"Local Recoding by Maximum Weight Matching for Disclosure Control of Microdata Sets"
We propose "local recoding" as a new technique for controlling disclosure risk of microdata sets. Compared to the technique of global recoding, where the observed values are grouped into broader intervals or categories throughout the data set, in local recoding different grouping is performed for each observation when necessary. As a means of performing local recoding we propose to form pairs of close individuals and recode observed values within each pair. For optimally forming pairs we can employ Edmonds' algorithm (Edmonds(1965)) of maximum weight matching. We illustrate the technique by applying it to the Japanese vital statistics data.
Year of publication: |
1999-02
|
---|---|
Authors: | Takemura, Akimichi |
Institutions: | Center for International Research on the Japanese Economy (CIRJE), Faculty of Economics |
Saved in:
freely available
Saved in favorites
Similar items by person
-
Some Superpopulation Models for Estimating the Number of Population Uniques
Takemura, Akimichi, (1997)
-
James-Stein Type Estimator by Shrinkage to Closed Convex Set with Smooth Boundary
Satoshi, Kuriki, (1997)
-
Takemura, Akimichi, (1999)
- More ...