List of Articles  

• Open Access Article
  • Abstract Page
  • Full-Text
  
  1 - Constructing New Features for Spam Detection on Twitter
  Arash  Erami Elham  Parvinnia
  Nowadays, by the growth of the internet, social networks are attracting unprecedented attention to themselves. Most people are at least active in one social network. Users in social networks follow their favorite people and topics to discover the latest news about the More

  Nowadays, by the growth of the internet, social networks are attracting unprecedented attention to themselves. Most people are at least active in one social network. Users in social networks follow their favorite people and topics to discover the latest news about them. This rising number of users has made social networks fertile grounds for advertising and finding the bait. Social networks also become celebrities’ popularity criterion. The problem is that some accounts created to spread malicious links, steal user’s information, and display advertising. These accounts are mainly controlled and supervised by an automatic program. Not only the increase in fake accounts has costs for social networks companies, but it also influences network quality. In this paper, we offer some new and low-cost features to distinguish spam accounts on Twitter. This paper offers some low cost and a new feature to distinguish spam accounts of Twitter. We apply machine learning algorithms to predestined datasets, and by looking at the characteristics of the accounts, then we anticipate class of users by the accuracy of 99.18%. Manuscript profile