Sorting the social network fakers from the movers and shakers

How can we detect fake profiles to preclude their disruptive and deleterious effects on social media and social networks? Writing in the International Journal of Information and Computer Security, Somya Ranjan Sahoo and B.B. Gupta of the National Institute of Technology at Kurukshetra in Haryana, India, discuss the issues and possible solutions.

Recent research in fake profile detection, they explain has focused on machine learning in order to reveal the kind of suspicious account activity that might betray a fake account. The team is now taking machine learning to big data to find a better way to distinguish the fakers from the movers and shakers, on the well-known social networking system, Facebook.

Facebook is an important part of life for many people, for organizations and other entities. There are some 2.5 billion monthly active users and approximately 1.7 billion people use a Facebook account every day. It is not known how many fake accounts lurk within those statistics. It is known that many malicious third parties hoping to gain access to personal, private, and other data with malicious intent will exploit loopholes in the Facebook system. That combined with social engineering confidence tricks and other exploits can provide them with sufficient data to access other people’s accounts and from there to steal personal information and then even break into other systems such as email and banking systems.

There have been many security exploits used to gain malicious access to information but the use of fake accounts can be the most successful especially when the person being attacked assumes the legitimacy or honesty of the fake account, accepts a friendship request or clicks on a malware phishing link, for instance.

The team’s tailored extension for the popular Google Chrome browser allows them to successfully spot fake accounts. This might be used by security experts as a third-party reporting tool to help Facebook cleanup its systems or ultimately perhaps by the company or users. The team is also now extending the approach to other popular networking sites such as Twitter and Google+.

Sahoo, S.R. and Gupta, B.B. (2020) ‘Fake profile detection in multimedia big data on online social networks‘, Int. J. Information and Computer Security, Vol. 12, Nos. 2/3, pp.303-331.