You’ve probably seen some samples of what I’d call “social data-mining” on tech sites and blogs. Usually, you feel like you’re in front of a spaghetti bowl that would have met Captain Fluo in the kitchen. Social network analysis is indeed a relatively new field where one can explore vast or little amounts of data to help show connections, influence, links, and give a meaning to what appears at first sight as a heap of dots and lines.
I’ve shown you here a previous sample of what could be done by mapping the Leisure conversation within the Singapore blogosphere, a work that proved interesting in many ways : the “virtual” links were really showing an existing interaction in real life, with food bloggers for instance being tightly connected.
Here’s two more samples on other networks.
The first one is a trial to show the main actors of a conversation on Twitter that happened during the IdeasInc conference in Singapore, in September (see wrap-up of this event hereÂ and there). A one-day talk and start-up pitches contest with, say, 300 to 400 hundreds visitors and a few tweeps gathering around the #ideasinc hashtag. I’ve been using NodeXL to import the data of this conversation and play a bit with it, but I’m not quite satisfied, as it shows only the main interactions within the talking tweeps that day, I’m not yet pro enough to show other facets of this conversation. But it will come soon I hope
The second one is an analysis of my Facebook friends’ network, including their male/female/other status. I did it as an assignment in a new course I took on Coursera, called “Social Network Analysis“, led by Lada Adamic of the University of Michigan. I exported my data on Facebook through GetNet, an app first designed by Bernhard Rieder, and tweaked by Lada to help us make this assignment an easy one. Then, I imported the data in Gephi to make it more readable.
This in only the first week of this course, so I hope to be able to show really more detailed stuff the forthcoming weeks
Category: Playing with data