Social media has become one of the biggest sources of information. Many people willingly share their experiences and knowledge through countless social media posts and updates. Even the media has started to rely on it when covering various events and breaking news. However, not everything you read on social platforms is true. Often, the information you find there ends up to be nothing more than a rumour. If only there was a way to set the truth from the lies! Well, that may soon become a reality. A group of researchers has created the very first lie detector for social media.
Scholars from the University of Sheffield have found a way to develop a system, which can assess the reliability of social media posts. Even though its algorithm requires some more work, researchers behind the project have been approved for an EU grant. Therefore, it may be ready faster than you think.
The goal of the social media lie detector is to offer a better and more accurate judgment than that of humans. This sounds hard to achieve given that the algorithm of the program was developed by humans. Nevertheless, the research team from the University of Sheffield has found a way to make it more accurate.
For that purpose, it came up with a classification of the most common types of web rumours. The team concluded that they are four main types of online rumours – misinformation, disinformation, speculation and controversy. Scientists plan to use this classification in their lie detector algorithm.
In addition, the complicated system will also take into account the source of the news and their authority. It will be able to tell whether the post was published by an expert, a journalist a media, an eyewitness or even by bots. How? According to the team behind the project, it will examine the background and the history of the given account. Well, we doubt that this is in correlation with the privacy policy of the top social media websites. Nevertheless, on theory its sounds logical.
Once the system has done its magic, it will show the results on a dashboard. Researchers promise that the first stage of the project would be ready by the end of 2015. After that, it will undergo different test with healthcare professionals and journalists.
The University of Sheffield plans to eventually release the system as a customized tool for journalists. Although this sounds quite interesting, transforming these ideas into practice would be the true challenge.
