摘要
针对传统电力用户核心大数据匿名化方法,存在用户识别精准度低、匿名效率差的问题,基于核熵成分分析研究了一种新的电力用户核心大数据匿名化方法。确定匿名信息来源方向,主要的来源方向有企业数据与9558信息处理数据两种,同时优化高性能计算技术、数据挖掘技术、数据可视化技术,基于三种技术实现电力用户核心大数据匿名化。为了验证方法的有效性,设定对比实验,结果表明,基于核熵成分分析的电力用户核心大数据匿名化方法,能够在短时间内精准地识别用户信息,实现匿名化处理。
Aiming at the problem of low accuracy of user identification and low efficiency of anonymity in traditional power user core big data anonymization method,a new method of power user core big data anonymization was studied based on kernel entropy component analysis.Determine the direction of anonymous information sources,including enterprise data and 9558 information processing data.Meanwhile,optimize high-performance computing technology,data mining technology and data visualization technology,and achieve the anonymity of core big data of power users based on three technologies.In order to verify the effectiveness of the method,comparative experiments were conducted.The results showed that the anonymization method of power users’core big data based on nuclear entropy component analysis could accurately identify user information in a short time and realize anonymization processing.
作者
张凡
张龙
宗震
ZHANG Fan;ZHANG Long;ZONG Zhen(State Grid Anhui Electric Power Co.,Ltd.Construction Company,Hefei 230071,China)
出处
《电子设计工程》
2020年第13期175-178,183,共5页
Electronic Design Engineering
基金
国家重点研究计划项目(2017YFA0500301)。
关键词
核熵成分分析
电力用户
核心大数据
匿名化研究
nuclear entropy component analysis
power users
core big data
anonymous research
