面向数据隐私差异的隐私保护数据发布方法

Privacy protection data publishing method for data privacy differences

针对关系型数据中多维敏感属性隐私差异所引起的隐私保护效用降低问题,提出了一种能有效表达多维敏感属性隐私差异的隐私保护数据发布方法.基于一种多维桶分组技术(MSB)对数据集的多维敏感属性隐私差异以及记录价值进行量化区分,给出记录分组优先级参数的计算方法,进而可实现基于记录分组优先级参数多维桶记录分组(TPSB)算法的隐私保护数据发布.实验结果表明:在权重参数合理赋值条件下,该方法在保证数据发布效率的同时可有效提升数据发布的质量.

Aiming at the problem of privacy protection utility caused by the difference of privacy of multi-dimensional sensitive attributes in relational data,a privacy protection data publishing method that can effectively express the privacy difference of multi-dimensional sensitive attributes was proposed.The method quantified the multi-dimensional sensitive attribute privacy difference and the record value of the data set,gave the calculation method of record packet priority parameter,which was based on a multi-dimensional bucket grouping technique called multi-sensitive bucketization(MSB).Then,privacy protection data publishing based on the record packet priority parameter multi-sensitive bucketization(TPSB)algorithm was implemented.Experiments show that under the condition of reasonable weighting parameters,this method can effectively improve the quality of data publishing while ensuring data publishing efficiency.