基于决策树的网络舆情类型识别模型研究

Research on network public opinion type recognition model based on decision tree classification method

网络舆情的合理归类是相关部门进行精确调控的重要依据。首先利用层次分析法构建网络舆情评估指标体系,然后通过K-均值聚类将网络舆情分为5种类型,最后使用决策树分类方法构建网络舆情类型识别模型。结果表明,本文构建的评估指标体系能够作为划分网络舆情类别的重要依据,网络舆情类型识别模型能够较为精确地识别处于潜伏期的网络舆情的类别,为相关部门提供具有针对性的调控策略。

The rational classification of network public opinion is an important basis for the relevant departments to make preciseadjustment and control. First, the evaluation index system of network public opinion is constructed by AHP, and then the networkpublic opinion is divided into five types by means of K-means clustering. Finally, the network public opinion type identificationmodel is constructed by using the decision tree classification method. The results show that the evaluation index system constructed inthis paper can be used as an important basis for classification of network public opinion. The network public opinion typeidentification model can accurately identify the categories of network public opinion in the latent period, and provide pertinentregulation strategies for the related departments.