search

iobis / obis-issues

Repository for all OBIS related issues and feature requests
4 stars 3 forks source link

OBIS manual lists incorrect use for DwC: type #172

Open albenson-usgs opened 4 years ago

albenson-usgs commented 4 years ago

The OBIS manual lists incorrect vocabulary for the DwC term type and needs to be updated.

Capture
pieterprovoost commented 4 years ago

Thanks, I have removed the reference to dcterms:type but I have kept the column as eventRemarks for clarity.

Maybe @bart-v should review this as I think EurOBIS uses type extensively.

bart-v commented 4 years ago

Hmm, this has quite an impact: we have been doing this for a few years now. But yes, dcterms:type was not correct, indeed.

Can't we find a better field? How about dwc:samplingProtocol (we recommend the actual sampling protocol info to go into the eMoF)

wardappeltans commented 4 years ago

At the time, this was what was agreed and recommended to define what type of event it was. Maybe eventRemarks is best. I don't think it is samplingProtocol.

skybristol commented 4 years ago

This is one of those areas where the ambiguities in the DwC drive me absolutely nuts! The documentation for samplingProtocol is vague enough to sail a ship through. We need to be moving toward a preference for the referral method where we point particularly vital things like observation protocols at a persistent and resolvable identifier like a documented protocol in protocols.io that the Microbe EOV group is using or the closer to home Ocean Best Practices System. Ideally, that would end up spooling out into a limited number of deeper structured data models with protocol details that can be leveraged for filtering data. At the very least, it would be useful to split concepts out in the DwC so that we have discrete properties for things that are resolvable identifiers vs. loose text; something that many other standards have gone toward.

In the near term, it would be useful to add an introspection routine into OBIS data processing for most/all of the fields we are indexing. I would look to do a couple of primary things with that:

One other reply on @wardappeltans idea on eventRemarks. We should not use properties from a standard that are intended for pure human consumption as prose to stuff in what would otherwise be controlled vocabulary terms/identifiers. That would be a functionally inappropriate application of the standard and would confuse uses of the data model. We should find a legitimate location for this information or work within the standards body to develop what we need.

albenson-usgs commented 4 years ago

I feel this question needs to be moved to the tdwg-qa as it seems like something we need to get input on from the broader TDWG community before making decisions. I can post it there if others agree?

albenson-usgs commented 4 years ago

https://github.com/tdwg/dwc-qa/issues/157

ClintBlight commented 1 year ago

Thanks, I have removed the reference to dcterms:type but I have kept the column as eventRemarks for clarity.

Maybe @bart-v should review this as I think EurOBIS uses type extensively.

I found this issue after noticing that that embedded tables in the current version of https://manual.obis.org/examples.html have an "eventRemarks" column but some of the text still refers to "type" being used. Posting here as it is still flagged as an open issue.

The couple of places in the text which might need a slight edit look to be:

2.2.4.11 Phytoplankton biomass & diversity "In this case, the the event type information is provided in type." 2.2.4.13 Zooplankton biomass & diversity "In this case, the the event type information is provided in type."