An in-depth analysis of abuse on Twitter

This paper was presented at VB2014 in Seattle, WA, USA. In this paper, we examine Twitter in depth, including a study of 500,000,000 tweets from a two-week period to analyse how it is abused. Most Twitter abuse takes the form of tweets with links to malicious and spam websites. These websites take many forms, including: spam websites; scam sites involved in compromising more Twitter accounts; phishing websites; and websites with malware or offering cracked versions of software. Many of the malicious tweets are sent from legitimate accounts that have been compromised, creating a range of problems for their owners. The scale of the threat is significant. Previous research (notably ‘@spam', Grier 2010) indicates that using URL blacklists is ineffective in detecting threats. Our research sho

Leave a Reply