一种基于Elman神经网络的网络舆情威胁估计方法  

Network Public Opinion Threat Assessment Based on Elman Neural Network

作　　者：王铁套[1] 王国营[1] 陈越[1] 

机构地区：[1]信息工程大学电子技术学院,河南郑州450004

出　　处：《信息工程大学学报》2012年第4期482-488,共7页Journal of Information Engineering University

基　　金：国家社科基金资助项目(09&ZD014)

摘　　要：在详细分析网络舆情影响因素和自身属性的基础上,建立了网络舆情指标体系。依据模糊综合评价法和神经网络构建了网络舆情威胁估计模型。采用模糊综合评价法确定指标体系中各舆情因素的数值化评价值,并结合改进的Elman神经网络模型对网络舆情进行综合评判,最终得到网络舆情的威胁等级。仿真实验表明了该模型的有效性和准确性。This paper constructs the index system of the network public opinion based on a detailed analysis of the factors affecting the network public opinion and its attributes. According to the fuzzy comprehensive evaluation and the neural network, the threat assessment model of the network public opinion is constructed. The model affirms the value of the public opinion factor of the index system by the fuzzy comprehensive evaluation, then evaluates synthetically the network public opinion com-bining with the improved Elman neural network, and last gets the threat levels of the network public opinion. The simulation experiment demonstrates the validity and accuracy of the model.

关 键 词：网络舆情 指标体系 威胁估计 ELMAN神经网络 层次分析法 模糊综合评价法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]