Everyone who deals with communication these days experience the challenges of information organization. It doesn't really matter whether it's email, memos, invoices, blogs or all of human knowledge, the challenge of categorizing information remains a stout one. Jon Udell has been writing lately about Dynamic Categories. Using the corpora of blogs he subscribes to, he is providing non-linear but structured search to the content. See Analyzing Blog Content for more.
Interestingly, Haystacks - The Universal Information Client is attempting to do the same across information types. Read User Interfaces for Supporting Multiple Categorization for some interesting insights into the challenges of categorizing information. I especially like this:
[we] found that that people who indiscriminately pile paper documents do so in order to skirt the problem of having to choose between several potentially overlapping categories
Ever look at my desktop.
The issue of information categorization was exactly what I was trying to get at with NewsHeap and virtual folders. It seems reasonable that a periodic reclassification of blog folders is not too much work, taking this down to content requires automation. It is interesting then that Udell's approach to content classification is pragmatic and idiomatic - what defines a book reference? An href that includes Amazon. Ubiquitous? Maybe not. Universal? Probably not. Practical? Eminently so.
See Content-aware Search for the continuing battle between the a priori vs explicit camps on information organization.