摘要
基于随机响应技术,提出了一种在保护隐私的关联规则挖掘中对包含三个属性值的序数型数据进行伪装的方法,设计了在伪装的数据集上进行挖掘的算法;分析了算法的安全性及复杂度。并通过实验表明,该算法在伪装的数据集上挖掘出来的规则与原规则相比,相对误差不超过5%。
A data disguised method for privacy preserving ordinal associate rule mining based on the randomized response techniques is introduced. The mining algorithm gets on the disguised item set and analyzes the security and complexity of this algorithm. The experiments show the rule that this algorithm gets fewer relative error which is less than 5% compared with the original rules.
出处
《科学技术与工程》
2006年第24期3863-3866,3870,共5页
Science Technology and Engineering
基金
甘肃省自然科学基金(3Z5051-A25-037)资助
关键词
随机响应
关联规则
数据挖掘
association rule randomized response data mining