Prediksi Penyebaran Hoaks di Media Sosial X Menggunakan Model SEIR dan Metode Runge-Kutta Orde 4

The rapid development of information technology has facilitated easier access to and dissemination of information through social media. However, this convenience also increases the spread of false information (hoaxes), which can cause misunderstandings, damage individual reputations, and influence public opinion. This study aims to analyze and predict the dynamics of hoax dissemination on social media X using the SEIR (Susceptible, Exposed, Infected, Recovered) model adapted from epidemiology and numerically solved using the fourth-order Runge–Kutta (RK4) method. The SEIR model represents the stages of hoax spread, including users who are vulnerable to false information (Susceptible), users who are exposed (Exposed), users who actively spread hoaxes (Infected), and users who stop spreading the information (Recovered). Simulations were conducted using initial values obtained from aggregated data and monthly data with a 12-week prediction period. The simulation results show that hoax dissemination follows a single-wave pattern, where the number of users spreading hoaxes increases in the early phase and reaches its peak in the second week with a maximum Infected compartment value of 0.1487, then gradually declines. At the end of the prediction period (week 12), the compartment values are S = 0.07118, E = 0.00002, I = 0.00008, and R = 0.92872, indicating that most users are in the recovered state and no longer spread hoaxes. A comparison of predictions based on aggregated and monthly data shows that although the initial values differ, the resulting dissemination dynamics exhibit consistent patterns. These results indicate that the SEIR model solved using the RK4 method is able to systematically describe the dynamics of hoax spread on social media X and can be used as an approach to understand and predict the behavior of false information dissemination.