APP下载

基于GA-BP算法的网络舆情预警机制研究

2021-04-03陈蒙李学志

现代信息科技 2021年19期

陈蒙 李学志

摘  要:近年来,互联网的快速发展带来了社交媒体网络的激增,广大民众可以在网上分享信息、知识和观点。然而,一旦出现突发事件,蜂拥而来的信息会对公众造成冲击,需要对信息的发展变化做出正确的预测并及时发现潜在的危机。有鉴于此,首先,构建一个基于突发事件的网络舆情预警指标体系,通过评论家算法计算各指标的权重,求得综合评价值,进而推断网络舆情预警水平;其次,利用基于遗传算法优化的BP神经网络构建网络舆情预警模型;最后,以突发事件河南水灾为例进行实证分析。

关键词:BP 神经网络;遗传算法;网络舆情;舆情预警;预警指标;批评家

中图分类号:TP391.3                  文献标识码:A文章编号:2096-4706(2021)19-0018-06

Research on Network Public Opinion Early Warning Mechanism Based

on GA-BP Algorithm

CHEN Meng, LI Xuezhi

(School of Information Engineering, Xinjiang Institute of Technology, Aksu  843100, China)

Abstract: In recent years, the rapid development of the Internet has led to the proliferation of social media networks, where vast numbers of people can share information, knowledge and opinions online. However, once an emergency occurs, the information flooding in will impact the public, so it is necessary to make a correct prediction of the development and change of information and timely discover the potential crisis. In view of this, firstly, build a network public opinion early warning index system based on emergencies, calculate the weight of each index through the critic algorithm, get the comprehensive evaluation value, and then infer the network public opinion early warning level; secondly, BP neural network based on genetic algorithm optimization is used to build network public opinion early warning model. Finally, take Henan flood as an example for empirical analysis.

Keywords: BP neural network; genetic algorithm; network public opinion; public opinion early warning; early warning index; critic

0  引  言

在互联网时代,互联网络对个人行为模式、观点、政治态度和情感倾向的影响越来越大。网络舆论在其整个传播过程中都会对社会产生巨大的负面影响。对任一实体任何意见/情绪平均倾向的计算和评估,有助于组织和个人了解当前形势或获得一个对不熟悉事物的正确认识。网络舆情预警研究已引起了学术界的广泛关注,实现网络舆情预警的关键在于建立评价指标体系和预测模型。现有的研究大部分重复性较强,过分强调指标体系的可解释性,往往导致预测精度不高。构建一个具有网络舆情预警作用的指标体系结构[1],需要综合考虑其预警水平和预测方法。为此,本文采用一种客观、严谨、可复制的方法——“批评家”方法,该方法不仅注重不同指标之间权重的影响,还兼顾预警指标冲突对预测效果的影响,结合遗传算法和BP神经网络构建网络模型,用以预测网络舆情水平[2]。……

登录APP查看全文