密度峰值聚类k匿名分布式网络数据隐私保护方法研究  

Research on Privacy Protection Method for Distributed Network Data with Density Peak Clustering K-anonymity

作　　者：郭艳红[1] GUO Yanhong(Communication University of Shanxi,Taiyuan 030013,China)

机构地区：[1]山西传媒学院,山西太原030013

出　　处：《数字通信世界》2025年第3期41-42,120,共3页Digital Communication World

摘　　要：由于分布式网络数据分散在多个节点上,导致数据隐私泄露的概率较大,为此,本文进行了密度峰值聚类k匿名的分布式网络数据隐私保护方法研究。其充分考虑了分布式网络环境自身的特点,引入了分布式k-NN查询算法,以找到其k个最近邻点,同时保证查询过程以不泄露数据隐私为目标,构建了针对分布式网络数据的k近邻匿名模型;利用密度峰值聚类算法识别具有高局部密度并且与更高密度点的距离较大的数据点作为聚类中心,对k近邻匿名模型中的节点进行聚类,实现数据保护。在测试结果中,设计方法在不同场景中的保护效果最好,对应的数据泄露概率始终稳定在0.2以下。Due to the dispersion of distributed network data across multiple nodes,there is a high probability of data privacy leakage.Therefore,this article conducts research on density peak clustering k anonymous distributed network data privacy protection methods.It fully considers the characteristics of distributed network environments and introduces the distributed k-NN query algorithm to find its k nearest neighbors while ensuring that the query process does not leak data privacy.It constructs a k-nearest neighbor anonymous model for distributed network data;Using density peak clustering algorithm to identify data points with high local density and larger distance from higher density points as clustering centers,clustering the nodes in k-nearest neighbor anonymous model to achieve data protection.In the test results,the design method showed the best protection effect in different scenarios,and the corresponding data leakage probability remained stable below 0.2.

关 键 词：密度峰值聚类 k匿名 分布式网络 数据隐私保护 分布式k-NN查询算法 k近邻匿名模型 局部密度 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]