(τ, m)‐slicedBucket privacy model for sequential anonymization for improving privacy and utility

Abstract

In a real-world scenario for privacy-preserving data publishing, the original data is anonymized and released periodically. Each release may vary in number of records due to insert, update, and delete operations. An intruder can combine i.e. correlate different releases to compromise the privacy of the individual records. Most of the literature, such as -safety, -safe (l, k)-diversity, have an inconsistency in record signatures and adds counterfeit tuples with high generalization that causes privacy breach and information loss. In this paper, we propose an improved privacy model ( , )-slicedBucket, having a novel idea of "Cache" table to address these limitations. We indicate that a collusion attack can be performed for breaching the privacy of -safe (l, k)-diversity privacy model, and demonstrate it through formal modeling. The objective of the proposed ( , )-slicedBucket privacy model is to set a tradeoff between strong privacy and enhanced utility. Furthermore, we formally model and analyze the proposed model to show that the collusion attack is no longer applicable. Extensive experiments reveal that the proposed approach outperforms the existing models.

A Taxonomy of Syntactic Privacy Notions for Continuous Data Publishing

“…Regarding anonymization methods or strategies, the generation of counterfeits (i.e., fake records) may imply a significant danger to the utility of the dataset. Searching for a solution without counterfeits that preserves high utility or investigating their necessity and the guarantees that can be only guaranteed with them are naturally two open strands of research of utmost importance, with initial proposals in [41] and [66].…”

Improved angelization technique against background knowledge attack for 1:M microdata

“…For future work considerations, the proposed algorithm can be extended to implement privacy in a dynamic data publishing scenario ( Xiao & Tao, 2007 ; Khan et al., 2020b ) for periodic or non-periodic updates. Similarly, the proposed work can be extended to a cluster based anonymization technique to more efficiently overcome the problem of privacy and utility paradigm ( Safi & Hwang, 2022 ).…”

m-Eligibility With Minimum Counterfeits and Deletions for Privacy Protection in Continuous Data Publishing

“…Another related work is [18], which propose an alternative implementation of τ -safety that improves utility and privacy by considering non-consecutive reinsertions of tuples. Their main proposal is the creation of a Cach table.…”

“…In continuous data publishing, the main syntactic mechanisms are based on the m-invariance notion [12] and their variations [13], [14], [15], [16], [17], [18]. This notion is deeply related to the m-eligibility property [2], that is, that the dataset has no more than 1 m fraction of tuples with the same sensitive value.…”

“…As an alternative to adding counterfeits, the authors of [18] present a new approach based on partial publication. The main idea is to extract one tuple for each distinct sensitive value in the dataset to create a so-called Cach table.…”