[Data-modeling] Which universe is that human from?
Jeff Fry
jfry at metaweb.com
Fri Mar 28 00:07:29 UTC 2008
There's been conversation before about how confusing it can be that
there are many
<http://qa.sfo1.metaweb.com/view/guid/9202a8c04000641f8000000000b003f6>
human
<http://qa.sfo1.metaweb.com/view/guid/9202a8c04000641f800000000025e3ad>
topics
<http://qa.sfo1.metaweb.com/view/guid/9202a8c04000641f80000000002c0fb0>,
most tied to different universes, and one
<http://qa.sfo1.metaweb.com/view/en/human> tragically tethered in our
non-fictional universe. Some likely merit it - Humans in the Star Wars
and the Dungeons & Dragons universes might well have different
properties, and certainly link to different Wikipedia articles.
The topic for our own non-fictional species is the first one in
relevance searches, and so many fictional characters (Marcie
<http://www.freebase.com/view/en/marcie>, Spider-Man
<http://qa.sfo1.metaweb.com/view/en/spider-man>, Nancy Drew
<http://qa.sfo1.metaweb.com/view/en/nancy_drew>, etc.) are all linked
back to this topic. To make things a bit sillier, someone noted that the
Human Character Species is found in the Firefly fictional universe...and
so now Ron Weasley, Captain Nemo, and Indiana Jones may well be caught
in a crossfire between the Anglo-Sino Alliance and the Reavers.
Cleanup is certainly needed, but I'm not really clear what the model is
at this point? Does every universe need it's own Human topic? Probably
not. If they don't, should those that don't have their own topic cluster
around the topic for the non-fictional species? Or around a generic
fictional human topic? And either way, should that topic list all or
none of the universes it belongs to? If all, then we might need to stop
having the Species property on Fictional Character list Universes that
contain it as a disambiguator.
What do you think?
Jeff
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.freebase.com/pipermail/data-modeling/attachments/20080327/e4365b2f/attachment.htm
More information about the Data-modeling
mailing list