Decoding Your Social Network

Friday, October 12, 2012: 8:55 AM
602 (WSCC)
Tyler McCormick, PhD , Sociology, Statistics, University of Washington, Seattle, WA
From Facebook to Google Plus, social networks are everywhere. Understanding social networking data is a challenging task. In this talk, I will discuss two interesting statistical problems that arise from social networking data.

You might have 1000 Facebook friends, but how many of them would you really recognize if you saw them on the street, much less call on the phone or invite to dinner? On the basis of your Facebook friends and your Facebook friends’ Facebook friends, is there a way that we can determine who your really close friends are? In this presentation, I will talk about statistical tools that can be used to better understand social network structure – for instance, to identify friendship groups. 

Social networking websites collect a lot of information about individuals beyond just their list of friends. For instance, on Facebook, a user has the option to enter information about their political views, religious beliefs, and favorite books, movies, and more. I’ll talk about how we can use information about a given individual’s Facebook friends in order to learn a lot about that individual – even things that he or she may have wanted to keep private!