Open GoogleCodeExporter opened 9 years ago
not ISO 19135, but OASIS EBRIM
Original comment by caru...@gmail.com
on 19 Nov 2008 at 4:59
Original comment by caru...@gmail.com
on 19 Nov 2008 at 6:03
(I'm not sure what the original intend was with this entry. But here is one
aspect
related with searches.)
Currently, (MMI Portal 1.5.0.alpha14 (20090831104811)), the search in the Web
VINE
interface *only* considers individuals (for example, those created as vocabulary
terms by voc2rdf). Enabling the search for classes and properties is rather
straightforward. However, it is not very clear how the corresponding relations
are
going to be handled/verified/exploited. For example, given two device
ontologies,
each having a class for Device, what would be the meaning of the following
mapping?
ont1:Device skos:<some>Match ont2:Device
As a real example, http://mmisw.org/portal/#http://mmisw.org/ont/ecs/device_map
contains various skos:broadMatch mappings intended to capture the fact that
certain
specific types of sensors have Sensor (in the second ontology) as a broader
concept.
Perhaps a better modeling approach would be to make the former subclasses of
the latter.
General discussion for a consensus is required. Feedback welcome!
Original comment by caru...@gmail.com
on 31 Aug 2009 at 10:30
I think the new SKOS concept matches are good enough for most basic uses; and
any specific relationships needed for particular kinds of searches can and
should be created (for now) by the particular user community who plan to build
tools around those searches. (In other words, let the community use the
semantic relations to do this job, don't build it into our code.)
This will be especially effective when issue 100/issue 285 get implemented.
A general ability to adapt a mapping interface to any kind of properties (not
just SKOS properties) would be pretty slick. But is a different topic then
we've initially addressed. (Actually, I think this may have been driven
originally by the desire to easily add mapping relations to the VINE interface.
But this request deserves a separate issue.)
So I recommend closing this as WontFix.
Original comment by jbgrayb...@mindspring.com
on 22 Aug 2014 at 5:26
Original issue reported on code.google.com by
caru...@gmail.com
on 19 Nov 2008 at 4:57