After some time trying to get a whole day (nearly 300,000
edges nodes and 41,000 edges) of the weblog dataset data to visualize, I finally produced the first ‘accurate’ picture. This is day 1, English blogs, run in GUESS with multiple edges and at 30% of the original size and before changing layouts (still does not want to do that). My next steps for the whole dataset are to remove the disconnected nodes and change the layout. I did not think it would be so difficult, but the dataset is being stubborn. Still, the process is exciting! Today is the first day that I have had much success with the entire layout and I wonder if it is due to using a larger screen. I am in the lab using the new 30″ Mac. SO NICE! Much faster and easier to see details. (Photo above: my new MacBook Pro 15.4 beside lab’s new 30″ Mac – heaven or what!) I have looked briefly at some of the larger networks and one interesting pattern I have seen is most of the larger clusters are based around livejournal users. I really don’t know a lot about this platform, but will look at it more carefully. This may not be so surprising as it can not be said that I am visualizing a network, rather what I think I am seeing are artifacts (traces) of conversation. More about what is actually visualized soon!
[tags] GUESS, HUMlab, weblogs, Mac [/tags]