n-cycle swapping for the American community survey

Authors:
Michael DePersio;Marlow Lemons;Kaleli A. Ramanayake;Julie Tsay;Laura Zayatz
Affiliations:
Center for Disclosure Avoidance Research, U.S.Census Bureau, Washington, D.C.;Center for Disclosure Avoidance Research, U.S.Census Bureau, Washington, D.C.;Center for Disclosure Avoidance Research, U.S.Census Bureau, Washington, D.C.;Center for Statistical Research & Methodology, U.S.Census Bureau, Washington, D.C.;Center for Disclosure Avoidance Research, U.S.Census Bureau, Washington, D.C.
Venue:
PSD'12 Proceedings of the 2012 international conference on Privacy in Statistical Databases
Year:
2012

Citing 3
Cited 0

k-anonymity: a model for protecting privacy

International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems
Data ShufflingA New Masking Approach for Numerical Data

Management Science
Why swap when you can shuffle? a comparison of the proximity swap and data shuffle for numeric data

PSD'06 Proceedings of the 2006 CENEX-SDC project international conference on Privacy in Statistical Databases

Quantified Score

Hi-index	0.00

Visualization

Abstract

Data from the American Community Survey (ACS) are collected under the authority of Title 13 of the United States Code which guarantees the confidentiality of all survey respondents. To maintain this confidentiality while still being able to release usable data, the U.S. Census Bureau applies several disclosure avoidance methods. For the ACS, data swapping techniques are used to protect records deemed "at risk". Households are identified as at risk if they are unique on attributes selected and predetermined by the confidentiality and survey management staff. After being uniquely identified (or "flagged"), data swapping is used to exchange the geographic information of a flagged household with another flagged household. This study compared the effectiveness of the pair-swapping method currently used with a proposed n-Cycle swapping method. Specifically, the goal was to maintain the same level of disclosure protection while outputting data with less perturbation.