In other words, we need better defined metadata for describing how
people project different persona into virtual social environments.
Consider one use case: people maintain multiple persona in virtual
worlds... in fact, when you eject a "griefer" from your event in
Second Life, they can show up a few minutes later as a completely new
and different avatar. It'd be quite useful to describe attributes
about the person behind that avatar: how long have they used this
avatar, what reputation do they have online in general, etc.
Consider another use case: in virtual worlds different players tend to
have specialties and expertise, particularly in the case of role
playing - or even in the case of in-world commerce. It would be useful
to allow for machine inference and recommendations, especially for
newcomers looking to find a tutor or guide.
It may turn out that FOAF can accommodate most of our needs - with a
few extensions, such as "Relationships". SIOC has some similar ideas,
but they focus on "User" and "Account" models, not so much about
persona, reputation, etc. In any case, there will need to be
extensions made and tested and agreed by a community.
We've floated this proposal with the name "RDF Avatarics Workgroup"...
RDF - as the framework for representing metadata and resources, though
in practice this will likely require more parts of the "Semantic Web
Stack", such as RDF Schema, OWL, etc.
Avatarics - as a term to describe technology related to avatars -
similar to the relation between "robotics" and "robots". See an
earlier proposal at: http://ceteri.blogspot.com/2006/12/three-laws-of-avatarics.html
Workgroup - because by nature this must be an interdisciplinary
effort, spanning across standards for online games, virtual words,
social networks, web-based communities, IM, reputation API, etc.
And so we are seeking your help, ideas, and participation.