基于随机游走模型的系统运行日志分布式查询方法  

Distributed query method of system operation log based on random walk model

作　　者：张晓慧 张千福 张才俊 林鸿 余锦河 ZHANG Xiaohui;ZHANG Qianfu;ZHANG Caijun;LIN Hong;YU Jinhe(State Grid Customer Service Centre,Tianjin 300300,China;Beijing China-Power Information Technology Co.,Ltd.,Beijing 100085,China)

机构地区：[1]国家电网有限公司客户服务中心,天津300300 [2]北京中电普华信息技术有限公司,北京100085

出　　处：《电子设计工程》2022年第21期100-103,108,共5页Electronic Design Engineering

摘　　要：目前研究的系统运行日志分布式查询方法查询数据匹配度较低、查询精度较差。为了解决上述问题,基于随机游走模型研究了一种新的系统运行日志分布式查询方法。建立随机游走模型对需要查询的数据进行特征相似性游走路线加权处理。对每个数据节点进行分析,查询到关键的节点,确定关键数据日志跳转矩阵。根据跳转矩阵,计算一段时间内此处节点的网络流量数据,进而根据公式推测出整体系统运行日志的网络流量数据。实验结果表明,基于随机游走模型的系统运行日志分布式查询方法能够有效提高数据匹配度,增强查询精度。At present,the distributed query method of system running log has low matching degree and poor query accuracy.In order to solve the above problems,a new distributed query method of system operation log based on random walk model is studied.A random walk model is established to weight the feature similarity of the data to be queried.Each data node is analyzed,the key nodes are found,and the key data log jump matrix is determined.According to the jump matrix,calculate the network traffic data of the node in a period of time,and then infer the network traffic data of the whole system operation log according to the formula.The experimental results show that the distributed query method based on the random walk model can effectively improve the data matching degree and enhance the query accuracy.

关 键 词：随机游走模型 系统运行日志 分布式查询 数据节点流量 

分 类 号：TN391[电子电信—物理电子学]