Academic
Publications
Mondrian Multidimensional K-Anonymity

Mondrian Multidimensional K-Anonymity,10.1109/ICDE.2006.101,Kristen Lefevre,David J. Dewitt,Raghu Ramakrishnan

Mondrian Multidimensional K-Anonymity   (Citations: 267)
BibTex | RIS | RefWorks Download
K-Anonymity has been proposed as a mechanism for pro- tecting privacy in microdata publishing, and numerous re- coding "models" have been considered for achieving k- anonymity. This paper proposes a new multidimensional model, which provides an additional degree of flexibility not seen in previous (single-dimensional) approaches. Often this flexibility leads to higher-quality anonymizations, as measured both by general-purpose metrics and more spe- cific notions of query answerability. Optimal multidimensional anonymization is NP-hard (like previous optimal k-anonymity problems). However, we introduce a simple greedy approximation algorithm, and experimental results show that this greedy algorithm frequently leads to more desirable anonymizations than exhaustive optimal algorithms for two single-dimensional models.
Cumulative Annual
    • ...We can adapt the greedy approach of MON-DRIAN [21] proposed fork-anonymization by starting with the vector (1;:::; 1) and gradually reducing the suppression probabilities until reducing any suppression probability further would violate privacy...

    Michaela Goetzet al. MaskIt: privately releasing user context streams for personalized mobi...

    • ...A wellknown technique to achieve k-anonymity, exploited by many of the methods proposed so far (e.g., [8], [18], [21], [22], [28]), is generalization...
    • ...Among these approaches, we can find methods based on perturbation [2], [5], [20] and k-anonymity [8], [15], [18], [21], [22], [28]...
    • ...(a) Varying QI. (b) Varying K. been proposed [8], [15], [18], [21], [22], [28]...

    Jianneng Caoet al. CASTLE: Continuously Anonymizing Data Streams

    • ...Mondrian [15] partitions the space recursively across the dimension with the widest normalized range of values and supports a limited version of local recoding...
    • ...An example of global recoding has already been shown in Fig. 2 .T he LRA algorithm uses an alternative local recoding generalization [2,15] that applies selected generalization rules to a subset of the transactions and results in a database where items are generalized at different levels in different transactions...

    Manolis Terrovitiset al. Local and global recoding methods for anonymizing set-valued data

    • ...However, the two extant schemes for t-closeness [18,19] are extensions of algorithms designed for k-anonymity; they employ either the Incognito [14 ]o r the Mondrian [15] technique for k-anonymization, merely adding to them the extra condition that the produced ECs should satisfy t-closeness...
    • ...The flexibility of local recoding allows for anonymizations of higher information quality [10,14,15]...
    • ...Furthermore, a single-dimensional recoding considers the domain of each QI attribute independently of the others [14] (hence forms a grid over the combined QI domains); on the other hand, a multidimensional recoding freely defines ECs over the combined domains of all QI attributes [15]...
    • ...Likewise, the second t-closeness-obtaining scheme [19] extends the Mondrian k-anonymization method [15]...
    • ...The same tree structure has also been employed in Mondrian [15]...
    • ...We compare SABRE against tIncognito [18] and tMondrian [19], i.e., the t-closeness schemes extended from Incognito [14] and Mondrian [15], respectively...

    Jianneng Caoet al. SABRE: a Sensitive Attribute Bucketization and REdistribution framewor...

    • ... to meet a privacy principle [k-anonymity (Sweeney 1997; Samarati 2001), l-diversity (Machanavajjhala et al. 2006), (α, k)-anonymity (Wong et al. 2006), t-closeness (Li and Li 2007)] using techniques such as generalization, suppression (removal), permutation and swapping of certain data values while minimizing certain cost metrics (Iyengar 2002; Wang et al. 2004; Meyerson and Williams 2004; Bayardo and Agrawal 2005; Fung et al. 2005; LeFevre et ...
    • ...We derive the estimated answer of a query using the approach explained in LeFevre et al. (2006a)...

    Xiaoxun Sunet al. Publishing anonymous survey rating data

Sort by: