I am preparing the analysis of rss4you.com (a RSS aggregator made by weblogger, only available in french). The idea of this piece of software, apart from the aggregation of posts, is to allow users to rate their blogroll. By doing so, people can know if somebody has a blogroll similar to another and hence discover new sites that could be of interest.

I would like to analyze the social network in rss4you. We could have a matrix like the following. Each grade correspond to the rating of a blog. For instance Nico gave 6 to Fab's blog.

Fab     Roby     Nico
Fab               5          5
Roby   x                    5
Nico    6         5

With this matrix, we can compute similarity measures (with multiple data scale/analysis) in order to visualize who is close to whom so as to get something like: