Over the last few months, I’ve become increasingly interested in learning how to analyze social networks to find new sources and the connections that exist between them. Unfortunately, I didn’t know much about social networking analysis. So I decided to start teaching myself some basics.
Why bother with this? Put simply: you may not realize you have a goldmine of new sources hiding in plain sight among the sources you already know about.
For example, read Kieran Healy’s Using Metadata to Find Paul Revere. By cross-referencing data of colonial Bostonians and their memberships to 18th century social clubs, he showed how the British could have found Paul Revere without knowing he even existed. He wasn’t well known to them, but he seemed to know everyone else involved at the start of the revolution, smack in the middle of the entire network.
I thought I’d start by experimenting with my own Facebook account, which currently has just under 1,800 friends. I was sure a lot of these folks knew each other, but would they fall into distinct groups? Do any of them serve as “bridge builders” between them, in a way that Paul Revere did?
To do this, I used two tools: Netvizz and Gephi. Netvizz is a Facebook app that looks at all of your FB friends and checks to see if any of them are friends with each other. It then saves the results in a file that can be imported into the open-source network analysis tool known as Gephi. You can download it and some basic tutorials at Gephi.org.
After a bit of tinkering, I came up with this map, representing 1,785 FB friends and more than 37,000 connections between each other. It turns out they form their own distinct network clusters.
Here’s a breakdown of the main clusters and the “Paul Reveres” among them.
