基于改进关联规则的报送信息大数据特征隐匿性加密算法

陈威宇1; 王泷1; 何建锋1; 苏怀振2; 徐延文1

基于改进关联规则的报送信息大数据特征隐匿性加密算法

引用本文：陈威宇1，王泷1，何建锋1，苏怀振2，徐延文1.基于改进关联规则的报送信息大数据特征隐匿性加密算法[J].计算技术与自动化,2024,(4):123-128

摘要点击次数: 236

全文下载次数: 0

作者	单位
陈威宇1，王泷1，何建锋1，苏怀振2，徐延文1	（1.国网甘肃省电力公司，甘肃兰州 730000；2.国网甘肃省电力公司定西供电公司，甘肃定西 743000）

中文摘要:报送信息大数据的安全性直接关系到“数字化纪检云平台”数据挖掘的准确性，一旦泄露或者被篡改，纪检工作可靠性将大大降低，由此，提出一种基于改进关联规则的报送信息大数据特征隐匿性加密算法。手工获取报送信息大数据，通过连接转换器实时连通异构数据库，对不符合规范要求的数据进行自动修正，包括缺失数据填补、离群值处理和数据标准化。利用改进遗传算法改进Apriori算法，挖掘数据项或属性之间隐含的关联规则，通过关联规则识别报送信息大数据特征。以报送信息大数据特征为输入，利用神经网络层级之间的权重矩阵作为密钥，将明文数据转变为密文数据。结果表明：所研究加密算法应用下，“三公经费”开支数据加密后的密文和明文相关性均在0~0.2之间，说明所研究算法的加密效果较好，保证了报送信息大数据安全。

中文关键词:改进关联规则 Apriori算法报送信息大数据特征识别神经网络隐匿性加密算法

Hidden Encryption Algorithm of Big Data Characteristics of Submitted Information Based on Improved Association Rules

Abstract:The security of big data of submitted information is directly related to the accuracy of data mining of "digital discipline inspection cloud platform". Once it is leaked or tampered, the reliability of discipline inspection work will be greatly reduced. Therefore, a feature hiding encryption algorithm of submitted information big data based on improved association rules is proposed. Manually obtain the big data of the submitted information, connect the heterogeneous database in real time through the connection converter, and automatically correct the data that does not meet the specification requirements, including missing data filling, outlier processing and data standardization. The improved genetic algorithm is used to improve the Apriori algorithm to mine the hidden association rules between data items or attributes, and identify the characteristics of big data of submitted information through association rules. Taking the big data characteristics of the submitted information as the input, and using the weight matrix between the neural network levels as the key, the plaintext data is transformed into ciphertext data. The results show that under the application of the encryption algorithm studied, the correlation between ciphertext and plaintext after the encryption of "three public funds" expenditure data is between 0 ~ 0.2, which shows that the encryption effect of the studied algorithm is good and ensures the security of big data of submitted information.

keywords:improved association rules Apriori algorithm submitted information big data feature recognition neural network hidden encryption algorithm

查看全文 查看/发表评论 下载pdf阅读器