基于RF算法的突发事件网络舆情演化预测分析

Prediction for the Evolution of Emergency Network Public Opinion Based on RF Algorithm

【目的/意义】突发事件类网络舆情演化情况的掌握对舆情监管部门而言至关重要。鉴于此,本文致力于构建能够准确预测舆情演化的模型,此预测模型的建立可拓宽舆情预测的渠道,为舆情参与主体和监管部门及时掌握舆情演化态势提供方法依据。【方法/过程】基于随机森林(RF)算法建立突发事件网络舆情演化预测模型,以微博和第三方舆情监测平台作为变量数据来源,以R语言为操作环境,然后进行模型的训练与预测。【结果/结论】实验表明,较之其它方法,本文构建的模型有更高的拟合度和更低的误差值。从结果来看,本模型的预测输出值与真实值最为接近,较好地实现了对舆情演化的预测,将RF算法应用在舆情预测的研究中具有一定的先进性。

【Purpose/significance】It is very important for public opinion supervision department to master the evolution of public opinion in emergency network.In view of this,this paper is committed to building a model that can accurately predict the evolution of public opinion.The establishment of this prediction model broadens the channels of public opinion prediction,and provides a method basis for the participants and regulatory departments to grasp the evolution situation of public opinion in time.【Method/process】Based on stochastic forest (RF) algorithm,a prediction model of public opinion evolution in emergency network is established.Weibo and third-party public opinion monitoring platform are used as variable data sources,R language is used as operating environment,and then the model is trained and predicted.【Result/conclusion】 Experiments show that the proposed model has higher fitting degree and lower error value than other methods.From the results,the predicted output value of this model is the closest to the real value,and the prediction of public opinion evolution has been well completed.The application of RF algorithm in the study of public opinion prediction has certain advancement.