the accuracy of the information.
Could online news soon be quickly verified on an International level? It looks like this Twitter lie detector test could be a big win for social validation.
Social interaction analysis has become a big business model for many enterprise platforms. While this new lie detector platform allows for false reporting, developers can already sort through sentiment analysis via the Twitter Search API. Dave Polykoff at Brand.com explains that, “Twitter actually makes it easy and fun to search for public tweets by referencing emoticons like frowns or happy faces, to indicate positive or negative sentiment.” Polykoff further explains that developers can also “use a simple question mark (?) to indicate tweets where people are asking questions.”
By examining a tweets sentiment and then determining its validity, brands and consumers can more readily determine the truthfulness of social signals.