[tdwg-guid] taxon concepts in dublin core
![](https://secure.gravatar.com/avatar/fc12e73649ec76224e65b0274a9b53d3.jpg?s=120&d=mm&r=g)
Dear all,
This is a bit OT, but at least about metadata: I am curently preparing some data in TCS schema and I plan to set up a OAI interface for data exchange. According to the OAI specifications the use of dublin (NOT darwin..) core formatted metadata is obligatory, so I will also try to offer oai_dc metadata.
I would like to ask the group if there have been any discussions on mapping TCS to DC before? Do you know of any attempts to model concept or name metadata with DC?
many thanks in advance, Robert
![](https://secure.gravatar.com/avatar/5e6ffc74a38184a3ac7494fe6f511065.jpg?s=120&d=mm&r=g)
Robert
I may be misuderstanding what you are trying to do here, but there is little point in trying to "map" TCS to Dublin Core. The theory is that if there were any "overlapping" properties in the two schemas then those elements/properties should have been ommitted from the TCS schema in the first place and dublin core elements promoted in their place. As Hannu suggested, you could attempt to put a few of the major properties, such as the taxon name or GUID, into dublin core elements, but there is no way you will "map TCS to Dublin Core" completely.
The reason for promoting DC in OAI is to ensure that there is at least one schema that can be queried on an OAI service that is relatively common. This schema would only be used for metadata about the service, and TCS used for the data.
As discussed on the LSID Vocabularies wiki (http://wiki.tdwg.org/twiki/bin/view/TAG/LsidVocs), the best approach is to build a model using elements from DC where possible, and create new elements for those not covered by DC elements. This has been the approach when creating the LSID RDF schemas, but probably not as much for the TCS schema.
Kevin
"Robert Huber" rhuber@wdc-mare.org 14/02/2008 2:55 a.m. >>>
Dear all,
This is a bit OT, but at least about metadata: I am curently preparing some data in TCS schema and I plan to set up a OAI interface for data exchange. According to the OAI specifications the use of dublin (NOT darwin..) core formatted metadata is obligatory, so I will also try to offer oai_dc metadata.
I would like to ask the group if there have been any discussions on mapping TCS to DC before? Do you know of any attempts to model concept or name metadata with DC?
many thanks in advance, Robert
![](https://secure.gravatar.com/avatar/fc12e73649ec76224e65b0274a9b53d3.jpg?s=120&d=mm&r=g)
Thank you Kevin!
Yes you are right, it makes no sense trying to map TCS to DC completely, I was just trying to find a way to offer something in DC which makes most sense.
I was thinking of something like this:
<metadata> <oai_dc:dc xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/oai_dc/ http://www.openarchives.org/OAI/2.0/oai_dc.xsd%22%3E dc:formattext/html</dc:format> dc:publisherStratigraphy.Net http://taxonconcept.stratigraphy.net </dc:publisher> dc:typeTaxon Concept</dc:type> dc:titleGloborotalia mendacis</dc:title> dc:date1969</dc:date> dc:creatorBlow</dc:creator> dc:identifier http://taxonconcept.stratigraphy.net/taxon_details.php?show_all=1&taxid=... </dc:identifier> dc:source Quilty, P.G.. (1976): Planctonic foraminifera DSDP Leg 34- Nazca Plata . DSDP initial reports Vol. 34 </dc:source> dc:relationGloborotalia (Turborotalia) mendacis Blow 1969</dc:relation> dc:subjectEukaryota</dc:subject> dc:subjectProtoctista</dc:subject> dc:subjectGranuloreticulosa</dc:subject> dc:subjectForaminifera</dc:subject> dc:subjectGlobigerinida</dc:subject> dc:subjectGloborotaliaceae</dc:subject> dc:subjectGloborotaliidae</dc:subject> dc:subjectGloborotalia</dc:subject> dc:subjectGloborotalia mendacis</dc:subject> </oai_dc:dc> </metadata>
where I put the informtion about the taxon in the title, creator and date fields, the subject fields contain the classification and the relation tag contains e.g synonymy info . The source tag would contain the reference of a taxon concept. Does this make sense?
Robert
2008/2/13, Kevin Richards RichardsK@landcareresearch.co.nz:
Robert
I may be misuderstanding what you are trying to do here, but there is little point in trying to "map" TCS to Dublin Core. The theory is that if there were any "overlapping" properties in the two schemas then those elements/properties should have been ommitted from the TCS schema in the first place and dublin core elements promoted in their place. As Hannu suggested, you could attempt to put a few of the major properties, such as the taxon name or GUID, into dublin core elements, but there is no way you will "map TCS to Dublin Core" completely.
The reason for promoting DC in OAI is to ensure that there is at least one schema that can be queried on an OAI service that is relatively common. This schema would only be used for metadata about the service, and TCS used for the data.
As discussed on the LSID Vocabularies wiki ( http://wiki.tdwg.org/twiki/bin/view/TAG/LsidVocs), the best approach is to build a model using elements from DC where possible, and create new elements for those not covered by DC elements. This has been the approach when creating the LSID RDF schemas, but probably not as much for the TCS schema.
Kevin
"Robert Huber" rhuber@wdc-mare.org 14/02/2008 2:55 a.m. >>>
Dear all,
This is a bit OT, but at least about metadata: I am curently preparing some data in TCS schema and I plan to set up a OAI interface for data exchange. According to the OAI specifications the use of dublin (NOT darwin..) core formatted metadata is obligatory, so I will also try to offer oai_dc metadata.
I would like to ask the group if there have been any discussions on mapping TCS to DC before? Do you know of any attempts to model concept or name metadata with DC?
many thanks in advance, Robert
-- Dr. Robert Huber,
WDC-MARE / PANGAEA - www.pangaea.de Stratigraphy.net - www.stratigraphy.net _____________________________________________ MARUM - Institute for Marine Environmental Sciences (location) University Bremen Leobener Strasse POP 330 440 28359 Bremen Phone ++49 421 218-65593, Fax ++49 421 218-65505 e-mail rhuber@wdc-mare.org
![](https://secure.gravatar.com/avatar/5e6ffc74a38184a3ac7494fe6f511065.jpg?s=120&d=mm&r=g)
Depends what you are trying to acheive here. Why not just use the TCS schema for anything other than the basic DC metadata? The caller would then have to specify the TCS schema to be used for the response.
"Robert Huber" rhuber@wdc-mare.org 14/02/2008 9:18 a.m. >>>
Thank you Kevin!
Yes you are right, it makes no sense trying to map TCS to DC completely, I was just trying to find a way to offer something in DC which makes most sense.
I was thinking of something like this:
<metadata> <oai_dc:dc xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/oai_dc/ http://www.openarchives.org/OAI/2.0/oai_dc.xsd%22%3E dc:formattext/html</dc:format> dc:publisherStratigraphy.Net http://taxonconcept.stratigraphy.net</dc:publisher> dc:typeTaxon Concept</dc:type> dc:titleGloborotalia mendacis</dc:title> dc:date1969</dc:date> dc:creatorBlow</dc:creator> dc:identifierhttp://taxonconcept.stratigraphy.net/taxon_details.php?show_all=1&taxid=...</dc:identifier> dc:source Quilty, P.G.. (1976): Planctonic foraminifera DSDP Leg 34- Nazca Plata . DSDP initial reports Vol. 34 </dc:source> dc:relationGloborotalia (Turborotalia) mendacis Blow 1969</dc:relation> dc:subjectEukaryota</dc:subject> dc:subjectProtoctista</dc:subject> dc:subjectGranuloreticulosa</dc:subject> dc:subjectForaminifera</dc:subject> dc:subjectGlobigerinida</dc:subject> dc:subjectGloborotaliaceae</dc:subject> dc:subjectGloborotaliidae</dc:subject> dc:subjectGloborotalia</dc:subject> dc:subjectGloborotalia mendacis</dc:subject> </oai_dc:dc> </metadata>
where I put the informtion about the taxon in the title, creator and date fields, the subject fields contain the classification and the relation tag contains e.g synonymy info . The source tag would contain the reference of a taxon concept. Does this make sense?
Robert
2008/2/13, Kevin Richards RichardsK@landcareresearch.co.nz: Robert
I may be misuderstanding what you are trying to do here, but there is little point in trying to "map" TCS to Dublin Core. The theory is that if there were any "overlapping" properties in the two schemas then those elements/properties should have been ommitted from the TCS schema in the first place and dublin core elements promoted in their place. As Hannu suggested, you could attempt to put a few of the major properties, such as the taxon name or GUID, into dublin core elements, but there is no way you will "map TCS to Dublin Core" completely.
The reason for promoting DC in OAI is to ensure that there is at least one schema that can be queried on an OAI service that is relatively common. This schema would only be used for metadata about the service, and TCS used for the data.
As discussed on the LSID Vocabularies wiki (http://wiki.tdwg.org/twiki/bin/view/TAG/LsidVocs), the best approach is to build a model using elements from DC where possible, and create new elements for those not covered by DC elements. This has been the approach when creating the LSID RDF schemas, but probably not as much for the TCS schema.
Kevin
"Robert Huber" rhuber@wdc-mare.org 14/02/2008 2:55 a.m. >>>
Dear all,
This is a bit OT, but at least about metadata: I am curently preparing some data in TCS schema and I plan to set up a OAI interface for data exchange. According to the OAI specifications the use of dublin (NOT darwin..) core formatted metadata is obligatory, so I will also try to offer oai_dc metadata.
I would like to ask the group if there have been any discussions on mapping TCS to DC before? Do you know of any attempts to model concept or name metadata with DC?
many thanks in advance, Robert
participants (3)
-
Hannu Saarenmaa
-
Kevin Richards
-
Robert Huber