摘要
当今社会进入大数据时代,其挖掘工作主要依托云计算平台。然而,云计算环境复杂多变,数据安全和隐私泄露问题日益显著。针对云计算环境下的数据安全问题,提出一种混合加密方案,根据数据敏感等级采用不同加密方法。同时提出了一种隐私保护的可识别性k-prototypes聚类算法,并利用信息熵对各数值属性进行权重分配进行改进,以解决大数据挖掘过程中的隐私泄露问题。结果显示,改进k-prototypes聚类算法的NMI值为0.284,准确度达到了94.95%,RI值为0.935,运行时间为924 ms。综合来看,该方案在提高数据加密效率的同时,确保了云环境下数据的安全性。
In today's society,which has entered the era of big data,its mining work mainly relies on cloud computing platforms.However,the cloud computing environment is complex and ever-changing,and data security and privacy leakage issues are becoming increasingly prominent.A hybrid encryption scheme is proposed to address data security issues in cloud computing environments,using different encryption methods based on data sensitivity levels.At the same time,a privacy preserving identifiable k-prototypes clustering algorithm was proposed,and the weight allocation of various numerical attributes was improved using information entropy to solve the privacy leakage problem in big data mining.The results show that the improved k-prototypes clustering algorithm has an NMI value of 0.284,an accuracy of 94.95%,an RI value of 0.935,and a running time of 924ms.Overall,this scheme ensures the security of data in the cloud environment while improving data encryption efficiency.
作者
尹飞
葛崇慧
YIN Fei;GE Chonghui(Jiangsu Frontier Electric Technology Co.,Ltd.,Nanjing 211134,China)
出处
《自动化与仪器仪表》
2024年第11期61-64,69,共5页
Automation & Instrumentation
基金
江苏省电力有限公司信息化项目《营销2.0客户数据安全防护能力提升服务》(1XXKS-231091-WF)。
