Be careful what you "like" on Facebook. It can tell people a lot more than you'd like them to know.
That's what a trio of boffins at Cambridge University in the U.K. discovered after analyzing the like activity of more than 58,000 Facebook users who volunteered to participate in their research project.
"This study demonstrates the degree to which relatively basic digital records of human behavior can be used to automatically and accurately estimate a wide range of personal attributes that people would typically assume to be private," wrote the researchers in the Proceedings of the National Academy of Sciences.
[See also: Facebook's Graph Search worries security experts]
According to the study, likes could be used to accurately predict:
- Race (African Americans vs. Caucasians) in 95% of the cases;
- Gender in 93% of the cases;
- Sexual orientation for males (88%) and females (75%);
- Political party (Democrat vs. Republican) in 85% of the cases;
- Religion (Christian vs. Muslim) in 82% of the cases;
- Substance use 73% of the time;
- And relationship status 65% of the time.
"Predicting users' individual attributes and preferences can be used to improve numerous products and services," said the researchers, Michal Kosinski, David Stillwell and Thore Graepel.
The trio was well aware of the dark side of their findings, too.
"Commercial companies, governmental institutions, or even one's Facebook friends could use software to infer attributes such as intelligence, sexual orientation, or political views that an individual may not have intended to share," they wrote. "One can imagine situations in which such predictions, even if incorrect, could pose a threat to an individual's well-being, freedom, or even life."
While Facebook members don't expect their likes to be vacuumed up and crunched into information about them by people they don't know, they may be less sensitive about inferences made from their likes than might be believed, according Jonathan Zittrain, co-founder and co-director of the Berkman Center for Internet & Society at Harvard University.
"[T]he function of prominently liking something on Facebook is not merely to follow updates about it in one's newsfeed, but to express one's identity," he said via email. "The sensitive inferences the researchers found they could make are to qualities that perhaps people are increasingly comfortable expressing -- or having inferred about themselves."
He maintained, however, that in the long run, it would be helpful if Facebook members knew about any automated, large-scale scraping of their data analytical purposes.
The study is a great example of how little things performed online can create a detailed picture of who you are, said Sarah A. Downey, a privacy analyst and attorney with Abine, an online reputation company in Boston.
"You may not think that a like here and there says anything about you, but they all add up -- especially with Facebook's new Graph Search that displays all your likes with a single search," she said via email. "If you can learn this much about a person through their Facebook likes, imagine how companies or governments could use -- or misuse -- that data,."
The kinds of information the researchers discerned from Facebook likes could result in real danger to a person, said Adi Kamdar, an activist with the Electronic Frontier Foundation (EFF) in San Francisco.
"It can be especially dangerous for people in repressive countries or even in some states where acknowledging something like sexual orientation could lead to some really awful consequences," he said.