Online public opinion dissemination model and simulation under media intervention from different perspectives

This study investigates the influence of different media interventions on the dissemination of online public opinion, analyzes its laws, and provides a theoretical basis for the control of relevant departments. This study adds the influence of network media and government media on susceptible-exposed-infectious-removed (SEIR) model and considers the interaction between different emotions to construct the susceptible-exposed-positive-emotional-infectious-and-negative-emotional-infectious-immune-removed (SEI2R1R2) online public opinion dissemination model under such dual intervention of network media and government media. The opinion propagation model is simulated in Python and the model is validated by mining text sentiment using text convolutional neural networks (CNN) to fit relevant parameters. Results show that the SEI2R1R2 online public opinion dissemination model is more consistent with reality. With the role of media from different perspectives, identifying and controlling emotional communicators and increasing the probability of converting positive and negative emotional communicators into quitters can effectively reduce the number of communicators, and thus control the spread of public opinion.