Evolution Prediction of Natural Disasters Network Public Opinion based on Random Forest Algorithm

The Internet has become the most important place for people to obtain information and express emotions.After the outbreak of natural disasters, the relevant news was discussed on the Internet, triggering online public opinion and even public opinion crisis. The authorities often try to improve the situation according to the development stage of public opinion, so it is very necessary to have a set of evolutionary prediction models with a high degree of fitting. This paper studies the development and evolution of natural disasters in networks from the perspective of information quantity in the social network. Based on the principle of random forest algorithm and the programming environment of the R language, the RF-based network public opinion evolution prediction model for natural disasters is built, and a new prediction method and model are introduced to analyze the propagation of natural disasters in the network. In order to measure the prediction efficiency of the model, the multiple linear regression model and BP neural network model were compared. The results show that, compared with other methods, the resolution of the proposed random forest algorithm model is as high as 98.85%, the fitting degree is high, and the error is small. This study provides technical support for government risk prevention and public perception of network natural disaster public opinion.