Simulation of Information Spreading Following a Crisis

In this chapter we consider how information about a crisis spreads. We consider scenarios, and models thereof, which are variants of the susceptible/infected model from epidemiology. The populace is initially unaware that a crisis has occurred. When the crisis begins, awareness that a crisis has occurred spreads throughout the populace via a combination of broadcast media and social feedback; eventually the entire populace becomes aware of the crisis. We investigate transitions in our models from a completely unaware populace to a completely aware populace, focusing particularly on the speed of the process and the relative impact of different media types. Our models’ behaviour depends heavily on the input parameters which dictate the strengths of different spreading mechanisms. As much as possible we draw values for these parameters from real data. These parameters vary significantly depending on the time of day. For example, the number of people who become aware almost immediately because they are tuned in to broadcast media when the crisis occurs ranges from about 2 % to about 47 %. In addition, the timescale on which an alert unfolds means that our models should incorporate dynamic parameters, i.e., parameters that change as the alert unfolds. With regard to the relative impact of different media types in our models, we note that, within our model, social media such as Facebook and Twitter are much less important than traditional media, primarily by virtue of their smaller audience and less frequent use. We also identify a critical timescale: the length of time it takes someone with the TV/Radio on to realize there is a crisis and then to relate it to someone else. This realize-and-relate timescale is likely to have an important role in shaping the early course of events in daytime crisis spreading.