一种改进的隐私保护关联规则挖掘算法  

Improved Algorithm of Privacy Preserving Association Rule Mining

作　　者：顾铖[1] 朱保平[1] 张金康[1] 

机构地区：[1]南京理工大学计算机科学与工程学院,南京210094

出　　处：《南京航空航天大学学报》2015年第1期119-124,共6页Journal of Nanjing University of Aeronautics & Astronautics

摘　　要：部分隐藏的随机化回答方法是基于关联规则数据挖掘的隐私保护算法,针对该算法在重构频繁项集支持度上的指数级时间复杂度导致算法执行效率下降的不足,采用分治策略和集合运算方法对该算法进行改进,消除重构数据的指数级运算。改进算法降低了算法的时间复杂度并有效提高了执行效率。仿真实验与分析表明了改进算法的有效性。Randomized response with partial hiding is a privacy preservation algorithm based on associa- tion rule mining. For the insufficient that the exponential time complexity of the algorithm in the recon- struction of frequent item set support reduces the execution efficiency of the algorithm, the divide and conquer strategy and the set operations are used to improve the algorithm, and the exponential computa- tion in the reconstruction is eliminated. The improved algorithm can reduce the time complexity and promote the efficiency effectively. Experiments and analysis indicate the effectiveness of the improved algorithm.

关 键 词：关联规则 隐私保护 分治策略 集合运算 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]