基于隐私保护的关联规则挖掘算法  

Association Rules Mining Algorithm Based on Privacy Preserving

作　　者：欧阳金亮[1] 陆黎明[1] 

机构地区：[1]上海师范大学信息与机电工程学院,上海200234

出　　处：《计算机与数字工程》2010年第8期55-57,61,共4页Computer & Digital Engineering

摘　　要：基于隐私保护的关联规则挖掘在挖掘项集之间的相关联系的同时,可以保护数据提供者的隐私。基于数据变换法,提出使用高效数据结构即倒排文件的隐私保护关联规则挖掘算法IFB-PPARM。针对特定的敏感规则以及给定的最小支持度和置信度,得到所需要修改的敏感事务并对其做适当的处理。算法只需对事务数据库做一次扫描,并且所有对事务的处理操作都在事务数据库映射成的倒排文件中进行。分析表明,该算法具有较好的隐私性和高效性。Association rules mining based on privacy preserving can protect privacy of data provider while mining relations between itemsets.Based on data-distort method,we propose privacy preserving association rules mining algorithm IFB-PPARM using efficient data structure namely inverted file.For specific sensitive rules and given minimum support and confidence,get the sensitive transactions needing to modify and make appropriate treatment on them.The algorithm only need scan the transaction database once,and all the transaction operations are carried out on the inverted file mapped from transaction database.Analysis shows that the algorithm has good privacy and efficiency.

关 键 词：隐私保护 关联规则 倒排文件 敏感规则 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]