摘要
为确保隐私保护数据挖掘中所存在的维数灾难问题得到有效解决,文章提出了将基于随机投影技术的一种数据挖掘隐私保护法。这种方法对攻击者能够以随机投影矩阵推测的方式重建原始数据进行了综合考虑,首先将安全子空空间概念提出来,再构建安全子空间映射,在低失真嵌入实现的同时,能够有效确保数据安全。通过实验证明,在对数据隐私予以保护的前提下,这种方法为数据质量提供有效保障。
In order to ensure the privacy of data mining in the presence of the curse of dimensionality issues are effectively addressed,this article will dig Privacy Protection Act proposed is based on a random data projection technology.In this way the attacker can be presumed in a random manner as the projection matrix to reconstruct the original data were taken into account,first,the proposed concept of the security sub-blank space,and then build the security sub-space mapping,while embedded achieve low distortion,it is possible to ensure an effective data security.Through experiments proved to be protected under the premise of data privacy,this method provides effective protection for data quality.
出处
《安顺学院学报》
2015年第1期129-130,共2页
Journal of Anshun University
关键词
数据挖掘
随机投影技术
隐私保护
data mining
stochastic projection technology
privacy protection