Extreme Events in Coupled Social Networks: Media Events as External Shocks

Pascal Jürgens

Short Abstract
This talk presents recent research in the field of social communication networks in order to shed light on the interaction between media and social communication systems. It will attempt to bring theoretical concepts of audience behavior and media choice into a network perspective on communication, offering explanations for the behavior of some observable extreme events in those networks.

Extended Abstract
There is a long-established tradition of theories about the way in which media information, people's minds and behavior interact. Perhaps one of the best known is the theory of agenda setting, which postulates several links through which the media are thought to define what people deem important and vice versa. The common characteristic of such models is that they consider long-term relations and work on an aggregate level.

With the broad-scale adoption of social communication channels on the internet, researchers for the first time gain insight into popular communication - unaggregated and with high temporal fidelity. Whereas we used to quantify beliefs and attitudes in the limited scope of surveys, we can now quantify behavior through non-reactive measurements.

The new opportunities for empirical research carry with them a shift in focus. Previous reliance on directed, deep and meaningful questions gives way to a superficial interpretation of traces of behavior. This shift can be seen as part of a larger trend away from the model of rational choice and deep beliefs towards a picture of individuals easily swayed by mental heuristics and external cues, as nobel prize winners Kahneman and Tversky have prominently painted it.

Armed with such data from social networks and other communication networks, scholars have started to look closer at the empirical reality of networked communication, uncovering new and unexpected aspects of human behavior. Of peculiar interest are extreme events - situations in which the behavior of a network or a significant subset changes drastically and often seemingly arbitrarily.

As far as some extreme events (such as political protests) go, mass communication research appears to offer a unique complement to network theory. Its theoretical models of event dynamics offer a compelling addition to the temporal analysis of one-mode networks. In this respect, media events can be regarded as external shocks which diffuse into an observable network, inducing and interacting with extreme events.

This talk presents recent research in the field of social communication networks in order to shed light on the interaction between media and social communication systems. It will attempt to bring theoretical concepts of audience behavior and media choice into a network perspective on communication, offering explanations for the behavior of some observable extreme events in those networks.