Home > Archive > 2017 > Volume 7 Number 6 (Dec. 2017) >
IJMLC 2017 Vol.7(6): 198-202 ISSN: 2010-3700
DOI: 10.18178/ijmlc.2017.7.6.646

Detecting Fake Followers in Twitter: A Machine Learning Approach

Ashraf Khalil, Hassan Hajjdiab, and Nabeel Al-Qirim

Abstract—Twitter popularity has fostered the emergence of a new spam marketplace. The services that this market provides include: the sale of fraudulent accounts, affiliate programs that facilitate distributing Twitter spam, as well as a cadre of spammers who execute large scale spam campaigns. In addition, twitter users have started to buy fake followers of their accounts. In this paper we present machine learning algorithms we have developed to detect fake followers in Twitter. Based on an account created for the purpose of our study, we manually verified 13000 purchased fake followers and 5386 genuine followers. Then, we identified a number of characteristics that distinguish fake and genuine followers. We used these characteristics as attributes to machine learning algorithms to classify users as fake or genuine. We have achieved high detection accuracy using some machine learning algorithms and low accuracy using others.

Index Terms—Twitter, security, machine learning, fake follower, social networks.

Ashraf Khalil and Hassan Hajjdiab are with Abu Dhabi University, Abu Dhabi, UAE (e-mail: ashraf.khalil@adu.ac.ae, hassan.hajjdiab@adu.ac.ae). Nabeel Al-Qirim is with United Arab Emirates University, Al Ain, UAE (e-mail: nalqirim@uaeu.ac.ae).


Cite: Ashraf Khalil, Hassan Hajjdiab, and Nabeel Al-Qirim, "Detecting Fake Followers in Twitter: A Machine Learning Approach," International Journal of Machine Learning and Computing vol. 7, no. 6, pp. 198-202, 2017.

General Information

  • ISSN: 2010-3700 (Online)
  • Abbreviated Title: Int. J. Mach. Learn. Comput.
  • Frequency: Bimonthly
  • DOI: 10.18178/IJMLC
  • Editor-in-Chief: Dr. Lin Huang
  • Executive Editor:  Ms. Cherry L. Chen
  • Abstracing/Indexing: Inspec (IET), Google Scholar, Crossref, ProQuest, Electronic Journals Library.
  • E-mail: ijmlc@ejournal.net

Article Metrics