Fake News Recognition applying Service-based Cross-Media Analytics

ended

The phenomenon of fake news is as old as human’s ability to communicate. Recently, technology has made it easy to spread fake news with unprecedented speed towards the unprotected public, which has resulted in undesired external influence in European elections and referenda (Brexit, Catalonia). Though a variety of fact checkers have arisen out of nowhere, current solutions fail on many aspects: they are primarily based on manual activities, focus mainly on politics and political statements and are highly country-/language specific. Due to the speed and sheer size of the Internet and Social Media platforms, manual defence against fake news is futile. Automated solutions are required involving automatic algorithms that quickly identify fake news and tag them as such, allowing Europeans to make educated, correct decisions. At SAIL LABS, we develop an alternative fact checker: TruthCheck. This combines available and novel technologies such as Machine-Learning algorithms, Trust Modelling, Natural Language Processing, and Speech-, Image- and Video-Recognition to verify feeds from TV, Radio, Newspapers, Internet and Social Media in 31 languages on a semi-automatic basis. TruthCheck will be available as free version to the public, while licenses are offered to enterprises, NGOs and governmental institutions. In absence of a clear market leader we intend to take the opportunity to become first-comer in this segment, thereby targeting the fast-growing market of media analysis. Through this project, we intend to assess and develop the required commercial and technical work that it needs to ensure successful and swift market launch.