An Agent-Based Model of Urgent Diffusion in Social Media.

During a crisis, understanding the diffusion of information throughout a population will provide insights into how quickly the population will react to the information, which can help those who need to respond to the event. The advent of social media has resulted in this information spreading quicker then ever before, and in qualitatively different ways, since people no longer need to be in face-to-face contact or even know each other to pass on information in an crisis situation. Social media also provides a wealth of data about this information diffusion since much of the communication happening within this platform is publicly viewable. This data trove provides researchers with unique information that can be examined and modeled in order to understand urgent diffusion. A robust model of urgent diffusion on social media would be useful to any stakeholders who are interested in responding to a crisis situation. In this paper, we present two models, grounded in social theory, that provide insight into urgent diffusion dynamics on social networks using agent-based modeling. We then explore data collected from Twitter during four major urgent diffusion events including: (1) the capture of Osama Bin Laden, (2) Hurricane Irene, (3) Hurricane Sandy, and (4) Election Night 2012. We illustrate the diffusion of information during these events using network visualization techniques, showing that there appear to be differences. After that, we fit the agent-based models to the observed empirical data. The results show that the models fit qualitatively similarly, but the diffusion patterns of these events are indeed quite different from each other.