I believe knowledge is sexy.
Facebook is part of our lives, for better or worse. But how much do we really understand about ourselves and those with whom we connect? The practice of social network visualization is hardly new, but it is becoming easier and easier to perform such visualization on an individual basis. Wolfram Alpha provides an excellent start, for instance. I’m on a complex network visualization kick personally, so this weekend I sat down to learn Gephi (a phenomenal open-source graph visualizing tool). Thanks to netvizz I was able to download a complete snapshot of my Facebook social network data, including information such as post count, like count, comment count, and such for my entire friend list. After massaging this data with Gephi, I got some rather beautiful and surprisingly informative graphs.
My Facebook social graph, clustered by Modularity using a modified Fruchterman-Reingold layout. Node size represents number of connections (known as Degree). Node color denotes communities as defined by a Modularity calculation.
In these graphs Gephi’s Modularity calculation does a hauntingly accurate job of partitioning my different communities (what Google might call Circles). Purple here represents roughly everyone I know from Apple Retail. Notice how highly interconnected this community is, with many nodes highly connected to others within the same community. This beautifully represents how “tight-knit” the Apple Retail community is socially; everyone is likely a friend of everyone else. Contrast this with my Texas Tech community (graphed in orange), which is much more loosely distributed. While I have almost as many Texas Tech friends as Apple friends, fewer of them know everyone else in my Texas Tech community. Off to the left a yellow group represents my friends from CCPA during my “high school” years. This is a similarly close-knit group, but much smaller than the other two.
Detail showing the region between my Texas Tech (orange) and Apple Store (purple) communities. The teal dot near the center is my good friend Kevin Saunders, who bridges these communities.
Between these communities, it is often easy to spot those who bridge different groups of people. For instance, I’ve known my friend Kevin Saunders throughout my Texas Tech years, and he later came to work with me at the same Apple Store. Therefore he bridges these two communities, plus a smaller sub-community of Texas Tech friends (represented in teal). Between the yellow and orange groups in the first image, you can spot my friend Shelby, who attended CCPA and later Texas Tech. I had always vaguely conceptualized these people as having a foot in both worlds, but seeing them represented here really made that pop.
Alternative circular layout, with nodes corrected so they don’t overlap. In this view it’s slightly easier to see the interconnectedness of the various communities. The difference in relative node size has also been increased.
The complex world of you
Seeing the world around us in the visual language of complex networks is an exciting and evolving practice, with applications far beyond social networks. I highly recommend starting with this phenomenal animation of Manual Lima’s talk on the Power of Networks. If you like it, buy his book. Then check out Visual Complexity and Visualizing.org, two wonderful sites curating complex network visualization projects. If you’d like more details on how I composed the visualizations here, feel free to comment, shoot me an email or check out this guide.