At 22:12 -0500 3/30/06, Kashyap, Vipul wrote:
I also think the use case for bringing the thesaurus to the Web
goes beyond the OWL stuff Vipal describes. Essentially, by
moving to RDFS (SKOS) you get an advantage different than reasoning -
the terms in your thesaurus become URIs that other people can point
to. It means that they can use your terminologies in their
applications, and links back to your terms can be maintained (rather
than "reverse engineered" by a search engine). Tools
people are playing with for SKOS (and OWL) include image annotation,
text/blog indexing, and database indexing/linking - and in those
cases, the ability to link to things outside the ontology space are
crucial (for example, imagine a lot of bloggers in the life science
area using your terms as the things they subscribe to via RSS - or
imagine being able to link your content to, for example, Nature's, by
having mappings between synonyms in each others' thesauri, with live
links to the content).
[VK] I agree with Jim that these would enable reuse of thesauri
concepts in a more significant way than otherwise.
However, as Jim himself
points out, the use case is that of bringing the thesaurus to the
Web
What would be interesting would be
to bring the thesaurus to the Semantic Web, that is make explicit the
semantic
structures in the thesaurus and
exploiting them using SW technologies. The catch of course is that it
requires
significant upfront
investment.
Vipul obviously means something different by "Semantic Web"
than I do -- but what do I know?
I think RDFS and RDFS+a few OWL constructs are as much (if not
more) Semantic Web than a standalone ontology even if it is in
OWL...
THat said, the benefit of having vocabularies in SKOS, and
using the linking of the Semantic Web is that then richer models (in
OWL) can be tied to terms in the ontology, and that will provide a lot
of new functionality... but the key is that ontologies published
on the Semantic Web (i.e. in RDF so that there is a URI for each term)
have a lot of power as yet unexplored, and provide a way to
incrementally add the power of OWL, without having to take the whole
thing into OWL in one fell swoop (if you'll pardon the pun). A
thesaurus of the size of the one that Anita describes would be very
expensive to port to OWL in one shot - but the port to SKOS is easy,
and then in the same, or different, document richer semantics can be
added - and that is new to KR, since incrementality of this kind (via
multiple rich subsets linked to a a single thesaurus) has not been
explored in the traditional KR&R space, certainly not "in the
wild" where search engines and the like can also b involved...
but then, I guess I have a vision for this stuff that isn't quite the
same as many people's these days -- my blog entry at [1] is a short
description of some of this...
-Jim H.
p.s. end of spiel, this isn't the place to argue visions...
this started with some practical advice about thesauri, which is what
I was aiming at...
[1]
https://www.mindswap.org/blog/2006/01/26/thnking-about-the-semantic-web/
--
Professor James Hendler
Director
Joint Institute for Knowledge Discovery
301-405-2696
UMIACS, Univ of Maryland
301-314-9734 (Fax)
College Park, MD 20742
http://www.cs.umd.edu/~hendler
Web Log: http://www.mindswap.org/blog/author/hendler
Joint Institute for Knowledge Discovery
UMIACS, Univ of Maryland
College Park, MD 20742
Web Log: http://www.mindswap.org/blog/author/hendler
