A study conducted by a team of researchers from the Massachusetts Institute of Technology (MIT) shows that only there or four pieces of information of the shopping history of a person using a credit card can unveil who that person is. The study challenges retailer’ claims that their batch of data shared with third parties is anonymized. “Anonymized” yes, but not anonymous, researchers say.
The study also reveals that the fragments about a person’s shopping habit can unveil not only the identity of that person, but also everything that he or she has bought using the same credit card.
As it turns out, every person has a specific shopping pattern, which includes price range, shops, products and hours of the purchase. This pattern makes the person easily recognizable among millions of others “anonymous” shoppers
Yves-Alexandre de Montjoye, lead-author of the study, said that their research wanted to show the world what were the real limits of anonymization of credit card data, but also mobile and internet data.
For their study, the authors used a 3-month-long history of credit-card transaction of more than 1 million people living in one of the highly industrialized countries. The researchers refused to get more specific on which country.
They picked a set of purchases of several individuals, anonymized them by removing details about names and bank accounts, and inserted them into a database. The database showed only approximate costs of the purchase because spending exactly $59.55 at Walmart on a Tuesday would give the shopper away almost instantly. Also, the database didn’t take into account payments of more than $22,000, for the same reason.
Soon afterward, researchers compared three pieces of shopping information related to location, date, and prices about a non-annonymized person with the database. By doing so, the MIT scientists were able to identify the annonymized purchases of that person, which were include in the data base.
“You bought a coffee at that coffee shop, and you bought jeans at that shop, and then you bought a pizza,”
said Prof Montjoye about one person’s unique pattern of doing shopping.
Researchers said that their database had 94 percent accuracy in revealing the identity of the shoppers. If the price wasn’t given as a clue, researchers needed to study four purchases to identify you, rather than only three. And the accuracy dropped to 90 percent.
The study also revealed that women were more identifiable by looking at their shopping patterns than men. Researchers couldn’t provide an explanation. Also, rich people who used credit cards were easier to spot than poor people.
Image Source: ForeclosurePulse