摘要
隐私保护的目标是要在不精确访问真实原始数据的条件下,如通过病历,对乳腺癌的确诊,得到准确的模型和分析结果。研究了分布式环境下构造贝叶斯网络结构的K2算法的评价函数,并且以密码学中的两方安全计算为基础对该函数进行改进。通过理论分析和实验结果说明了改进的评价函数能够得到与原始的评价函数相同的贝叶斯网络结构。达到了数据垂直分布情况下隐私保护的目的。
The purpose of privacy preservation was to discover accurate patterns without precise access to the original data.For example,the result of breast-cancer could be obtained according to diseases data.The paper researched the score function of K2 Algorithm for constructing the Bayesian Networks in the distributed settings,and improved the score function based on the two-party computation of the cryptographic preliminaries.It could generate the same structures whether original score function or improved score function by the theory analysis and experimental results.The purpose of privacy preserving was reached on vertically partitioned data.
出处
《武汉理工大学学报》
CAS
CSCD
北大核心
2008年第6期140-142,共3页
Journal of Wuhan University of Technology
关键词
隐私保护数据挖掘
贝叶斯网络
评价函数
privacy-preserving data mining
bayesian networks
score function