隐私保护的时序规则分布挖掘  

Temporal Rule Distribution Mining of Privacy-preserving

作　　者：耿波[1] 仲红[1] 彭俊[1] 王大刚[1] 

机构地区：[1]安徽大学计算机信息学院,合肥230039

出　　处：《计算机工程》2008年第24期69-70,共2页Computer Engineering

基　　金：国家自然科学基金资助项目(60773114);安徽省自然科学基金资助项目(070412051);安徽高校省级重点自然科学研究基金资助项目(KJ2007A43)

摘　　要：针对多方分别拥有一部分数据记录,并想在不泄露自己隐私数据的情况下联合对各自的时序序列进行分割的问题,提出基于半可信第三方的隐私保护的时序规则分布挖掘方法。将联合计算时序规则各频度的问题转化成多方秘密比较数大小的问题,从而设计一个简单的基于半可信第三方的算法,解决多方联合计算时序规则频度的问题。This paper brings forward such a problem： there are multiple partners, one of them owns part of records, and no one has all of the records, then they want to divide their temporal sequence without leaking their private information. For this, it proposes a method for finding temporal sequence rule distribution based on untrusted third-party. Then the problem of calculating the frequency by multiple partners together will be converted into the problem of the comparing size of figures. Accordingly it designs an arithmetic based on untrusted third-party to resolve the problem of calculating the frequency.

关 键 词：时序序列 规则分布 隐私保护 半可信第三方 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]