Steve, the Artful Tagger

Folksonomies - the ad hoc tagging by anyone of anything - sound terribly democratic compared to your top-down authoritarian imposition of taxonomies, but it's easy to see why people are sceptical about them: how can anything useful arise out of something so chaotic?

Del.icio.us is one example of how such folksonomies can be really useful, and here's another (and note the groovy .museum domain - the first time I've seen this):

"Steve” is a collaborative research project exploring the potential for user-generated descriptions of the subjects of works of art to improve access to museum collections and encourage engagement with cultural content. We are a group of volunteers, primarily from art museums, who share a common interest in improving access to our collections. We are concerned about barriers to public access to online museum information. Participation in steve is open to anyone with a contribution to make to developing our collective knowledge, whether they formally represent a museum or not.

Very cool - both in terms of adding metadata to objects, and as far as getting the public involved with art. Indeed, this idea should really be extended to everything - imagine a database of public places that people could tag.

Great idea, then, but why "Steve"?

A Moot of Folksonomies

Being a rigorous sort of chap, I was sceptical about folksonomies - ad-hoc tags. But over time I've come to appreciate their power - and the fact that once people start using them routinely, the combined body of folksonomic knowledge becomes quite impressive.

But the obvious question is: what lies beyond the simple tag? Myabe this kind of thing:

GroupMe! extends the idea of social tagging systems like del.icio.us, Flickr or BibSonomy by introducing the group dimension. The foundation of social tagging systems are so-called folksonomies, which describe how users (folks) tag resources (e.g. photos, videos, publications, etc.). In technical terms a folksonomy is just a collection of tag assignments:

(User, Tag, Resource) = User has tagged Resource with Tag at a particular time.

Over time it is likely that semantics emerge, e.g. tags that are often assigned to same resources may be synonyms. Hence, folksonomies are promising to improve (web) search, etc. With GroupMe!'s approach of taggable groups we extend tag assignments with a group dimension:

(User, Tag, Group, Resource) = User has tagged Resource with Tag in a certain Group at a particular time.

This prompts the next question: what do we call these groups? I vote a "moot".