The Kauwa-Kaate Fake News Detection System

Abstract

Fake news spread via social media is a major problem today. It is not easy with current-generation tools to check if a particular article is genuine or contains fake news. While there are many Web sites today that debunk viral fake news, checking if a particular article has been debunked or is true is not easy for an end-user. Search engines like Google do not make it easy to check a complete article since they limit the number of query keywords. In this paper, we outline the architecture of the Kauwa-Kaate system for fact-checking articles. Queried articles are searched against articles crawled from fact-checking sites, as well as against articles crawled from trusted news sites. Our system supports querying based on text as well as on images and video; the latter features are very important since many fake news articles are based on images and videos. We also describe the user interfaces which we will use to demonstrate the Kauwa-Kaate system.