Characterizing super-spreading in microblog: An epidemic-based information propagation model

Area of Science:

• Social Network Analysis
• Information Diffusion Modeling
• Computational Social Science

Background:

• Microblogging services drive rapid information spread, creating "super-spreading events".
• Super-spreaders significantly influence information propagation dynamics on Online Social Networks (OSNs).
• Understanding these phenomena aids tasks like hot topic detection and harmful content monitoring.

Purpose of the Study:

• To develop a parameterized model, the SAIR model, for characterizing super-spreading events in tweet propagation.
• To analyze the analogy between information diffusion and contagious disease spread.
• To identify and model the role of super-spreaders in information dissemination.

Main Methods:

• Adaptation of well-known epidemic models (SIR) to create the SAIR model.
• Statistical empirical observations on a real-world Weibo dataset.
• Steady-state analysis for equilibrium and validation against real-world data.

Main Results:

• The SAIR model demonstrates superior performance in characterizing super-spreading events compared to the conventional SIR model.
• Empirical validation on Weibo data confirms the SAIR model's effectiveness.
• Numerical simulations reveal parameter sensitivity in information propagation.

Conclusions:

• The SAIR model offers a more promising approach to understanding super-spreading phenomena on social media.
• The study highlights the importance of super-spreaders in information diffusion patterns.
• Findings contribute to better social media data mining and intervention strategies.