IEEE AOC'12The 6th IEEE WoWMoM Workshop on
Autonomic   and   Opportunistic  Communications

Keynote Speech

WHAT CAN SOCIAL NETWORKS TELL US ABOUT THE EXTERNAL WORLD?

Seth Myers

Abstract
Individuals interact with members of their social circle on a regular basis, whether it be meeting up with a friend for coffee or posting an amusing link on their Facebook profile. While these interactions often reveal much about the individual, they can also tell us plenty about the rest of the world. I will provide two specific examples. First, we will examine location-based social networks, which are social networks via mobile phone apps in which users can announce their current location to all of their friends in the network. I will present a probabilistic model that identifies which of an individual's movements are periodic (such as commuting to and from work) and which are social (such as going to a party at a friends house). Not only can this model predict an individual's future movements with alarming accuracy, but it can also reveal local geography, population densities, and major commute patterns, as well as detect large social gatherings such as concerts. Second, we examine the spread of information between users on the micro-blogging site Twitter. As users share news articles and funny videos with other users, single pieces of content can cascade across large portions of the Twitter social network. This process, however, is heavily influenced by sources outside the Twitter network, such as television and other forms of mass media. By modeling the decision process each user goes through as they are exposed to new content, we can quantify the otherwise unobservable influx of information from outside the network. We can then use this model to detect external events such as civil unrest in Egypt, as they happen, only by observing the patterns of information spread on Twitter.

Bio
Seth Myers works at the Institute for Computational and Mathematical Engineering of the Stanford University. His research interests are primarily in information diffusion in social networks and location-based social networks. He works under the supervision of Prof. Jure Leskovec. Before Stanford, he did his undergraduate work at Northwestern University, and his primary research was on complex systems, specifically the synchronization of networks of oscillators.