动态网络舆情异常数据准确监测算法仿真  

The Simulation of Abnormal Data of Accurate Monitoring Algorithm of Dynamic Network Public Opinion

作　　者：罗玉婷[1] 熊秋娥[1] LUO Yu-ting;XIONG Qiu-e(Informatization Center,Nantong University,Nantong Jiangsu 226019,China)

机构地区：[1]南通大学信息化中心,江苏南通226019

出　　处：《计算机仿真》2024年第9期470-474,共5页Computer Simulation

摘　　要：网络舆情信息的形式多样,既有文本也有图片、视频等,复杂度较高,且包含很多谣言、恶意攻击、虚假信息等。如何从众多信息中排除无价值和虚假信息是动态网络舆情监测的难点之一。为此,改进贝叶斯算法,提出一种新的动态网络舆情异常数据监测方法。利用网络爬虫技术采集动态网络舆情数据,通过改进多小波变换系数相关去噪算法,滤除数据中的噪声。利用TF-IDF算法提取动态网络舆情异常数据关键特征,引入LSI算法获取高频数据。基于此,应用改进的贝叶斯算法构建贝叶斯模型,将全部特征输入模型中,实现动态网络舆情异常数据监测。实验结果表明,所提方法可以精准获取动态网络舆情异常数据,且监测时间更短,监测相对误差低于0.04%。At present,the forms of network public opinion information are diverse,including texts,images and videos.The complexity is high.Network public opinion information contains a lot of rumors,malicious attacks,and false information.How to remove valueless and false information from numerous information is one of the difficulties in dynamic network public opinion monitoring.In order to address this issue,based on an improved Bayesian algorithm,a method of monitoring abnormal data in dynamic network public opinion was presented.Firstly,dynamic network public opinion data were collected through web crawlers,and then the noise was filtered out through an improved multi-wavelet transform coefficient algorithm.Secondly,the key features of abnormal data in dynamic network public opinion were extracted by the TF-IDF algorithm.Meanwhile,high-frequency data was obtained by the LSI algorithm.On this basis,an improved Bayesian algorithm was applied to construct a Bayesian model.Moreover,all features were input to the model.Finally,abnormal data monitoring of dynamic network public opinion was achieved.Experimental results show that the proposed method can accurately obtain THE abnormal data IN dynamic network public opinion,with shorter time.The relative error is less than 0.04%.

关 键 词：改进贝叶斯算法 动态网络 数据去噪 舆情异常数据 监测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]