(α, k)-anonymity: an enhanced k-anonymity model for privacy preserving data publishing

By Raymond Chi-Wing Wong, Jiuyong Li, Ada Wai-Chee Fu, Ke Wang. Knowledge Discovery and Data Mining, 2006

728 citations recorded by the source corpus.

Privacy preservation is an important issue in the release of data for mining purposes. The k-anonymity model has been introduced for protecting individual identification. Recent studies show that a more sophisticated model is necessary to protect the association of individuals to sensitive information. In this paper, we propose an (α, k)-anonymity model to protect both identifications and relationships to sensitive information in data. We discuss the properties of (α, k)-anonymity model. We prove that the optimal (α, k)-anonymity problem is NP-hard. We first presentan optimal global-recoding method for the (α, k)-anonymity problem. Next we propose a local-recoding algorithm which is more scalable and result in less data distortion. The effectiveness and efficiency are shown by experiments. We also describe how the model can be extended to more general case.

Subfields: Artificial Intelligence

View Paper: https://www.semanticscholar.org/paper/c4f0bb8795e57d9f14a7e229e98350a02c4cf074

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

Citations in the corpus

Location Privacy Protection Based on Differential Privacy Strategy for Big Data in Industrial Internet of Things (2018, methodology): “[9] proposed the k-anonymity method and also a large number of methods based on k-anonymity [10]–[12].”
Slicing: A New Approach to Privacy Preserving Data Publishing (2009, background): “Finally, privacy measures for attribute disclosure protection include ℓ -diversity [23], ( α, k )-anonymity [34], t - closeness [20], ( k, e )-anonymity [16], ( c, k )-safety [25], privacy skyline [5], m -conﬁdentiality [33] and ( ǫ, m )- anonymity [19].”
Anonymizing Classification Data for Privacy Preservation (2007, methodology): “The efcienc y of the bottom-up cell generalization algorithm in [ 15 ] is comparable to TDR when k = 2; 10, but they did not report the efcienc y for larger k. A cell generalization algorithm in [20] took about 60 seconds to mask the data.”
On the tradeoff between privacy and utility in data publishing (2009, methodology): “First, one chooses a specific privacy requirement, such as k-anonymity [25, 26], -diversity [21], (α, k)anonymity [29], t-closeness [19], and δ-presence [23], based on intuitions of what privacy means.”
δ-Presence without Complete World Knowledge (2010, background): “Privacy definitions such as ‘-diversity [6], t-closeness [11], ð�;k Þ-anonymity [ 12 ], ... prevent linking individuals to specific values in a sensitive attribute (e.g., linking students to GPA information) by enforcing diversity in sensitive values of indistinguishable tuples.”
Data and Structural k-Anonymity in Social Networks (2009, neutral): “Similar models such as p-sensitive k-anonymity [26], l-diversity [17], (α, k)-anonymity [30], and t-closeness [14] exist for microdata.”
Privacy-preserving data publishing: A survey of recent developments (2010, methodology): “2005, 2007] VS IGPL minimal Progressive Local Recoding [Wong et al. 2006] CG MD minimal i-Diversity Incognito [Machanavajjhala et al. 2007] FG,RS MD,DM optimal InfoGain Mondrian [LeFevre et al. 2006b] MG IG minimal Anatomy [Xiao and Tao 2006a] AM heuristics minimal (k,e)-Anonymity Permutation…”
Differentially private data release for data mining (2011, background): “Professional [18-65) Professional [18-65) Professional [18-65) Professional [18-65)”
Technical Privacy Metrics (2015, background): “onymity extendsk-anonymity with the additional requirement that in any equivalence class E (rows that have the same quasi-identifier values), the frequency of a sensitive value s has to be less than α[145, 148]. As a result, no single sensitive attribute can be dominant in an equivalence class. priv”