What is the `type` field?

[SimonGoring]

The type field for the Occurrence is causing some questions. Do we use the controlled terminology based on the Dublin Core terminology as listed here or is Occurrence an accepted field? How do we record actual physical specimens?

http://tools.gbif.org/dwca-validator/extension.do?id=dwc:Occurrence

[SimonGoring]

On the Neotoma (e.g., @SimonGoring or @IceAgeEcologist) side we had elected to use "Dataset", part of the Dublin Core controlled vocabulary, but as @mmcclenn pointed out, it may be better suited to an "Occurrence" type.

@tucotuco, any chance you might weigh in on this?

[tucotuco]

Hi folks,

The Dublin Core type vocabulary applies strictly to the dcterms:type term. Occurrence is a Darwin Core class and not a valid value for that term. To be more specific than dcterms:type allows, Darwin Core has the basisOfRecord term (http://rs.tdwg.org/dwc/terms/index.htm#basisOfRecord), whose range includes the Darwin Core classes PreservedSpecimen, LivingSpecimen, FossilSpecimen, MaterialSample, MachineObservation, and HumanObservation. The range of the dcterms:type term include Event, Dataset, PhysicalObject, Sound, StillImage, MovingImage, and Text.
For fossil specimens, then, we would use the combination of dcterms:type=PhysicalObject and dwc:basisOfRecord=FossilSpecimen. By contrast, if a biodiversity data record was something taken out of a manuscript (no physical specimen), then we would use dcterms:type=Text dwc:basisOfRecord=HumanObservation.

Does that help?

[SimonGoring]

Thanks @tucotuco, I think that makes sense. Makes it a bit more complicated for us though :)

I think we need to split the type and basis of record by datasettype then. I wonder what we should do about records in the PaleoDB that have come in from DeepDive, or that were entered from published records.

I assume most fossil occurrences come from physical specimens, as do our pollen records (for example), but some must come from only a text record without knowledge of the actual physical specimen. Is that correct @mmcclenn @cambro or @markuhen ?

So we'd need some sort of scheme to differentiate between:

dcterms:type=PhysicalObject and dwc:basisOfRecord=FossilSpecimen
-vs-
dcterms:type=Text and dwc:basisOfRecord=HumanObservation

Do I have the gist of this right?

[tucotuco]

To me, yes, that is the correct use of the type fields.

On Fri, Jul 29, 2016 at 12:22 AM, Simon notifications@github.com wrote:

Thanks @tucotuco https://github.com/tucotuco, I think that makes sense.
Makes it a bit more complicated for us though :)

I think we need to split the type and basis of record by datasettype then.
I wonder what we should do about records in the PaleoDB that have come in
from DeepDive, or that were entered from published records.

I assume most fossil occurrences come from physical specimens, as do our
pollen records (for example), but some must come from only a text record
without knowledge of the actual physical specimen. Is that correct
@mmcclenn https://github.com/mmcclenn @cambro
https://github.com/cambro or @markuhen https://github.com/markuhen ?

So we'd need some sort of scheme to differentiate between:

dcterms:type=PhysicalObject and dwc:basisOfRecord=FossilSpecimen
-vs-
dcterms:type=Text and dwc:basisOfRecord=HumanObservation

Do I have the gist of this right?

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
#1 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/AAcP6xnaqvdjPAJSej53P09mZaQbWqlJks5qaSudgaJpZM4JS1NF
.

[markuhen]

All,

Yes. For the most part, all PBDB records are ultimately based on a
physical specimen. I could imagine that some could be characterized as
observations, but very few. For instance, I'm thinking of dinosaur
footprints that are left in the field. They are still based on a
specimen though, even though it wasn't collected and placed in a repository.

Thanks,

Mark

On 7/28/16 6:22 PM, Simon wrote:

Thanks @tucotuco https://github.com/tucotuco, I think that makes
sense. Makes it a bit more complicated for us though :)

I think we need to split the type and basis of record by datasettype
then. I wonder what we should do about records in the PaleoDB that have
come in from DeepDive, or that were entered from published records.

I assume most fossil occurrences come from physical specimens, as do our
pollen records (for example), but some must come from only a text record
without knowledge of the actual physical specimen. Is that correct
@mmcclenn https://github.com/mmcclenn @cambro
https://github.com/cambro or @markuhen https://github.com/markuhen ?

So we'd need some sort of scheme to differentiate between:

dcterms:type=PhysicalObject and dwc:basisOfRecord=FossilSpecimen
-vs-
dcterms:type=Text and dwc:basisOfRecord=HumanObservation

Do I have the gist of this right?

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
#1 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/ASpsTC7CzYRXdr2BBB_R6TKuxcO8cjx4ks5qaSudgaJpZM4JS1NF.

Mark D. Uhen
Assistant Professor
George Mason University
AOES Geology
MSN 6E2
Fairfax, VA 22030
Phone: 703-993-5264
Fax: 703-993-3535

[SimonGoring]

I agree with that, I'm just wondering if we need to make a distinction between records that were entered where we know the specimen, or its general location (museum, or whatever) vs. something that was pulled in programmatically.

[markuhen]

Simon,

Yes. I think we need to distinguish between these two types of records.
There are LOTS of reasons to want to know the difference, and it's the
reasons I can't think of that make me want to keep track of the
different types!

Thanks much,

Mark

On 8/1/16 11:27 AM, Simon wrote:

I agree with that, I'm just wondering if we need to make a distinction
between records that were entered where we know the specimen, or its
general location (museum, or whatever) vs. something that was pulled in
programmatically.

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
#1 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/ASpsTJWnBtugerKdrzh1exRCRNVOZjsWks5qbhBPgaJpZM4JS1NF.

Mark D. Uhen
Assistant Professor
George Mason University
AOES Geology
MSN 6E2
Fairfax, VA 22030
Phone: 703-993-5264
Fax: 703-993-3535