Most of us think we have control over our online privacy with the security settings of social platforms – and some less trusting people opt out of social media altogether.
But two new studies, published in the journal Nature Human Behaviour, show you can’t even have a contacts list or listen to music without revealing personal information to Big Brother.
In the first study, led by James Bagrow from the University of Vermont in Burlington, US, researchers used information theory to amalgamate data from over 30 million Twitter posts by nearly 14,000 users.
Results showed they can predict our activities and interests from just eight to nine of our contacts with 95% accuracy.
That means it’s not just our own online activity that is telling – the study found that posts written by our online friends can predict just as much about our behaviour.
“It shows that information about you is embedded in your interactions with friends,” explains co-author Lewis Mitchell from the University of Adelaide in Australia.
“Just like if I overhear one side of a phone conversation it tells me something about the person on the other end of the line, so do your friends’ social media posts tell me something about you.
“And if you delete your account, it doesn’t necessarily help, because you can still in principle be profiled from the digital traces left by your interactions with friends.”
If a user avoids or leaves an online social network, a “shadow profile” can be created from information posted by social contacts. Indeed, Mark Zuckerberg admitted that Facebook collects information on people who are not users.
Other research has revealed that data derived from friends can be used to predict such private information as friendships, religion, whereabouts and sexual preferences, David Garcia writes in a related editorial.
“This all means that there is no place to hide in online social networks like Facebook and Twitter,” Mitchell says.
Although the digital age has yielded rich data for researchers, the privacy implications of these findings are a source of concern for users and non-users alike, the authors write.
The second study shows that even musical preferences can reveal personal information, including our age, where we’re from, and how we’re feeling.
Minsu Park and co-authors from Cornell University, US, analysed 765 million online music pieces streamed on Spotify by a million people in 51 countries to identify people’s emotional patterns.
“Big Music Data” lends itself particularly well to analysis, with estimates that people spend on average 44% of their waking hours listening to it.
“The omnipresence of music affords a singular opportunity to identify diurnal and seasonal patterns in listener’s musical choices,” the authors write.
Music can shed more light on people’s emotions than other sources because it reflects moods that people choose as well as their emotional status; “musical choice both shapes and reflects mood”.
Overall, they found that people prefer more relaxing music in the evening and more energetic music during the day – including the mid-afternoon slump.
There were differences between ages and groups – young people listen to more intense music, for instance, and “night owls” listen to less intense music. Latin American music is more stimulating while music played in Asia is more relaxing.
While the results “show a remarkable similarity to results based on sentiment analysis of Twitter messages”, there are differences.
“Positive emotion in Twitter messages dips around 15:00 while the consumption of arousing music does not, suggesting that music can also be used as a mid-afternoon stimulant.”
The implications of the latter research for individual privacy remain to be seen.
Nevertheless, tapping into people’s data is a social, rather than individual issue, says Mitchell.
“We need to stop thinking about individual privacy control,” writes Garcia, “and switch to a paradigm of networked privacy.