Impeding individual user profiling in shopper loyalty programs

  • Authors:
  • Philip Marquardt;David Dagon;Patrick Traynor

  • Affiliations:
  • Converging Infrastructure Security (CISEC) Laboratory, Georgia Tech Information Security Center (GTISC), School of Computer Science, Georgia Institute of Technology, Atlanta, GA;Converging Infrastructure Security (CISEC) Laboratory, Georgia Tech Information Security Center (GTISC), School of Computer Science, Georgia Institute of Technology, Atlanta, GA;Converging Infrastructure Security (CISEC) Laboratory, Georgia Tech Information Security Center (GTISC), School of Computer Science, Georgia Institute of Technology, Atlanta, GA

  • Venue:
  • FC'11 Proceedings of the 15th international conference on Financial Cryptography and Data Security
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

Shopper loyalty club programs are advertised as a means of reducing prices for consumers. When making a purchase, a customer simply scans their keyring tag along with the items they intend to buy and is granted a reduction in the total price. While the use of these cards results in a visible reduction in price, customers are largely unaware of the privacy implications of such discounts. In particular, the ability to link all purchases made by an individual customer allows retailers to develop detailed profiles that may reveal sensitive information, especially if leaked or sold to third parties. In this paper, we present ShopAnon, a mobile phone-based infrastructure designed to help consumers partake in shopper loyalty programs without allowing their transactions to be linked by a retailer. ShopAnon displays legitimate but random barcodes for specific retailers on each execution, and provides a number of operational modes that respond to the changing availability of resources and the specific privacy concerns of the user. Communications between the application and the database storing the barcodes occurs using an Oblivious Transfer protocol to prevent our system from exposing the barcode received by a requester. We design, implement and characterize the behavior of our application on the iPhone mobile platform, and demonstrate its practical efficiency (i.e., the ability to render random tags in less than 0.25 seconds via 802.11 links and approximately 3.9 seconds via a 3G cellular connection). Through this, we provide a powerful tool through which customers can improve their privacy in a retail environment.