“If you look unique in how you make phone calls, it is possible to connect that to where you’ve made credit card purchases,” said researcher Chris Riederer of Columbia.
The team developed an algorithm that compares geotagged posts on Twitter with posts on Instagram and Foursquare to link accounts held by the same person.
It works by calculating the probability that one person posting at a given time and place could also be posting in a second app, at another time and place.
The algorithm can also identify shoppers by matching anonymous credit card purchases against logs of mobile phones pinging the nearest cell tower.
This method, they found, out-performs other matching algorithms applied to the same data sets.
“Many people choose not to identify themselves online,” said Professor Augustin Chaintreau of Columbia. “If I now tell you that your location data makes you recognisable across all of your accounts, how does that change your behavior? This is a question we now have to answer.”
Mobility data may also reveal someone’s age, gender and ethnicity.
“What this really shows is that simply removing identifying information from large-scale data sets is not sufficient,” MIT Media Lab researcher Yves-Alexandre de Montjoye told Columbia University after reading the findings. “We need to move to a model of privacy-through-security. Instead of anonymising data and making it public, there should be technical controls over who gets access to the data, how it is used, and for what purpose.”
Riederer, with Columbia undergraduates Danny Echikson and Stephanie Huang, has built a tool, called ‘You Are Where You Go’, that lets individuals audit their own social media trail.
It retraces steps on Twitter, Instagram and Foursquare and algorithms then make “relatively accurate”, said Columbia, inferences about your age, ethnicity, income, and whether you have kids.
The team is presenting its research today at the World Wide Web conference in Montreal.