Delicious Clusters

Clustering Interface for delicious data. The method is simple but it provides a much nicer interface for certain data than any other tool. Currently I am exploring the use for del.icio.us, flickr, and mail, and found a special subset of hierarchical clustering to yield better results than k-means (of course you pay for it in time). Anyway, hopefully helping to solve the problem of tag clouds being inefficient interfaces for semi-large sets.