差分隐私模糊聚类位置保护方法

doi:10.16180/j.cnki.issn1007-7820.2022.11.010

电子科技 ›› 2022, Vol. 35 ›› Issue (11): 64-71.doi: 10.16180/j.cnki.issn1007-7820.2022.11.010

差分隐私模糊聚类位置保护方法

林静,胡德敏,王揆豪

上海理工大学光电信息与计算机工程学院,上海 200093

收稿日期:2021-03-28 出版日期:2022-11-15 发布日期:2022-11-11
作者简介:林静(1996-),女,硕士研究生。研究方向:隐私保护、信息安全。|胡德敏(1963-),男,博士,副教授。研究方向:计算机网络、分布式计算、云计算|王揆豪(1994-),男,硕士研究生。研究方向:计算机视觉、深度学习。
基金资助:
国家自然科学基金(61170277);国家自然科学基金(61472256);上海市教委科研创新重点项目(12zz137);上海市一流学科建设项目(S1201YLXK)

Differential Privacy Fuzzy Clustering Location Protection Method

LIN Jing,HU Demin,WANG Kuihao

School of Optical-Electrical and Computer Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China

Received:2021-03-28 Online:2022-11-15 Published:2022-11-11
Supported by:
National Natural Science Foundation of China(61170277);National Natural Science Foundation of China(61472256);Key Project of Scientific Research and Innovation of Shanghai Municipal Education Commission(12zz137);Shanghai First Class Discipline Construction Project(S1201YLXK)

摘要/Abstract

摘要：

针对现有差分隐私聚类位置保护方法存在初始值敏感、离散数据不适用、误差较大的问题,文中提出了一种差分隐私模糊聚类位置保护方法。首先,通过高斯核函数将点映射到特征空间,由于核函数计算量相对较小,计算效率有了显著提升;然后,将差分隐私与改进的模糊C均值聚类算法相结合,使得每一组输入数据不再仅隶属于某一特定的类,而是以隶属程度来表现;最后,文中将满足差分隐私约束的拉普拉斯噪声添加到聚类集合的质心点中,得到每个点的扰动位置,并使用扰动位置进行查询。实验结果表明,在保障位置隐私安全的前提下,差分隐私模糊聚类位置保护方法降低了查询误差,提升了算法效率。

关键词: 差分隐私, 隐私保护, 拉普拉斯机制, 核函数, 位置保护, 聚类算法, DPK-F, KFCM

Abstract:

In view of the problems of sensitive initial value, inapplicability of discrete data and large error in the existing differential privacy clustering location protection methods, a differential privacy fuzzy clustering location protection method is proposed in this study. Firstly, the points are mapped to the feature space by Gaussian kernel function, and the computational efficiency is significantly improved due to the relatively small amount of computation. Secondly, the differential privacy is combined with the improved fuzzy C-means clustering algorithm, so that each group of input data no longer belongs to a specific class, but to the degree of membership. Finally, the Laplacian noise satisfying the differential privacy constraint is added to the centroid of the clustering set to obtain the disturbance position of each point, and the disturbance position is used to query. The experimental results show that the differential privacy fuzzy clustering location protection method reduces the query error and improves the efficiency of the algorithm under the premise of ensuring the location privacy security.

Key words: differential privacy, privacy preserving, Laplacian mechanism, kernel function, location preserving, clustering algorithm, DPK-F, KFCM

中图分类号:

TP309

林静,胡德敏,王揆豪. 差分隐私模糊聚类位置保护方法[J]. 电子科技, 2022, 35(11): 64-71.

LIN Jing,HU Demin,WANG Kuihao. Differential Privacy Fuzzy Clustering Location Protection Method[J]. Electronic Science and Technology, 2022, 35(11): 64-71.

图/表 7

图1

图2

图3

图4

图5

图6

图7

参考文献 18

[1]	Bao J, He T, Ruan S, et al. Planning bike lanes based on sharing-bikes' trajectories[C]. Halifax: Proceedings of the Twenty-third ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2017.
[2]	李凤华, 张翠, 牛犇, 等. 高效的轨迹隐私保护方案[J]. 通信学报, 2015, 36(12):114-123.
	Li Fenghua, Zhang Cui, Niu Ben, et al. Efficient scheme for user's trajectory privacy[J]. Journal on Communication, 2015, 36(12):114-123.
[3]	赵婧, 张渊, 李兴华, 等. 基于轨迹频率抑制的轨迹隐私保护方法[J]. 计算机学报, 2014, 37(10):2096-2106.
	Zhao Jing, Zhang Yuan, Li Xinghua, et al. A trajectory privacy protection via trajectory frequency suppression[J]. Chinese Journal of Computers, 2014, 37(10):2096-2106.
[4]	朱维军, 游庆光, 杨卫东, 等. 基于统计差分的轨迹隐私保护[J]. 计算机研究与发展, 2017, 54(12):2825-2832.
	Zhu Weijun, You Qingguang, Yang Weidong, et al. Trajectory privacy preserving based on statistical differential privacy[J]. Journal of Computer Research and Development, 2017, 54(12):2825-2832.
[5]	Nergiz M E, Atzori M, Saygin Y. Towards trajectory anonymization: A generalization-based approach[J]. Transactions on Data Privacy, 2008, 2(1):52-61.
[6]	Domingo-Ferrer J, Trujillo-Rasua R. Microaggregation and permutation-based anonymization of movement data[J]. Information Sciences, 2012, 20(8):55-80.
[7]	Feng Y, Liu P, Zhang J. A mobile terminal based trajectory preserving strategy for continuous querying LBS users[C]. Hangzhou: Proceedings of IEEE the Eighth International Conference on Distributed Computing in Sensor Systems, 2012.
[8]	Dwork C. Differential privacy: A survey of results[C]. Heidelberg: Proceedings of the International Conference on Theory and Applications of Models of Computation, 2008.
[9]	Hua J, Gao Y, Zhong S. Differentially private publication of general time-serial trajectory data[C]. Hong Kong: Proceedings of the IEEE Conference on Computer Communications, 2015.
[10]	徐启元, 陈珍萍, 付保川, 等. 基于差分隐私的混合位置隐私保护[J]. 计算机应用与软件, 2019, 36(6):296-301.
	Xu Qiyuan, Chen Zhenping, Fu Baochuan, et al. Hybrid location privacy protection based on differential privacy[J]. Computer Applications and Software, 2019, 36(6):296-301.
[11]	Yan Y, Hao X, Zhang L. Hierarchical differential privacy hybrid decomposition algorithm for location big data[J]. Cluster Computing, 2019, 22(4):9269-9280. doi: 10.1007/s10586-018-2125-z
[12]	杨理皓, 谷科, 李威. 基于差分隐私机制的位置数据隐私保护策略[J]. 计算机应用研究, 2018, 35(3):895-900.
	Yang Lihao, Gu Ke, Li Wei. Location data privacy protection strategy based on differential privacy mechanism[J]. Application Research of Computer, 2018, 35(3):895-900.
[13]	霍峥, 孟小峰. 轨迹隐私保护技术研究[J]. 计算机学报, 2011, 34(10):1820-1830. doi: 10.3724/SP.J.1016.2011.01820
	Huo Zheng, Meng Xiaofeng. A survey of trajectory privacy-preserving techniques[J]. Chinese Journal of Computer, 2011, 34(10):1820-1830. doi: 10.3724/SP.J.1016.2011.01820
[14]	王晓滨. 基于差分隐私的位置信息隐私保护研究[D]. 北京: 北京交通大学, 2019.
	Wang Xiaobin. Research on location information privacy protection based on differential privacy[D]. Beijing: Beijing Jiaotong University, 2019.
[15]	胡柳, 王梅, 邓杰, 等. 大数据环境下的隐私保护研究[J]. 电子科技, 2017, 30(7):159-161.
	Hu Liu, Wang Mei, Deng Jie, et al. Research on big data privacy protection[J]. Electronic Science and Technology, 2017, 30(7):159-161.
[16]	杨飞, 朱志祥. 基于特征和空间信息的核模糊C-均值聚类算法[J]. 电子科技, 2016, 29(2):16-19.
	Yang Fei, Zhu Zhixiang. Kernelized fuzzy C-means clustering algorithm based on features and spatial information[J]. Electronic Science and Technology, 2016, 29(2):16-19.
[17]	Khanlari M, Ehsanian M. An improved KFCM clustering method used for multiple fault diagnosis of analog circuits[J]. Circuits, Systems, and Signal Processing, 2017, 36(9):3491-3513. doi: 10.1007/s00034-016-0479-0
[18]	Wang Q, Zhang Y, Xiao Y, et al. Kernel-based fuzzy C-means clustering based on fruit fly optimization algorithm[C]. Stockholm: Proceedings of the International Conference on Grey Systems and Intelligent Services, 2017.

差分隐私模糊聚类位置保护方法

Differential Privacy Fuzzy Clustering Location Protection Method

RichHTML

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

图/表 7

参考文献 18

相关文章 13

Metrics

本文评价

推荐阅读 0

[1]	刘琛,马驷俊,倪雪莉. 基于属性的物联网感知层访问控制方案[J]. 电子科技, 2019, 32(9): 55-59.
[2]	蓝机满. 基于云计算的数据挖掘系统设计[J]. 电子科技, 2019, 32(8): 70-74.
[3]	张长青,杨楠. 基于车联网大数据分析的实时路况检测系统[J]. 电子科技, 2019, 32(8): 66-70.
[4]	张翠翠,阮树骅. 用于短频繁项的隐私保护关联规则挖掘方法[J]. , 2016, 29(5): 88-.
[5]	龚楷椿,邬春学. 基于邻接矩阵的自适应图像分割算法研究[J]. , 2016, 29(2): 66-.
[6]	李健坤 1，宋寿鹏 1，李建平 2，李翔 1，丁楠 1. 基于支持向量机的贯流风叶叶片粘连缺陷诊断[J]. , 2016, 29(11): 154-.
[7]	王虹,孙红. 基于混合聚类算法的客户细分策略研究[J]. , 2016, 29(1): 29-.
[8]	张应辉,郑东,马华. 具有隐私保护的云存储访问控制方案[J]. , 2015, 28(6): 61-.
[9]	穆新亮. 基于混合核函数的快速KPCA人脸识别算法[J]. , 2015, 28(2): 46-.
[10]	江楠,徐秦. 数据流聚类算法在数据处理中的应用[J]. , 2015, 28(1): 155-.
[11]	安健, 张扬. 基于Otsu和模糊核聚类算法的极化SAR图像分类[J]. , 2014, 27(2): 42-.
[12]	郝晓鹏,赵洪业,高强. 基于KPCA的FPSO流程故障检测方法[J]. , 2013, 26(5): 109-.
[13]	井塬塬. 基于C均值聚类算法的雷达信号分选方法[J]. , 2011, 24(4): 4-.