Key points are not available for this paper at this time.
The problem of statistical database confidentiality in releasing microdata is addressed through the use of approximate data-swapping. Here a portion of the microdata is replaced with a database that has been selected with approximately the same statistics. The result guarantees the confidentialityof the original data, while providing microdata with accurate statistics. Methods for achieving such transformations are considered and analyzed through simulation.
Steven P. Reiss (Tue,) studied this question.