Revealing the sentiment propagation under the conscious emotional contagion mechanism in the social media ecosystem: For public opinion management

In public opinion events, the breeding of negative sentiment has a serious negative impact on the network environment and even offline lives. Hence, establishing an emotion-based propagation dynamic model to catch the sentiment development patterns is essential for helping public opinion management. We propose an E-SLFI (Emotion-driven Susceptible-Latent-Forwarding-Immune) model, which describes the dynamics of the sentiment propagation of ternary polarities under the promoting effect of the conscious emotional contagion mechanism. An empirical case composed of 16,354 pieces of forwarding information and two phases verifies the effectiveness of the proposed sentiment propagation dynamic model, due to the fitting optimization indicator MAPE equals 0.0942 % and 0.0066 % respectively. Further, we simulate the model and implement sensitivity analysis of the important parameters of the model. Combining the results of experiments, we find that enhancing the emotional consensus of recipients and inducers can decide the main sentiment in the system, and anti-emotional consensus can improve the existence of weak sentiments. Our work here is conducive to designing online public sentiment guidance strategies to manage public opinion and calming the network atmosphere to a certain extent.