[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: XML/RDF

• To: "Paul Prescod" <paul@p...>
• Subject: Re: XML/RDF
• From: "Jonathan Borden" <jonathan@o...>
• Date: Mon, 11 Nov 2002 00:46:31 -0500
• Cc: <xml-dev@l...>, "Mike Champion" <mc@x...>

[[
If the ontology problem is solved and standardized, then why are people
still working on it? Also, according to the SNOMED website, SNOMED is
changing (mostly growing) quickly:

http://www.snomed.org/milestones_txt.html
]]

Several points, first, the _part_ of the ontology problem termed
"description logic" (DL) is well characterized, coded, and somewhat
operational. SNOMED is developed using DL tools. Inferences are made using
SNOMED classified data using DL reasoners etc. What OWL gives us is:

1) a web based _interchange language_ for exchanging ontologies for example
between different DL reasoners, between ontology authoring tools and
inferencing/reasoning tools etc.

2) OWL is not just DL, rather the OWL DL subset corresponds to DL. OWL Lite
is intended to be both simpler to implement and less "powerful" than DL. OWL
Full is RDF(S) whose reasoners are still in development. see
http://www.w3.org/TR/owl-guide/ for a better description.

3) SNOMED is a proprietary ontology classification. SNOMED intends anyone
and everyone who *uses* SNOMED terms to pay SNOMED for this -- I cannot
support this in the least, regardless of the amount of payment. I am totally
opposed to this type of language. I have publically suggested to SNOMED that
they ought to open the language and make their $$$ by selling guides, texts,
tools, etc. I don't consider the SNOMED _classification_ proprietary to the
extent that this mirrors _common medical knowledge_. However, since SNOMED
considers this proprietary, I've not actually looked at SNOMED itself though
since it claims to be a DL vocabulary I've mentioned it in talks etc.

As an alternative to SNOMED consider OpenGALEN http://www.opengalen.org and
note that this is also DL based.

So these are some of the problems that we face prior to our ability to use
ontologies -- we need ontologies that we can freely use... sigh...

Jonathan

http://www.jonathanborden-md.com
http://www.erieneurosurgery.com
http://www.openhealth.org

• Follow-Ups:
  • Re: XML/RDF
    • From: Mike Champion <mc@x...>

• Prev by Date: Re: XML/RDF
• Next by Date: Re: XML/RDF
• Previous by thread: Re: XML/RDF
• Next by thread: Re: XML/RDF
• Index(es):
  • Date
  • Thread