Newer posts are loading.
You are at the newest post.
Click here to check if anything new just came in.
Click here to check if anything new just came in.
September 08 2010
interesting stuff. though i'd be even more curious about analyses that go in the opposite direction: instead of taking the categories for granted and then clustering the content based on these categories, start out with some variable, cluster its manifestations, and then look at the categories that emerge in the form of clusters, as well as at how sharply defined the cluster-categories are.
