| Abstract: |
| We propose a susceptible-infected-recovered-susceptible (SIRS) model to study the user adoption and abandonment dynamics of online social networks (OSNs). Two types of abandonment dynamics are considered in our model: (1) infectious abandonment due to internal influences by individuals who already quit the network and (2) noninfectious abandonment due to mass media, commercials, or emergence of new products. The well-poseness of the model is first verified and the existence and stability of its equilibria are then discussed. Extensive numerical simulations are performed to validate the theoretic results. The effectiveness of the model is demonstrated by fitting it to the historical Facebook daily active user data, and finally, the fitted model parameter values are used to predict the numbers of future Facebook active users. |
|