[rdfweb-dev] Revisiting Groups, a plea for data

Jim Ley jim at jibbering.com
Fri Aug 22 01:26:33 UTC 2003


Describing Groups in the FOAF world:

Defining a group is no problem I think, the current spec version does this
well, and weblog, mbox (for a mailing list for the group, although a
subProperty for actual mailing lists might be good so we can distinguish a
contact email address from the group, and a list they all read.), homepage
etc. can all be used for Group smushing as much as Person smushing.

There's a lot of inference group creating we can do (works at the same
place, collaborates on etc.) but I think it would be a good idea to try and
get some instance data out there to start looking at tools, generating a
member of FoafDevelopers, ecademy, and other places should be pretty cheap
for us to set up, aswell as Libby's existing swad folk will then give us a
data set to start looking at.  If we can get that sample set we can start
creating tools, I'm certainly quite motivated to add some layering into
foafnaut showing groups, their member and their relationships.

On that we also need a "knows" equivalent to groups, the foafnaut developers
group and the FOAF developers group are related, we need to be able to say
that, as far as I can see there are sub-groups, and related groups, I'm not
sure of the best way for this, so I'll leave that to the rest of you.

sbp came up with the important point that it would be good to be able to
model membership of the group just for a certain time-period (so that you
can know that someone was a member in the past etc.) so that complicates
things more than the current design, but I'm happy to think about that
sometime soon, but now I'd just like some data, can we start creating some
on the current modelling?


More information about the foaf-dev mailing list