基于弱无回溯模式匹配算法的敏感大数据安全分级检测方法  

Sensitive Big Data Security Hierarchical Detection Method Based on Weak Backtracking Pattern Matching Algorithm

作　　者：张海明 王艺霏 那琼澜 王森 温馨 姚艳丽 ZHANG Haiming;WANG Yifei;NA Qionglan;WANG Sen;WEN Xin;YAO Yanli(State Grid Jibei Information and Telecommunication Company,Beijing 100053)

机构地区：[1]国网冀北电力有限公司信息通信分公司,北京100053

出　　处：《计算技术与自动化》2025年第1期171-176,共6页Computing Technology and Automation

摘　　要：随着信息技术的广泛应用,敏感大数据在传输和存储过程中存在隐私泄露、数据滥用、安全风险等问题。为了提高敏感大数据的安全性,提出了一种基于弱无回溯模式匹配算法的敏感大数据安全分级检测方法。建立协方差矩阵构建敏感大数据样本库,设定约束条件同构嵌入敏感大数据子图,采用弱无回溯模式匹配算法回溯匹配敏感大数据查询图索引,降低计算复杂度。利用核函数映射查询图特征向量,提高敏感大数据识别准确度,以剪枝的方式完成敏感大数据特征相似性匹配。通过相似度指数函数计算敏感因子,划分敏感大数据安全等级,完成自动化分级检测。实验结果表明:所提方法的敏感大数据安全分级准确度为100%、特征匹配耗时为2.7 s、收敛速度为0.44。由此证明,所提方法能够有效保护敏感大数据安全,在数据传输和存储中具有较高的实际应用价值。With the wide application of information technology,there are some problems in the transmission and storage of sensitive big data,such as privacy leakage,data abuse and security risks.In order to improve the security of sensitive big data,this paper proposes a hierarchical detection method of sensitive big data security based on weak backtracking pattern matching algorithm.Establishing covariance matrix and constructing sensitive big data sample database.Set constraints to isomorphically embed sensitive big data subgraphs,and use weak backtracking pattern matching algorithm to backtrack the index of sensitive big data query graph,which reduces the computational complexity.Kernel function is used to map the feature vector of query graph to improve the recognition accuracy of sensitive big data,and the similarity matching of sensitive big data features is completed by pruning.The sensitivity factor is calculated by similarity index function,and the security level of sensitive big data is divided,thus completing automatic hierarchical detection.The experimental results show that the accuracy of sensitive big data security classification is 100%,the time for feature matching is 2.7 s,and the convergence speed is 0.44.This proves that the proposed method can effectively protect the security of sensitive big data and has high practical application value in data transmission and storage.

关 键 词：弱无回溯模式匹配算法 协方差矩阵 相似度指数函数 敏感大数据 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]