.png){kind=icon}
.png){kind=icon}
Return
Efficient privacy preserving algorithms for hiding sensitive high utility itemsets
.png){kind=icon}
PRE
AI
DOI：10.1016/j.cose.2023.103360.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Utility-driven mining is a powerful data mining technique that aims to extract significant and valuable knowledge from different kinds of datasets. Nevertheless, analyzing datasets with sensitive or private information may raise security and privacy concerns. To balance utility maximization and privacy preservation, Privacy Preserving Utility Mining (PPUM) has been presented. The primary objective of PPUM algorithms is to conceal the sensitive knowledge that can be found via the application of utility mining algorithms to sensitive data. However, the current PPUM literature shows a paucity of privacy preserving algorithms scalable and efficient enough to handle large and dense datasets. Based on this perspective, this paper proposes three heuristic-based algorithms, namely Selecting the Most Real item sensitive utility First (SMRF), Selecting the Least Real item sensitive utility First (SLRF), and Selecting the most Desirable Item First (SDIF), to efficiently conceal all Sensitive High utility Itemsets (SHIs) while mitigating the projected detrimental impact on the non-sensitive information. The proposed algorithms rely on a novel concept, called Real Item Sensitive Utility (RISU), to effectively select a definite victim item for each SHI during the whole sanitization process. Furthermore, a new sensitive dataset sorting technique is proposed to reduce the time needed to find suitable transactions for sanitization. Through comprehensive experimental evaluations with state-of-the-arts, the viability of the proposed three algorithms was testified. The acquired findings clearly demonstrate the efficacy of the proposed algorithms in terms of reducing the sanitization time and side effects. & COPY; 2023 Elsevier Ltd. All rights reserved.
Keywords:
Privacy preserving
Data mining
Utility mining
Sensitive pattern
Privacy metrics
Data utility metrics
Journal
C
IF:
5.4
Papers:
4.6K
Citations:
1.4W