Indiana University’s Truthy Project attempts to visualize and analyze the diffusion of information on Twitter.
In particular, researchers are investigating how memes spread, and how valid the information in those memes actually is such as in the above graphic of @barackobama mentions (the orange cluster) and posts (the rest).
As New Scientist wrote last fall of the group’s Twitter analysis of the 2010 election:
The Indiana team is using these and other examples to train software to recognise unusual patterns of tweets. The system is still being tested, but [principal investigator Filippo] Menczer says it can already detect around 85 per cent of the questionable tweets he has collected.
The investigation at that time surrounded automated retweet bots and astroturfing through multiple accounts operated by the same users.
Conservative groups, Truthy discovered, had the most dubious activity.
This isn’t because the left doesn’t employ similar tactics, Menczer notes, but because ”there is so much more traffic from the right. The left is just not generating enough traffic for us to notice.”