大数据环境下跨组织间协同优化决策的隐私保护算法  

Privacy-preserving Algorithm for Cross-organizational Collaborative Optimization Decisions Based on Big Data

作　　者：刘洪伟[1] 刘智慧[1] 朱慧[1] 陆涛[1] 

机构地区：[1]广东工业大学管理学院,广东广州510520

出　　处：《广东工业大学学报》2014年第3期21-26,共6页Journal of Guangdong University of Technology

基　　金：国家自然科学基金资助项目(70971027);广东省哲学社会科学规划项目(GD10YGL09);广东省普通高校人文社会科学重点研究项目(12ZS0112)

摘　　要：组织协同决策分析的数据具有大数据的分布性、异构性和隐私性等典型特征.安全多方计算是一种基于协同机制或协议的隐私保护算法,但它一些常用的单调张成等方法却无法挣脱计算复杂性的困扰.本文主要研究组织间两种结构的协同优化决策问题,提出针对决策变量与约束参数隐私保护的安全多方计算协议,并给出相对应的安全证明.研究表明对于本文构造的SMC协议,可以降低优化协同决策的计算复杂度,部分隐私信息无须加工传送也可以完成计算任务.The cross-organizational data has the typical characteristics of big data in collaborative optimi-zation decision-making , such as distributedness , heterogeneity , and privacy etc .Secure multi-party com-putation（SMC） is a privacy-preserving algorithm, based on collaborative mechanisms or protocols .How-ever , the methods typically used , such as monotone span program , cannot get rid of the computational complexity .It discussed two kinds of problems in collaborative optimization decision , and proposed se-cure multi-party computation protocols for the privacy-preserving of constraint parameters and decision variables.Then, it gave security proof.The results show that the SMC protocols can reduce the computa-tional complexity of collaborative optimization decisions , and computation of some private information can be completed without transferred processing .

关 键 词：大数据 协同优化 隐私保护 安全多方计算 

分 类 号：TP309.2[自动化与计算机技术—计算机系统结构] F224.31[自动化与计算机技术—计算机科学与技术]