http://www.bloggersblog.com/cgi-bin/blo ... og=1024072
and if you want to be difficult ...
http://www.cs.cmu.edu/~jure/blogs/blogs-uc-pa.html
A naive, intuitive solution would be to select the big, well-known blogs. However, these usually have a large number of posts, and are time-consuming to read. We show, that, perhaps counterintuitively, a more cost-effective solution can be obtained, by reading smaller, but higher quality, blogs, which our algorithm can find.