Rocket ship or Blimp? - Implications of Malicious Accounts removal on Twitter

In this study we investigate how the removal of malicious accounts that follow legitimate accounts owned by popular people impacts the popularity of tweets posted by celebrities and politicians. Using retweet counts, we analyze to what extent malicious accounts contribute to amplification of tweets across the network. We organize tweets into three broad categories (Rocket Ship, Jet or Blimp) and investigate how the distribution of tweets is influenced by a cleanup of malicious accounts. To understand how the suspension of malicious accounts impacts the propagation of messages on Twitter, we conduct a descriptive statistical analysis of retweets of a total of 464 Donald Trump tweets. We find a statistically significant difference in the mean count of retweets and favorites before and after the malicious account removal. Preliminary results of our analysis show that the implications of Twitter’s cleanup initiatives, which targeted malicious accounts, are visible in the narrowing amplitudes of retweet values. However, the distribution of tweet categories based on the number of retweets remains unchanged.