Data Swapping: A Risk-Utility Framework and Web Service Implementation (2003)

Introduction

Data swapping is a statistical disclosure limitation (SDL) technique that works at the microdata (individual data record) level. Confidentiality protection is achieved by selectively modifying a fraction of the records in the database by switching a subset of attributes between selected pairs of records. In this paper, we describe:

• A risk–utility (RU) framework for selecting among candidate swapped releases corresponding, for example, to different choices of the swapping attributes and swap rate;
• NISS WebSwap, a Web service that performs data swapping, uploading the original data file and swapping specification from the user’s computer and downloading the file containing the (possibly) swapped records, as well as a log file.
• The NISS WebSwap graphical user interface (GUI), which will be embedded in a data swapping toolkit (DSTK) currently under development at the National Institute of Statistical Sciences (NISS).