当前位置: X-MOL 学术Future Gener. Comput. Syst. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Efficient algorithms for victim item selection in privacy-preserving utility mining
Future Generation Computer Systems ( IF 7.5 ) Pub Date : 2021-10-19 , DOI: 10.1016/j.future.2021.10.008
Shalini Jangra 1 , Durga Toshniwal 1
Affiliation  

High-utility itemset mining has evolved as an essential and captivating research topic. It aims to extract the patterns/itemsets having high utility value; hence, they are called high utility itemsets (HUIs). From a business perspective, a utility can be the benefit associated with the sale of a particular item or the usefulness or satisfaction that a customer experiences from a product. The economic utilities are helpful to evaluate the drivers behind a customer’s purchase decision. The advances in information technology have enabled us to access the datasets related to various domains like health care, stock market, market-basket, education and bioinformatics. Companies strive to increase the utility value of their products and share their customer’s transactions data to extract high utility patterns to achieve global customer insights. However, this can lead to massive security and privacy risk if their competitors misuse the patterns that can disclose their confidential information. Privacy-preserving utility mining (PPUM) is a branch of privacy-preserving data mining (PPDM) that presents various algorithms which intend to hide sensitive high utility itemsets (SHUIs) and maintain a balance between utility-maximizing and privacy-preserving. In this paper, two SHUIs hiding algorithms, MinMax and Weighted, are proposed with three variants of each algorithm. Experiments on various datasets show that proposed algorithms perform better than the existing SHUIs hiding algorithms as fewer distortions of non-sensitive knowledge occur. This study uses six performance evaluating metrics to assess the proposed algorithms against compared algorithms.



中文翻译:

隐私保护效用挖掘中受害者项目选择的有效算法

高效用项集挖掘已经发展成为一个重要且引人入胜的研究课题。旨在提取具有高实用价值的模式/项集;因此,它们被称为高效用项集(HUI)。从商业角度来看,效用可以是与特定商品的销售相关的收益,也可以是客户从产品中体验到的有用性或满意度。经济效用有助于评估客户购买决定背后的驱动因素。信息技术的进步使我们能够访问与医疗保健、股票市场、市场篮子、教育和生物信息学等各个领域相关的数据集。公司努力提高其产品的效用价值并共享其客户的交易数据以提取高效用模式以实现全球客户洞察。然而,如果他们的竞争对手滥用可能泄露其机密信息的模式,这可能会导致巨大的安全和隐私风险。隐私保护效用挖掘(PPUM)是隐私保护数据挖掘(PPDM)的一个分支,它提出了各种算法,旨在隐藏敏感的高效用项集(SHUI)并在效用最大化和隐私保护之间保持平衡。在本文中,提出了两种 SHUI 隐藏算法 MinMax 和 Weighted,每种算法的三种变体。在各种数据集上的实验表明,所提出的算法比现有的 SHUI 隐藏算法性能更好,因为非敏感知识的失真更少。本研究使用六个性能评估指标来评估所提出的算法与比较算法。

更新日期:2021-10-28
down
wechat
bug