摘要
隐私数据可能以多种形式存在,且可能经过加密或混淆处理。识别这些隐私数据并提取其特征是一个复杂的过程。此外,大数据中包含的噪声数据和无效数据也会影响隐私泄露检测结果的准确性。为此,本研究提出了一种基于改进粒子群优化算法-可能性模糊C均值聚类算法(PSO-PFCM)聚类算法的隐私泄露检测方法。首先,设计爬虫技术功能模块,自动化获取数据流量,再设计数据解密流程,为后续特征提取奠定基础;然后,通过提取大数据平台数据的静态特征,更有效地识别和分类隐私数据;最后,将改进后的粒子群优化算法应用到聚类中心的选择中,去除无效数据和噪声数据,减少噪声数据对聚类结果的影响。通过聚类结果与隐私泄露检测结果对比来判断是否存在隐私泄露问题。实验结果表明:针对不同的数据类型,该方法始终能够有效降低隐私泄露检测的错误率,提高检测效率和准确性,能够为大数据平台的安全运营提供保障。
Privacy data may exist in various forms and may be encrypted or obfuscated.Identifying these private data and extrac-ting their features is a complex process.In addition,the noise and invalid data contained in big data can also affect the accuracy of privacy leak detection results.Therefore,this study proposes a privacy leak detection method based on an improved particle swarm optimization-possibility fuzzy C-Means(PSO-PFCM)clustering algorithm.Firstly,design the crawler technology functional module to automatically obtain data traffic,and then design the data decryption process to lay the foundation for subsequent feature extraction;Then,by extracting the static features of big data platform data,privacy data can be more effectively identified and classified;Final-ly,the improved particle swarm optimization algorithm will be applied to the selection of clustering centers,removing invalid and nois-y data,and reducing the impact of noisy data on the clustering results.Compare the clustering results with the privacy leak detection results to determine whether there is a privacy leak issue.The experimental results show that this method can effectively reduce the er-ror rate of privacy leakage detection for different data types,improve detection efficiency and accuracy,and provide guarantees for the secure operation of big data platforms.
作者
郭舒扬
朱大智
陈诗
钟晓文
GUO Shuyang;ZHU Dazhi;CHEN Shi;ZHONG Xiaowen(Information and Communication Branch,Hainan Power Grid Co.,Ltd.,Haikou 570203;Southern Grid Digital Platform Technology(Guangdong)Co.,Ltd.,Guangzhou 510000)
出处
《自动化与仪器仪表》
2025年第5期269-272,共4页
Automation & Instrumentation
基金
个性化资产管理数字化应用建设(投资计划及项目管理域数据迁移)项目(072900HQ42210004)。
