Hi,
Well, I do understand your concern. what we choose has much to do with how we read (interpreted) the recommendations.
"Recommended best practice is to use a controlled vocabulary such as the list of Darwin Core classes. Examples: "PreservedSpecimen", "FossilSpecimen", "LivingSpecimen", "HumanObservation", "MachineObservation". "
I read examples, and the only term which is corrupted is (the one we invented, years ago already..) is "literatureObservation" which, in my thinking did follow the line of the examples. I did not figured out that event:text en basisOfRecord: occurrence would de. Also here, occurrence is not in the list...
That is the reason, not because we wanted to do something contrary. :)
Also, I have no problem in using controlled vocabularies, when they suit my need. And reading the DwC explanation is states "Recommended best practice is to use a controlled vocabulary "
Chrs, Dimi
On Fri, Jun 26, 2015 at 10:35 AM, John Wieczorek tuco@berkeley.edu wrote:
I understand on one hand, but on the other, why not follow the recommendations to begin with, or make an effort to get them changed if they did not suffice? Not trying to be snotty, just wondering.
On Fri, Jun 26, 2015 at 9:30 AM, Dimitri Brosens dimitri.brosens@inbo.be wrote:
All,
This means that everybody who used 'basisOfRecord' as defined in the Darwin Core (best to use controlled voc:) and republishes data could reach this issue in IPT (enforced to use controlled voc).... can encounter this problem. I don't know if I 'm happy with this. I need to change all the published data because we used literatureObservation whenever occurrences originated in literature... And it will eventually change the records which were published before...
Chrs, Dimi
On Thu, Jun 25, 2015 at 5:56 PM, John Wieczorek tuco@berkeley.edu wrote:
Quite some time ago the Dublin Core type vocabulary and Darwin Core type vocabulary were relegated to control distinct terms in Darwin Core. This meant that Dublin Core type values were recommended only for dcterms:type and not for basisOfRecord. Thus, a record of a camera trap image would have dcterms:type=StillImage and basisOfRecord=MachineObservation. Though the Darwin Core type vocabulary was deprecated, the recommendation (and the strength of it) remains. Following the recommendation (which GBIF's IPT implements and enforces), non-standard values are not allowed.
My recommendation for the non-standard basisOfRecord values is to change to the following combinations:
'literatureObservation' - dcterms:type=Text, basisOfRecord=HumanObservation 'literature' - dcterms:type=Text basisOfRecord=Occurrence (if you can't be more specific) 'fieldObservation' - dcterms:type=Event, basisOfRecord=HumanObservation 'unknown' - dcterms:type=null, basisOfRecord=Occurrence
On Thu, Jun 25, 2015 at 5:13 PM, Paul J. Morris mole@morris.net wrote:
fieldObservation probably maps sanely onto HumanObservation.
unknown maps sanely onto Occurrence "NOTE: this value is ambiguous and hence should only be used when the when the resource type is unknown"
This type vocabulary doesn't include a sane type for data derived from the literature. This type vocabulary claims to extend the Dublin Core type vocabulary, so data sourced from the literature should probably be able to map to Text http://purl.org/dc/dcmitype/Text. Whether IPT, on a technical level, treats the DarwinCore type vocabulary as an extension of the DublinCore type vocabulary and treats values from dcmitype as valid is another question.
-Paul
On Thu, 25 Jun 2015 16:58:42 +0200 Kyle Braak kbraak@gbif.org wrote:
Dear Dimitri,
To satisfy the IPT’s requirements, you can translate your source’s values into values that match the Darwin Core Type vocabulary [1]. More information about the translation feature is available in the mapping section of the user manual [2].
Best regards, Kyle
[1] http://rs.gbif.org/vocabulary/dwc/basis_of_record.xml [2]
https://code.google.com/p/gbif-providertoolkit/wiki/IPT2ManualNotes?tm=6#Dar...
On 25 Jun 2015, at 16:21, Dimitri Brosens dimitri.brosens@inbo.be wrote:
Dear,
I was wondering, in DwC it is stated:
Recommended best practice is to use a controlled vocabulary such as the list of Darwin Core classes. Examples: "PreservedSpecimen", "FossilSpecimen", "LivingSpecimen", "HumanObservation", "MachineObservation". For discussion see http://terms.tdwg.org/wiki/dwc:basisOfRecord
As far as I understand, this is not mandatory and we have been using terms which are not in the 'supposed' controlled vocabulary on the tdwg site....
Publishing version #22.1 of resource belgian-coccinellidae-inbo-occurrences failed: Archive generation for resource belgian-coccinellidae-inbo-occurrences failed: Can't validate DwC-A for resource belgian-coccinellidae-inbo-occurrences. Each row in the occurrence file(s) must have a basisOfRecord, and each basisOfRecord must match the Darwin Core Type Vocabulary (please note comparisons are case insensitive) Continue to resource overview.
My problem is that IPT refuses to republish my previously published datasets where we use terms like: 'literatureObservation' , 'literature', fieldObservation or 'unknown
What to do?
Chrs, Dimi
-- Belgian Biodiversity Platform
Dimitri Brosens
Biodiversity Research Liaison Officer
Research Institute for Nature and Forest Kliniekstraat 25 1070 Brussels
ORCID: 0000-0002-0846-9116
www.inbo.be www.biodiversity.be www.beescommunity.be
WATCH OUR BELGIAN BIODIVERSITY PLATFORM MOVIES ON: http://vimeo.com/114955090 (data publication activities) http://vimeo.com/114955160 (science-policy activities) http://vimeo.com/114955193 (our general mission)
IPT mailing list IPT@lists.gbif.org http://lists.gbif.org/mailman/listinfo/ipt
-- Paul J. Morris Biodiversity Informatics Manager Harvard University Herbaria/Museum of Comparative Zoölogy mole@morris.net AA3SD PGP public key available _______________________________________________ IPT mailing list IPT@lists.gbif.org http://lists.gbif.org/mailman/listinfo/ipt
IPT mailing list IPT@lists.gbif.org http://lists.gbif.org/mailman/listinfo/ipt
--
Belgian Biodiversity Platform
Dimitri Brosens
Biodiversity Research Liaison Officer
Research Institute for Nature and Forest Kliniekstraat 25 1070 Brussels
ORCID: 0000-0002-0846-9116
www.inbo.be www.biodiversity.be www.beescommunity.be *WATCH OUR BELGIAN BIODIVERSITY PLATFORM MOVIES ON:* http://vimeo.com/114955090 (data publication activities) http://vimeo.com/114955160 (science-policy activities) http://vimeo.com/114955193 (our general mission)