Even as Facebook moves to improve and simplify privacy options for the social network’s 750 million users, a group of independent researchers is proposing even greater security measures, specifically for those social networks that place control and responsibility squarely in the hands of users.
The researchers presented their findings at the ACM Conference on Knowledge Discovery and Data Mining in San Diego. They’ve created an index that lets users determine how vulnerable their Facebook friends are to the myriad of attacks occurring on social networks and plan to develop an app based on their research.
The researchers looked at two million Facebook users and assigned a vulnerability index to each account based not only on the individual’s behavior and privacy settings but those of their friends as well. The upshot is that an individual’s privacy is only as secure as the weakest link (person) in that person’s network of friends.
“The solution,” says Pritam Gundecha, a doctoral student in computer science at Arizona State University and one of the authors of the study, “is to unfriend those with questionable behaviors or friends who have not set their privacy controls to acceptable levels.”
“If you don’t wish to unfriend someone, you can make them aware of their vulnerabilities and to ask them to address them,” says Gundecha, whose studies focus on social media security.
Professor Huan Liu of the School of Computing, Informatics, and Decision Systems Engineering at ASU and other researchers hope to develop a Facebook application that will let users see the privacy attributes of their friends.
“The work is based on a relatively simple mathematical model that uses public information,” says Gundecha.
Some types of malicious activities occurring on social networks:
- Malicious scripts
- Hacked accounts
- Malicious tagging
- Hacking into anonymous data to extract personal user information
- Sybil attacks that involve the creation of false IDs to carry out malicious activities