用户终端隐私大数据交互式保护算法
Interactive Protection Algorithm for User Terminal Privacy Big Data
赵振兴 1段富2
作者信息
- 1. 新泰市人民政府办公室,山东 新泰 271200;太原理工大学计算机科学与技术学院,山西 太原 030000
- 2. 太原理工大学计算机科学与技术学院,山西 太原 030000
- 折叠
摘要
用户终端隐私数据泄露,大量隐私信息被不法分子利用,导致用户隐私安全受到严重威胁.为了提高用户终端隐私数据的安全性,提出一种用户终端隐私大数据交互式保护算法.采用互补集合经验模态分解(Complementary Ensemble Empirical Mode Decomposition,CEEMD)算法结合小波阈值函数对用户终端隐私数据去噪处理,并通过启发式搜索算法和最大信息数建立特征选择模型,获取隐私数据的特征;采用差分隐私保护算法对隐私数据保护,通过K-means算法划分特征并计算敏感度,使用拉普拉斯机制将噪声加入敏感度较高的特征中,实现隐私数据的保护.实验结果表明,所提算法的数据隐私性强、可用性高,且对数据加密所用时间短.
Abstract
The leakage of privacy data in user terminal and the exploitation of privacy information by illegal ele-ments cause a serious threat on privacy and security.In order to improve the security of privacy data in user terminal,this paper put forward an interactive protection algorithm for privacy big data in user terminal.Firstly,we combined the Complementary Ensemble Empirical Mode Decomposition(CEEMD)algorithm with wavelet threshold function to denoise the privacy data,and built a feature selection model through heuristic search algorithm and maximum informa-tion count,thus obtaining the features of privacy data.Moreover,we used differential privacy protection algorithm to protect private data,and then adopted the K-means algorithm to partition features and calculate the sensitivity.Final-ly,we added noise to the features with higher sensitivity by Laplace mechanism,thus achieving the privacy data pro-tection.The experimental results show that the proposed algorithm has strong data privacy,high availability,and short encryption time for data.
关键词
网络终端/小波阈值函数/隐私大数据/差分隐私保护算法Key words
Network terminal/Wavelet threshold function/Privacy big data/Differential privacy protection algo-rithm引用本文复制引用
出版年
2024
NETL
NSTL
万方数据