A Method for Simulating the Propagation of Network Hot Topics Based on Combined SEIRS-ARIMA Model

Abstract

: The spread of hot topics on the internet can easily attract public attention and discussion. However, the spread of rumours can cause social instability, while positive media guidance can generate economic value. Therefore, developing a model that can understand the propagation of popular topics on the internet is a worthwhile pursuit. In this study, a hybrid model is developed by combining the SEIRS (Susceptible-Exposed-Infected-Recovered-Susceptible) infectious disease model to simulate the overall trend and the ARIMA (Autoregressive Integrated Moving Average) model to simulate noise, to explain internet hot topic propagation. Using the wordle game as a case study, the model exhibits a notable capability in conforming to the general pattern and the majority of the minor variations, while demonstrating a high degree of interpretability.