IA Summit Folksonomies Panel
I'm back from IA Summit in Montreal and just starting to collect my thoughts about it. For now, here are the presenations from the Folksonomies Panel on Saturday morning:
- Introduction (PDF, 2.2MB) - Gene Smith
- Folksonomies: Better than Nothing? (PDF, 5.8MB) - Peter Morville
- Metadata for the Masses (PDF, 1.2MB) - Peter Merholz
- Folksonomies; A Wrappers' Delight (PDF, 468KB) - Thomas Vander Wal
I thought the panel went well overall. Enough friction to keep the discussion interesting, smart presentations from the panelists, and good questions from the audience helped keep things rolling.
David Weinberger's quote "the old way makes trees... the new way rakes leaves together" was a recurring theme in the discussion. Peter Morville responded to the quote by saying (this is paraphrased):
And we know what happens to leaves when we rake them together. They rot. And become food for new trees.
And at the beginning the his presentation Peter Merholz said (again, paraphrased):
And sometimes the trees get really big and block out the light and kill off everything on the ground. So you have to chop the trees down.
Or, you know, there's always the option of a prescribed burn. The tree metaphor got a bit overworked, but I think it reinforces the idea that there's a metadata ecosystem where tags and folksonomies can influence more formal information architectures.
The other feeling I got from the post-panel Q & A was that people weren't sure where they could use folksonomies. Just after the panel I thought up the folksonomy formula:
Login + Large Content Set = Folksonomy Opportunity
In the case of del.icio.us the large content set is the web, but it could just as easily be a legal or medical database, an intranet or a staff directory. The folksonomy piece complements a larger architecture plan, but it also allows the users to interact with and influence the architecture.
An IA can look at tagging patterns to help determine labeling, structure, search and best bets in the same way as query term logs. But unlike query term logs, tags have persistent value for the user because they are exposed and re-usable. (And, I'd argue, more closely approximate real user knowledge than search terms.) Tags also enable the creation of communities around classification. In fact, the folksonomy is a ridiculously low-cost kind of community that's nothing more than a beneficial side effect of people tagging documents for their own future recall.
I think that sums up my post-panel thoughts. I'd like to say a public "thanks" to the panelists--Peter, Peter and Thomas--who made the whole thing worth doing.
