On 12/05/2013 09:49 AM, Damian Steer wrote:
On 5 Dec 2013, at 13:52, Thomas Steiner <to...@google.com> wrote:
Dear Public-LOD,
Thank you all for your very helpful replies. Following your joint
arguments, owl:sameAs is _not_ an option then.
You could use dc:hasFormat to link them:
"A related resource that is substantially the same as the
pre-existing described resource, but in another format." [1]
<http://ex.org/video.mp4> dc:hasFormat <http://ex.org/video.ogv> .
This is an excellent example in which perfection may be the enemy of the
good. But it depends on the application.
For applications that do not semantically distinguish between mp4 and
ogv formats of a work, owl:sameAs would be fine, because within the
perspective of those applications, there would be no discernible
difference between them. But for other applications that *do* care
about different formats, there would indeed be a difference and
owl:sameAs should *not* be used. So one might conclude that, to be
safe, dc:hasFormat should always be used, or a proxy object should be
used, because that would preserve the most information. But notice
that there is a cost in using dc:hasFormat or a proxy object instead of
owl:sameAs : it requires a more sophisticated RDF application to
understand the data. owl:sameAs is completely generic, and doing
owl:sameAs inference is probably one of the most common kinds of
inference after subsumption (subclass) inference. In contrast,
dc:hasFormat is far more special-purpose, and the proxy object approach
would require much more specific knowledge of the data to enable an
application that treats mp4 and ogv formats as equivalent to recognize
that -- for its purposes -- they are indeed equivalent.
Thus, for *some* purposes dc:hasFormat or a proxy object would be an
overall better choice, and for *other* purposes owl:sameAs would be an
overall better choice. If you are publishing this data for others to
use, you could offer both choices, putting the owl:sameAs statements in
a separate document for those who want to use them.
David
