Cloud-Based System for Fake Tweet Identification

摘要

Twitter, as a popular social media platform, has great influence on society. Unfortunately, studies have observed some unreliable or fake information spread via Twitter with harmful consequences. In this paper, we present a cloud-based system to identify tweets with fake news by using tweet text characteristics, machine learning technique, reverse image research, and fake news source comparison. Major components in the system have been deployed as web services, which can be easily integrated or extended in other applications. In addition, web-based interface is developed for general users to validate tweet credibility via web browsers.