期刊名称：International Journal of Innovative Research in Science, Engineering and Technology
印刷版ISSN：2347-6710
电子版ISSN：2319-8753
出版年度：2013
卷号：2
期号：4
页码：1230
出版社：S&S Publications
摘要：Several anonymization techniques, likegeneralization and bucketization, have been intendedfor privacy preserving microdata publishing. currentwork has shown that generalization loses significantamount of information, particularly forhigh-dimensional data. on the other hand,Bucketization does not prevent membershipdisclosure and does not apply for data that do not havea clear separation between quasi-identifying attributesand sensitive attributes. In this paper, we present anew technique called slicing, in that data is partitioninto both horizontally and vertically. We demonstratethat slicing preserves better data utility thangeneralization and can be used for membershipdisclosure protection. Another main advantage ofslicing is that it can handle high-dimensional data. Weillustrate how slicing can be used for attributedisclosure protection and build up an efficientalgorithm for computing the sliced data that obey theℓ-diversity requirement. Our workload experimentsverify that slicing preserves better utility thangeneralization and is more effective than bucketizationin workloads involving the sensitive attribute. Ourexperiments also show that slicing can be used toprevent membership disclosure.
