Rattle wrote:

] Before anyone sends me any feedback, please keep in mind,
] whats posted here is my proof of concept.

I think the graphs are very interesting regardless of the polish level.

] There is one minor error in Decius's description. This graph
] does not take topics into account at all. I threw out all
] that information when generating these. The links do convey
] direction, so thats why some nodes have two links between
] them. If I wasn't throwing out topic information, there would
] be a number of nodes with over 6 links connecting them.

So why then, in the power-law graph, are there nodes with over 300 inbound links?

] In particular.. In this format, when private information is
] aggerated, does it become statistical information that can be
] desiminated without infringing on user's privacy? And can
] anyone think of a way that an individual user's reputation
] information could be mined from this graph?

I don't think there is anything you can learn from this that you wouldn't learn from careful analysis of the blogs. I think this only gets sketchy if you start revealing who is who...

Tom

RE: Visualizing MemeStreams