关联规则隐藏算法的研究  

Research of Association Rule Hiding Algorithm

作　　者：丁小刚[1] 黄伟伟[1] 柏文阳[1] 

机构地区：[1]南京大学计算机科学与技术系计算机软件新技术国家重点实验室,江苏南京210093

出　　处：《计算机应用研究》2006年第6期28-30,共3页Application Research of Computers

基　　金：国家"863"计划资助项目(2002AA141091)

摘　　要：数据挖掘能从不同角度、不同抽象层上看待数据,这将潜在地影响数据的私有性和安全性。着重介绍了关联规则数据挖掘中的规则隐藏算法,提出了一个改进的关联规则隐藏算法OSA,该算法综合采用项的添加和约束方法来降低关联规则的支持度和置信度,从而达到规则隐藏的目的。Data mining may pose a threat to privacy and data security since from non-sensitive data one is able to infer sensitive information, including personal information, facts or even patterns that are not supposed to be disclosed. With the increasing popular use of the World Wide Web and recent advances in data mining techniques and machine learning, there has been much interest recently on privacy preservation in data mining and issues related to data mining and security have been recognized and investigated. This paper first introduces the development of the privacy preservation and the principle of association rule and its realization by general algorithms, then proposes some improvements of the algorithm. Some experiments show these improvements proposed in this paper can realize the association rule hiding algorithm efficiently and quickly.

关 键 词：数据挖掘 关联规则挖掘 频繁项集 敏感规则隐藏 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]