图书情报工作 ›› 2013, Vol. 57 ›› Issue (15): 27-33.DOI: 10.7536/j.issn.0252-3116.2013.15.004

• 专题一:网络舆情的形成、发展与预测研究 • 上一篇    下一篇

利用灰色预测与模式识别方法构建网络舆情预测与预警模型

杜智涛1, 谢新洲2   

  1. 1. 中国青年政治学院新闻与传播系;
    2. 北京大学新闻与传播学院
  • 收稿日期:2013-04-24 修回日期:2013-07-09 出版日期:2013-08-05 发布日期:2013-08-05
  • 作者简介:杜智涛,中国青年政治学院新闻与传播系讲师,E-mail:ddzztt@sina.com;谢新洲,北京大学新闻与传播学院教授,博士,博士生导师。
  • 基金资助:

    本文系国家自然科学基金资助项目"基于互联网网民言论信息的口碑监测、分析与管理研究" (项目编号:71073006)研究成果之一。

The Establishment of Public Opinion Forecasting and Early-warning Model with the Methods of Grey Forecasting and Pattern Recognition

Du Zhitao1, Xie Xinzhou2   

  1. 1. Journalism & Communication Department, China Youth University for Political Sciences, Beijing 100089;
    2. School of Journalism & Communication, Peking University, Beijing 100871
  • Received:2013-04-24 Revised:2013-07-09 Online:2013-08-05 Published:2013-08-05

摘要:

针对网络舆情主题多样、内容复杂、数据海量等特点,构建一个由三级指标构成的多层次网络舆情研判指标体系;在此基础上,运用灰色预测方法建立网络舆情预测模型,运用模式识别方法建立网络舆情预警模型,并选取60个网络舆情事件为训练样本,确立舆情警级分类规则,对该模型的可用性进行验证;鉴于网络舆情预测与预警所需要的数据量大、计算复杂,提出网络舆情预测与预警系统建设的整体架构,以进行数据处理。

关键词: 灰色预测, 模式识别, 网络舆情, 预测, 预警

Abstract:

As the network public opinion possesses the features of various topic, complex content and large amount data etc., this paper constructs a multi-level index system of network public opinion analysis which is consisted of three level indicators. And then, network public opinion forecasting model is established with the grey forecasting method, and the early-warning model is established through the pattern recognition method. This paper selects 60 network public opinion events as training samples, establishes the classifying discipline of network public opinion level, and tests the usability of the forecasting and early-warning model. The forecasting and early-warning need large amount of data and complicated calculation, then it puts forward the whole frame of constructing the system to process data.

Key words: grey forecasting, pattern recognition, network public opinion, forecasting, early-warning

中图分类号: