数据挖掘中匿名化隐私保护研究进展

 随着信息技术的发展,如何在保证数据高可用性的同时,不泄露数据主体的隐私信息,已日益引起国内外研究者的高度关注.隐私保护技术主要有数据加密、数据失真以及数据匿名化技术,其中匿名化技术是数据挖掘中隐私保护的最主要技术手段.围绕匿名技术的研究,国内外学者提出了多种匿名隐私保护模型,通过对其中4种主要模型,即k-匿名模型、l－多样性模型、(α,k)－匿名模型和t-closeness模型的分析比较,指出每种匿名模型的特点及优、缺点,并归纳了常用的匿名技术,总结了当前主要的匿名化质量的度量方法.未来匿名化技术作为数据挖掘中隐私保护的主要手段,还将面临着需要进一步解决的问题,对数据挖掘中匿名隐私保护的下一步研究方向进行了展望.

Progress in Anonymous Privacy-Preserving in Data Mining

With the development of information technology, an important issue is to ensure the high usability of the data and to protect the privacy. The privacy-preserving technology is related with the data encryption, the data distortion and the data anonymity in data mining. Among them, the most primary technology is the anonymous privacy-preserving technology. In that respect, a various privacy preserving models were proposed This paper focuses on the k-anonymous model, the l-diversity model, the (α, K)-anonymous model and the t-closeness model, and it is pointed out that each anonymous model has its, advantages and disadvantages. The commonly used anonymity technology and the major anonymous quality measurement methods are reviewed. In the future, the anonymity technology will face new problems, and the privacy-preserving will be further considered in data mining.