Tackling Fake News

Fake news is not always fake. In fact, accusations of fakery are most typically levelled at news that is biased: stories that cherry-pick from the information; that use leading lexicalisation; that emphasise, de-emphasise, exaggerate and focus to advance a partisan position. But then, every news source has to be selective in telling a concise story; particular words must be chosen to effectively communicate the story; emphasis and de-emphasis must be used to simplify the message. So when do brevity, clarity and simplicity become bias? It turns out that theories of reasoning and argument might be what’s required to explain these relationships. Early results from fields as diverse as mathematics, linguistics, philosophy and psychology have shown that relationships between propositions used in human reasoning are far more nuanced than was previously thought, and by unpicking these nuances it becomes possible to engineer software systems that support and enhance human interactions. This project will develop and test the underlying theory to lay the foundation for developing unique solutions for dealing with the challenges presented by fake news.