基于大数据聚类的工程项目审计信息化管理体系建设  

Construction of Information Management System for Engineering Project Audit Based on Big Data Clustering

作　　者：王伟 燕超源 聂顺林 王国娟 王利军 文洪昌 WANG Wei;YAN Chao-yuan;NIE Shun-lin;WANG Guo-juan;WANG Li-jun;WEN Hong-chang(State Grid Information&Telecommunication Group Co.,Ltd.,Beijing 100010 China;Beijing Guodiantong Network Technology Co.,Ltd.,Beijing 100070 China;State Grid Qinghai Electric Power Company,Xining 810000 China)

机构地区：[1]国网信通产业集团,北京100010 [2]北京国电通网络技术有限公司,北京100070 [3]国网青海省电力公司,青海西宁810000

出　　处：《自动化技术与应用》2024年第8期113-116,135,共5页Techniques of Automation and Applications

基　　金：国网信息通信产业集团有限公司科技项目(5400-202219120A-0-0-00)。

摘　　要：为提高审计效率,基于大数据聚类算法设计工程项目审计信息化管理体系。通过清洗与选择预处理审计数据,提取局部异常因子,分析数据的可聚类属性。对审计大数据聚类分析,计算数据相似度,进一步匹配待筛选的审计指标,并对其进行降序的排列,得到不同财务指标的算术平均值以及离差系数,建立输入节点与输出节点之间的权重连接机制,建立基于大数据聚类的工程项目审计模型,通过计算任意样本与节点之间的距离,得到工程项目审计结果。实验结果可知:设计算法能够准确地得到财务审计结果,具备较好的应用性。In order to improve the audit efficiency,the project audit information management system is designed based on big data clustering algorithm.By cleaning and selecting the pre-processing audit data,the local abnormal factors are extracted and the clustering attributes of the data are analyzed.It clusters and analysises of audit big data,calculates the data similarity,further matches the audit indicators to be screened,and arranges them in descending order to obtain the arithmetic mean value and dispersion coefficient of different financial indicators,establishes the weight connection mechanism between input nodes and output nodes,establishes the project audit model based on big data clustering,and obtains the project audit results by calculating the distance between any sample and nodes.The experimental results show that the designed algorithm can get the financial audit results accurately and has good applicability.

关 键 词：大数据 工程项目审计 信息化管理体系 聚类 相似度匹配 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]