Estimating Identification Disclosure Risk Using Mixed Membership Models

Statistical agencies and other organizations that disseminate data are obligated to protect data subjects' confidentiality. For example, ill-intentioned individuals might link data subjects to records in other databases by matching on common characteristics (keys). Successful links are particul...

Celý popis

Uloženo v:

Podrobná bibliografie
Vydáno v:	Journal of the American Statistical Association Ročník 107; číslo 500; s. 1385 - 1394
Hlavní autoři:	Manrique-Vallier, Daniel, Reiter, Jerome P
Médium:	Journal Article
Jazyk:	angličtina
Vydáno:	United States Taylor & Francis Group 01.12.2012 Taylor & Francis Ltd
Témata:	Algorithms Applications and Case Studies Bayesian analysis Bias Censuses Confidentiality Contingency table Contingency tables data collection Data sampling Datasets Disclosure Estimate reliability Estimates Estimation methods False positive errors Grade of membership Latent class Linear analysis Linear models Markov analysis Markov chain Membership Membership groups Modeling Monte Carlo simulation Population Population estimates prediction Probability Public use Random sampling risk Risk assessment Statistics U.S.A Uniqueness United States United States Contingency table Grade of membership Latent class Disclosure Confidentiality
ISSN:	1537-274X, 0162-1459, 1537-274X
On-line přístup:	Získat plný text
Tagy:	Přidat tag Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!

Popis
Shrnutí:	Statistical agencies and other organizations that disseminate data are obligated to protect data subjects' confidentiality. For example, ill-intentioned individuals might link data subjects to records in other databases by matching on common characteristics (keys). Successful links are particularly problematic for data subjects with combinations of keys that are unique in the population. Hence, as part of their assessments of disclosure risks, many data stewards estimate the probabilities that sample uniques on sets of discrete keys are also population uniques on those keys. This is typically done using log-linear modeling on the keys. However, log-linear models can yield biased estimates of cell probabilities for sparse contingency tables with many zero counts, which often occurs in databases with many keys. This bias can result in unreliable estimates of probabilities of uniqueness and, hence, misrepresentations of disclosure risks. We propose an alternative to log-linear models for datasets with sparse keys based on a Bayesian version of grade of membership (GoM) models. We present a Bayesian GoM model for multinomial variables and offer a Markov chain Monte Carlo algorithm for fitting the model. We evaluate the approach by treating data from a recent U.S. Census Bureau public use microdata sample as a population, taking simple random samples from that population, and benchmarking estimated probabilities of uniqueness against population values. Compared to log-linear models, GoM models provide more accurate estimates of the total number of uniques in the samples. Additionally, they offer record-level predictions of uniqueness that dominate those based on log-linear models. This article has online supplementary materials.
Bibliografie:	http://dx.doi.org/10.1080/01621459.2012.710508 SourceType-Scholarly Journals-1 ObjectType-Feature-1 content type line 14 ObjectType-Article-2 content type line 23 ObjectType-Article-1 ObjectType-Feature-2
ISSN:	1537-274X 0162-1459 1537-274X
DOI:	10.1080/01621459.2012.710508