Communications of the ACM
Ict Professors Explore Deception in Social Media
Credit: Adventory Ltd.
The explosion of the Internet and social media has literally put the world at our fingertips, revolutionizing the way people connect and share information. However, for all the positives social media provides, it can also open the door to deception, potentially wreaking havoc on people's lives both personally and professionally.
For instance, you may receive a 'friend request' on Facebook from someone you vaguely remember from your childhood but how do you determine if the person making the request is the genuine person or someone masquerading as such in order to obtain personal information from your Facebook page? Assistant Professor Michael Tsikerdekis and Associate Professor Sherali Zeadally, in the Information Communication Technology (ICT) program at the University of Kentucky College of Communication and Information, are currently exploring the area of deception in social media.
In their recent article, "Multiple Account Identity Deception Detection in Social Media Using Non-Verbal Behavior" published in August 2014 by the peer reviewed IEEE Transactions on Information Forensics and Security, they used Wikipedia as an experimental case to demonstrate the high accuracy of their method over previous deception detection methods.
"There are many fake accounts being created on Wikipedia with the sole intent to add biases to articles," Tsikerdekis says. "These accounts are discovered and banned but it could take up to a year or more to get them banned."
The method tracks non-verbal user activity in order to distinguish fake accounts from real ones. Just like with non-verbal behavior in the real world, such as body movement, in social media services we all leave signs that betray our behavior. The speed that we type on a keyboard, the choice on whether to send a message or view a profile, and the number of comments we make to certain pages, all reveal things about us. The method measured this non-verbal behavior on the early days of each account on Wikipedia. The results indicated a divergence between the number of revisions people made in the first days after creating an account, the time it took between revisions, the mean number of bytes added or removed in each revision, and, the places on which revisions were made. The last case was in particular a strong clue that an account was created with bad intention since this behavior betrayed where people spend most of their time on Wikipedia. Similarly, with social media such as Facebook and Twitter, a photo of someone can be used to set up a fake account.
"A fake Twitter account exists for UK President Capilouto that is going strong with over 7,000 followers. The account does disclose that it is a fake account, but what if it did not? How would you be able to tell the difference?" Tsikerdekis says.
In a follow-up study entitled "Online Deception in Social Media" published in the September 2014 peer-reviewed Communications of the ACM, Tsikerdekis and Zeadally compare traditional offline deception to online deception using social media and focus on why online deception is so much easier than traditional deception. Tsikerdekis and Zeadally argue that ICT knowledge is an important factor because not only can it be used as a tool for deceivers but also as a defensive mechanism for victims. Deceivers, especially those possessing the technical know-how, are likely to look for easy targets that are less technologically-inclined and therefore easier to deceive, the authors say.
"Simply detecting deception, while a step in the right direction in this emerging area, is clearly not sufficient" Zeadally says. "Our long term goal in the area of deception in social media at UK is to develop novel techniques that can be deployed and used by software designers as well as social media users to prevent deception in the first place."
No entries found