Full Citation

Title: An Algorithm for l-diversity based on Randomized Addition of Sensitive Values

Citation Type: Journal Article

Publication Year: 2015

ISBN:

ISSN:

DOI:

NSFID:

PMCID:

PMID:

Abstract: Consideration of privacy is necessary when sharing an information database about an individual with other companies. l-In general anonymization technology such as diversity, an identifier that identifies an individual is excluded from the database, and the pseudo identifier (QID) is generalized to allow an attacker to estimate the attribute value of each individual. prevent. Anonymization is usually performed only once, and an anonymized database can be shared by multiple data users. Therefore, there is a possibility that a QID that a certain data user wants to analyze is completely generalized and cannot be analyzed at all. In this study, l-diversity is realized by adding dummy elements to sensitive attributes without generalizing QID. Therefore, each data user can freely analyze based on their favorite QID. Simulations using real data show that the proposed method has a higher trade-off between privacy and effectiveness than existing l-diversity methods. Existing anonymization techniques remove identifiers and generalize quasi-identifiers (QIDs) from the database.By doing so, adversaries cannot specify each individual's values ​​of the sensitive attributes. Because the database is anonymized based on one-size-fits-all measures in usual, it is possible that QIDs that a data user primarily on are all generalized, and the anonymized database has no value for the user.In this study, we proposed a new technique for l-diversity, which keeps QIDs unchanged so that data users can analyze it based on QIDs they focus on.Through simulations of real data sets,we prove that our proposed method can result in a better tradeoff between privacy and utility of the anonymized database.

Url: https://uec.repo.nii.ac.jp/?action=pages_view_main&active_action=repository_view_main_item_detail&item_id=9098&item_no=1&page_id=13&block_id=21

User Submitted?: No

Authors: Yuichi, Sei; Akihiko, Ohsuga

Periodical (Full): 情報処理学会論文誌

Issue: 5

Volume: 56

Pages: 1377-1387

Data Collections: IPUMS USA

Topics: Methodology and Data Collection, Other

Countries: