Is the stackoverflow community a scale-free or a small world network?

Question

I am a graph/network enthusiast and this just for my curiosity :)

I am trying to model the StackOverflow community as a graph/network. Assume that the people in the SO community are nodes and that the answers given to any of the question establishes a relationship between these nodes. The relationship can be assumed to be directed(link from answer -> question) or undirected. The graph could be weighted and that the weights of the nodes could represented number of vote-ups/downs (normalized on the scale of 0 to 1).

What kind of graph/network does one end up with at any given snapshot of time? Is it scale-free? Is it a small-world? The graph is continuously evolving over a period of time and i would like to understand its structure and dynamics.

Is there a way where can i retrieve this relationship data from - may be SO APIs or some one from SO can help me out with (sample) data?

Clarification edit: Scale-free network: A network whose degree distribution asymptotically follows a power law Small-world: A network that has sub-networks characterized by presence of connections between almost any two nodes within them and most pairs of nodes are connected by at least one short path.

Answer 1

To the second part of your question:

Is there a way where can i retrieve this relationship data from - may be SO APIs or some one from SO can help me out with (sample) data?

Try these questions instead. There are a lot of plans to implement an API to access SO data. Some things are in change, but there are possibilities to screen-scrape the data or access them via JSON (afaik).

• Is there a guide to accessing StackOverflow data programmatically?
• What would you want to see in a StackOverflow API?
• Are there plans for a StackOverflow API?

Try it out. Good luck!

Answer 2

What kind of graph/network does one end up with at any given snapshot of time? Is it scale-free? Is it a small-world? The graph is continuously evolving over a period of time and i would like to understand its structure and dynamics.

It takes only a few links between remote clusters to turn a random network into a small world one, so it's quite likely to be small world.

As to whether it's scale free, that would require there to be a few posters with lots of answers and many with only one or two. I seem to recall Jeff saying that there were lots with only one question in one of the pod-casts; you might be better off asking the question there rather than here, as he will have the data.