Comments on: Enterprise Taxonomy – An XML schema for Publishing a Taxonomy http://blog.leeromero.org/2009/01/14/enterprise-taxonomy-an-xml-schema-for-publishing-a-taxonomy/ On Content, Collaboration and Findability Tue, 23 Feb 2010 18:35:33 -0600 http://wordpress.org/?v=2.8.6 hourly 1 By: Lee Romero http://blog.leeromero.org/2009/01/14/enterprise-taxonomy-an-xml-schema-for-publishing-a-taxonomy/comment-page-1/#comment-2345 Lee Romero Thu, 15 Jan 2009 15:16:59 +0000 http://blog.leeromero.org/?p=246#comment-2345 Thanks again, Leonard - I looked at the model you'd linked to in the mailing list post but hadn't looked closely enough to realize what the differences were. And, yes, that looks much closer to the structure we use (or, at least I can more easily see how it would translate to it). Thanks again, Leonard – I looked at the model you’d linked to in the mailing list post but hadn’t looked closely enough to realize what the differences were. And, yes, that looks much closer to the structure we use (or, at least I can more easily see how it would translate to it).

]]> By: Leonard Will http://blog.leeromero.org/2009/01/14/enterprise-taxonomy-an-xml-schema-for-publishing-a-taxonomy/comment-page-1/#comment-2344 Leonard Will Thu, 15 Jan 2009 15:02:39 +0000 http://blog.leeromero.org/?p=246#comment-2344 Lee - I'd encourage you to look at the draft ISO model that I referred to rather than the one on the BSI site, as the later model has the important addition of a "conceptGroup" structure. This is more appropriate for development into a classification scheme than the "thesaurusArray" structure. The idea of an array is that it is a group of sibling concepts, sharing a common parent to which they have a BT/NT relationship. These concepts must therefore necessarily all be of the same fundamental type, i.e. must all be in the same facet. The conceptGroup was introduced to allow concepts to be grouped without this restriction - all concepts relating to a subdivision of the subject area, for example - these groups are sometimes called "microthesauri", "themes" or the like. You will find more information in DD8723, available from the British Standards Institution, but unfortunately they charge for it. The ISO draft is more up-to-date, but will not be available for public comment until later this year, we hope. Again we are frustrated by the policy of these standardising bodies in not allowing us to distribute drafts for comment until they are issued officially to those who choose to pay for them. :-( If you don't mind sitting through audio recordings with accompanying PowerPoint slides you might find it useful to look at the record of a meeting of ISKO UK held last year at: http://www.iskouk.org/SKOS_July2008.htm Leonard Lee -

I’d encourage you to look at the draft ISO model that I referred to rather than the one on the BSI site, as the later model has the important addition of a “conceptGroup” structure. This is more appropriate for development into a classification scheme than the “thesaurusArray” structure. The idea of an array is that it is a group of sibling concepts, sharing a common parent to which they have a BT/NT relationship. These concepts must therefore necessarily all be of the same fundamental type, i.e. must all be in the same facet. The conceptGroup was introduced to allow concepts to be grouped without this restriction – all concepts relating to a subdivision of the subject area, for example – these groups are sometimes called “microthesauri”, “themes” or the like.

You will find more information in DD8723, available from the British Standards Institution, but unfortunately they charge for it. The ISO draft is more up-to-date, but will not be available for public comment until later this year, we hope. Again we are frustrated by the policy of these standardising bodies in not allowing us to distribute drafts for comment until they are issued officially to those who choose to pay for them. :-(

If you don’t mind sitting through audio recordings with accompanying PowerPoint slides you might find it useful to look at the record of a meeting of ISKO UK held last year at:
http://www.iskouk.org/SKOS_July2008.htm

Leonard

]]> By: Lee Romero http://blog.leeromero.org/2009/01/14/enterprise-taxonomy-an-xml-schema-for-publishing-a-taxonomy/comment-page-1/#comment-2343 Lee Romero Thu, 15 Jan 2009 14:32:21 +0000 http://blog.leeromero.org/?p=246#comment-2343 Hi Leonard - Thank you very much for providing insight on another alternative (one I missed :-/). I just took a look (admittedly not as deep as I could) at the Draft on http://schemas.bs8723.org/. It does seem to present a good foundation. Applying it to the needs I've had to address, it does appear that the availability of the ThesaurusConcept structure (I'd relate that to the "value" concept I use) and the ThesaurusArray structure (which I'd roughly equate to the "Classification" and probably the "Classification Level" concept I use) cover all of the needs we would have. The hierarchy we need to express between values is inherent in the ThesaurusConcept (which, in fact, is *far* more general than the hierarchy we have in use). Not being deeply familiar with this or SKOS - can you share some insight (a pointer to a document, perhaps?) that provides details on how this differs? Thanks again for the pointer! Hi Leonard – Thank you very much for providing insight on another alternative (one I missed :-/).

I just took a look (admittedly not as deep as I could) at the Draft on http://schemas.bs8723.org/. It does seem to present a good foundation.

Applying it to the needs I’ve had to address, it does appear that the availability of the ThesaurusConcept structure (I’d relate that to the “value” concept I use) and the ThesaurusArray structure (which I’d roughly equate to the “Classification” and probably the “Classification Level” concept I use) cover all of the needs we would have. The hierarchy we need to express between values is inherent in the ThesaurusConcept (which, in fact, is *far* more general than the hierarchy we have in use).

Not being deeply familiar with this or SKOS – can you share some insight (a pointer to a document, perhaps?) that provides details on how this differs?

Thanks again for the pointer!

]]> By: Leonard Will http://blog.leeromero.org/2009/01/14/enterprise-taxonomy-an-xml-schema-for-publishing-a-taxonomy/comment-page-1/#comment-2341 Leonard Will Thu, 15 Jan 2009 10:06:42 +0000 http://blog.leeromero.org/?p=246#comment-2341 In developing the British Standard for structured vocabularies, BS8723, we produced a data model for a thesaurus, as well as an XML schema derived from it. These were published in DD8723 (produced as a "Draft for Development" rather than a British Standard as we did not consider that it was sufficiently developed at that stage). This is available at http://schemas.bs8723.org/, and comments are welcome. Subsequently, on work for the corresponding international standard ISO25964, we developed the model further. As the ISO standard is still at draft stage we cannot make it public (regrettably!) but you can see the current draft data model attached to a message I sent to the SKOS mailing list at http://lists.w3.org/Archives/Public/public-esw-thes/2008Dec/0003.html We think that this model represents all that is needed for a standard thesaurus, including several elements which are not covered by SKOS. It may be expanded later to cover other types of structured vocabulary such as classification schemes and other systems using pre-coordinated headings. Do you think that this gives a sound foundation on which XML schemas or other exchange formats can be built? Leonard Will In developing the British Standard for structured vocabularies, BS8723, we produced a data model for a thesaurus, as well as an XML schema derived from it. These were published in DD8723 (produced as a “Draft for Development” rather than a British Standard as we did not consider that it was sufficiently developed at that stage). This is available at http://schemas.bs8723.org/, and comments are welcome.

Subsequently, on work for the corresponding international standard ISO25964, we developed the model further. As the ISO standard is still at draft stage we cannot make it public (regrettably!) but you can see the current draft data model attached to a message I sent to the SKOS mailing list at http://lists.w3.org/Archives/Public/public-esw-thes/2008Dec/0003.html

We think that this model represents all that is needed for a standard thesaurus, including several elements which are not covered by SKOS. It may be expanded later to cover other types of structured vocabulary such as classification schemes and other systems using pre-coordinated headings.

Do you think that this gives a sound foundation on which XML schemas or other exchange formats can be built?

Leonard Will

]]>