Improved Micro-blog Classification for Detecting Abusive Arabic Twitter Accounts

The increased use of social media in Arab regions has attracted spammers seeking new victims. Spammers use accounts on Twitter to distribute adult content in Arabic-language tweets, yet this content is prohibited in these countries due to Arabic cultural norms. These spammers succeed in sending targeted spam by exploiting vulnerabilities in content-filtering and internet censorship systems, primarily by using misspelled words to bypass content filters. In this paper we propose an Arabic word correction method to address this vulnerability. Using our approach, we achieve a predictive accuracy of 96.5% for detecting abusive accounts with Arabic tweets.