Original HC term recording or measurement agent type definition:
The agent taking the recording. This can be "human" or "machine" (e.g., camera traps)
(examples wrongly and only included in definition above).
This term can be tricky. It may be kept "simple" if we only want to denote if it was human or machine taking the record (?). But a more complicated path can be considered, in relation with definition/use of basisOfRecord. I'll list here just some considerations:
The Event Core does not include the term basisOfRecord. basisOfRecord allows values that are the DwC classes (including "Event"), and therefore the term is probably not included in the Event Core because it would be redundant.
However, an event can be of other of the "bases" (e.g., MachineObservation, HumanObservation, etc.).
Including basisOfRecord in the extension could be considered, but how would we not default it to value "Event"?
basisOfRecord does not currently have an ontology behind it.
Note that even though recommended vocab are the DwC classes, other values are currently in broad use, e.g., Literature and Unknown (see GBIF basisOfRecord vocab)
As from the first review round, we identified 3 possible routes to follow:
Keep this term in the HC extension but in a very simple way, something that basically would only indicate if a camera trap was involved. This would have to be done carefully to not confuse with basisOfRecord as current, and would probably overlap with sampling protocol.
Deal with basisOfRecord once and for all, for DwC in general. This would mean reviewing its definition, probably defining an ontology, etc. LONG process, with many challenges, but probably many BIG gains all over.
Propose a term within the HC extension that would momentarily take care of the issue, and which could be deprecated in the future when the basisOfRecord big issue is resolved. For this option a proposal is to create a basisOfEvent term, with definition in the lines of:
The nature of the sampling event(s) referred to in the record, regarding the type of agent that performed the sampling.
Original HC term recording or measurement agent type definition:
The agent taking the recording. This can be "human" or "machine" (e.g.,
camera traps)
(examples wrongly and only included in definition above).
This term can be tricky. It may be kept "simple" if we only want to denote
if it was human or machine taking the record (?). But a more complicated
path can be considered, in relation with definition/use of basisOfRecord.
I'll list here just some considerations:
The Event Core does not include the term basisOfRecord. basisOfRecord
https://dwc.tdwg.org/terms/#dwc:basisOfRecord allows values that are
the DwC classes (including "Event"), and therefore the term is probably not
included in the Event Core because it would be redundant.
However, an event can be of other of the "bases" (e.g.,
MachineObservation, HumanObservation, etc.).
Including basisOfRecord in the extension could be considered, but
how would we not default it to value "Event"?
basisOfRecord does not currently have an ontology behind it.
As from the first review round, we identified 3 possible routes to follow:
1.
Keep this term in the HC extension but in a very simple way, something
that basically would only indicate if a camera trap was involved. This
would have to be done carefully to not confuse with basisOfRecord as
current, and would probably overlap with sampling protocol.
2.
Deal with basisOfRecord once and for all, for DwC in general. This
would mean reviewing its definition, probably defining an ontology, etc.
LONG process, with many challenges, but probably many BIG gains all over.
3.
Propose a term within the HC extension that would momentarily take
care of the issue, and which could be deprecated in the future when the
basisOfRecord big issue is resolved. For this option a proposal is to
create a basisOfEvent term, with definition in the lines of:
The nature of the sampling event(s) referred to in the record,
regarding the type of agent that performed the sampling.
I agree that avoiding basisOfRecord is a sensible way to be able to progress this.
I wonder if we might want to consider alternatives to basisOfEvent? I suspect basisOfRecord was misunderstood because it was so generic and we might be repeating previous mistakes.
If the intention is to capture the type of "thing" creating the samples or measurements perhaps we should be specific in that; such as typeOfRecorder with a controlled vocabulary (camera trap, human, radar, laser, microphone etc)?
Original HC term
recording or measurement agent typedefinition:(examples wrongly and only included in definition above).
This term can be tricky. It may be kept "simple" if we only want to denote if it was human or machine taking the record (?). But a more complicated path can be considered, in relation with definition/use of
basisOfRecord. I'll list here just some considerations:As from the first review round, we identified 3 possible routes to follow:
Keep this term in the HC extension but in a very simple way, something that basically would only indicate if a camera trap was involved. This would have to be done carefully to not confuse with basisOfRecord as current, and would probably overlap with sampling protocol.
Deal with basisOfRecord once and for all, for DwC in general. This would mean reviewing its definition, probably defining an ontology, etc. LONG process, with many challenges, but probably many BIG gains all over.
Propose a term within the HC extension that would momentarily take care of the issue, and which could be deprecated in the future when the basisOfRecord big issue is resolved. For this option a proposal is to create a
basisOfEventterm, with definition in the lines of:For practical plus forward-pointing reasons, I like option 3. I don't think this Task Group should have to solve or wait on the bigger issue.
On Sat, Feb 27, 2021 at 6:33 AM Paula Zermoglio notifications@github.com wrote:
I agree that avoiding
basisOfRecordis a sensible way to be able to progress this.I wonder if we might want to consider alternatives to
basisOfEvent? I suspectbasisOfRecordwas misunderstood because it was so generic and we might be repeating previous mistakes.If the intention is to capture the type of "thing" creating the samples or measurements perhaps we should be specific in that; such as
typeOfRecorderwith a controlled vocabulary (camera trap, human, radar, laser, microphone etc)?The term dwc:eventType, organized in the dwc:Event class, has been added to Darwin Core.