------------------------------------------------------------------------ Date: Fri, 21 May 2004 10:56:55 +0100 From: Andy Powell Subject: Re: Fwd: Approved qualifiers in Dublin Core To: DC-USAGE@JISCMAIL.AC.UK ------------------------------------------------------------------------ On Thu, 20 May 2004, Diane Hillmann wrote: > Folks: > > Any suggestions for an answer to this? My answer would be the same as the one I gave to the NetInsert people, i.e. something along these lines... --- cut --- One approach is that you 'name' your encoding scheme using a URI within your own namespace, e.g. http://wwwcf.nlm.nih.gov/class/Class If you do this, then your XHTML encoding would be along the lines of This is in line with the XHTML encoding guidelines at http://dublincore.org/documents/dcq-html/ and doesn't require any registration of 'NLMClass' with DCMI. Note that there are other possibilities for choosing the URI that you use to name your taxonomy - for example, you could use a PURL http://purl.org/NLM/Class in which case the 'rel="schema' line above will need to be changed. Note also that there are psuedo-religious issues around whether you should use a URIref rather than a URI to name your scheme (i.e. whether it should have a '#' in it or not). The above approach can also be used if you choose a future encoding based on XML, where the instance metadata will look something like WA 288 given an appropriate namnespace declaration for the 'nlm' prefix. It will also work in RDF/XML (though the syntax is not as shown above). You might also like to consider assigning a URI (or URIref) to each of the terms in your taxonomy? If so, then a possible way of doing this is to use the 'info' URI scheme (though note that this isn't a registered URI scheme yet). For details see http://www2.elsevier.co.uk/~tony/info/info.html Some examples are at http://www2.elsevier.co.uk/~tony/info/info.html#ex_ddc --- cut --- Andy > >From: "Boehr, Diane (NIH/NLM)" > >To: "'dih1@cornell.edu'" > >Subject: Approved qualifiers in Dublin Core > >Date: Wed, 19 May 2004 16:18:41 -0400 > > > >Diane, > > > >How would NLM go about getting the NLM classification to be one of the > >approved qualifiers for DC.Subject? MeSH has already been approved and we > >would like the value NLMClass to also be added, with the following URL > >link http://wwwcf.nlm.nih.gov/class/ > > > >I couldn't find an official contact for a request like this on the DCMI > >site, so I hope you can help. > > > >Diane Boehr > > > >Diane Boehr > >Cataloging Unit Head > >National Library of Medicine > >8600 Rockville Pike > >Bethesda, MD 20894 > >301-435-7059 > >boehrd@mail.nlm.nih.gov > ------------------------------------------------------------------------ Date: Fri, 21 May 2004 13:14:26 +0200 From: Thomas Baker Subject: Re: Fwd: Approved qualifiers in Dublin Core To: DC-USAGE@JISCMAIL.AC.UK ------------------------------------------------------------------------ Diane Boehr of NLM wrote: > > >How would NLM go about getting the NLM classification to be one of the > > >approved qualifiers for DC.Subject? MeSH has already been approved and we > > >would like the value NLMClass to also be added, with the following URL > > >link http://wwwcf.nlm.nih.gov/class/ Diane Hillmann asked: > > Any suggestions for an answer to this? Andy responded: > This is in line with the XHTML encoding guidelines at > > http://dublincore.org/documents/dcq-html/ > > and doesn't require any registration of 'NLMClass' with DCMI. Andy's response provides guidance on using the URI (which they already have) in encodings, but Diane Boehr seems to be requesting that DCMI provide approval for NLM classification. Two responses seem possible: 1) Traugott, who may be out of email contact just now, is going to put up a simple Web page listing Subject vocabularies. 2) We could consider whether the DCMI Usage Board can and should assign status to this externally maintained term, perhaps in the form of an endorsement of an assertion by NLM that http://wwwcf.nlm.nih.gov/class/ qualifies dc:subject. ------------------------------------------------------------------------ Date: Wed, 23 Jun 2004 16:23:04 -0400 From: "Rebecca S. Guenther" Subject: Re: "Approval" for NLM encoding scheme? To: DC-USAGE@JISCMAIL.AC.UK ------------------------------------------------------------------------ Okay, I talked to Diane yesterday. She didn't quite understand everything I was telling her I don't think. It seems that they are using Dublin Core elements, but as HTML, not as XML. You can see what they're using at: http://www.nlm.nih.gov/tsd/cataloging/metafilenew.html. You'll see they're using that dot syntax. They've used "NLMDC.Subject.NLMClass" as their element for this encoding scheme. I explained to her the various options, particularly about declaring a persistent URI for the NLM classification and/or registering an info: scheme. (She did say that they are using that NLM classification URI as a persistent one.) This would work in the context of XML schema or RDF/XML but didn't have too much meaning for her since they seem to be using the dot syntax. So she still seemed to be more attracted to having some sort of registration of NLM classification comparable to what already exists in dcmi:terms for MESH and LCC. I suppose how I interpret it is that she would like the option that Tom mentions below: "record that assertion in a machine-processable form so that the endorsement of the NLM term shows up in our Web documents and RDF schemas;" I guess that means a fast-track sort of registration/approval by the UB. Alternatively, I could talk to someone else there who knows more about any plans they have for setting up namespaces, schemas and the like. Rebecca > On Sat, May 22, 2004 at 08:12:22AM +0200, Thomas Baker wrote: > > On Fri, May 21, 2004 at 04:07:05PM -0400, Rebecca Guenther wrote: > > > I'm not sure I understand how this relates to NLM. The assertions have to > > > do with asserting whether each term in the relators list is a subproperty > > > of Contributor. NLM is asking about how to use their encoding scheme with > > > the DC element Subject. Or maybe I'm missing something on this late > > > Fri. afternoon. > > > > NLM has a term which -- they would like to assert -- qualifies > > a DCMI term. But unless I have misunderstood, Diane's point is > > that NLM is _also_ looking for some sort of acknowledgement, > > approval, recommendation, or dissemination by DCMI of that > > assertion. > > > > The MARC Relators, to my way of thinking, are a test case for > > working out the mechanics and the etiquette of making and > > recognizing such assertions, so we should consider whether > > it provides a model for handling the NLM qualifier. > > > > If NLM already has a term -- and it already has a persistent, > > institutionally supported URI (still an "if" I believe) -- > > and they assert that it qualifies dc:subject, we could do > > one or more of the following: > > > > -- simply take mental note; > > -- alert DC-General (once); > > -- record the fact on Traugott's planned Web page; > > -- review the assertion in UB and endorse or recommend it in some form; > > -- record that assertion in a machine-processable form so that the > > endorsement of the NLM term shows up in our Web documents and RDF > > schemas; > > -- ensure that the term appears in the DCMI Registry.