search

ioos / ioos-metadata

Documentation site for the IOOS Metadata Profile
https://ioos.github.io/ioos-metadata
1 stars 9 forks source link

Indigenous metadata considerations #35

Open MathewBiddle opened 2 years ago

MathewBiddle commented 2 years ago

This partner is closely engaged with First Nations and had some questions about the metadata. These are based on the metadata profile 1.2, mostly focused in the attribution section.

  1. For the controlled "sector" options, "tribal" is understandably U.S. focused. However, a change from "tribal" to "indigenous" and including "Tribes and First Nations" in the description would be more inclusive.
  2. For "country" fields, the partner that we are working with said the country should be the First Nation, rather than Canada. This raised the question in our minds around if a First Nation was listed, it would perhaps create some confusion to some users who may think it was a typo or aren't familiar with the First Nation (I am certainly not fully aware of these). Anyways, would that be allowable? Or would IOOS rather quality "country" with "Western"? There are likely other solutions as well, but I'm curious if IOOS has started thinking about this yet or not.

@mwengren any thoughts?

MathewBiddle commented 2 years ago

@emiliom any experience with this from your NANOOS days?

mwengren commented 2 years ago

@MathewBiddle The only legacy connection I know of to the sector attribute we use in the IOOS Metadata Profile is to the IOOS SOS/SWE schema/guidelines. That can be found here (look for operatorSector).

AFAIK, sector was carried over from the SOS/SWE guideline to the IOOS Metadata Profile so that a netCDF dataset that met the requirements of the original 1.1 version of the IOOS Metadata Profile could be converted to match the SOS/SWE conventions XML formats using ncSOS software.

Since we're no longer requiring SOS/SWE, I think we could drop sector attribute entirely in a future version of the IOOS Metadata Profile if it was no longer important to us, or we could update the vocabulary in MMISW if we have access to do that. Either option seems reasonable.

emiliom commented 2 years ago

We did use the "tribal" sector value for a couple of assets in NANOOS. It didn't raise issues with our partners, BUT, they probably were not aware of the existence of this metadata. Still, this being a fully US context, that part would probably not have been an issue. Changing it to "indigenous" would make sense, though I suggest that would be an opportunity to engage tribal partners about this terminology and associated descriptions. I'd be happy to facilitate some connections if you'd like. You'd also have to think about what to do with existing uses of "tribal".

I believe the motivation for having a sector controlled vocabulary at all came from the IOOS asset inventory needs. If that need is still there (I imagine it is), I'd think you'd want to retain it in the IOOS Metadata Profile.

The issue with "country" seems much trickier, and one I never dealt with. I imagine that not having US or Canada (as the case may be) in a machine actionable form under "country" could easily lead to unintended results when querying and inventorying assets. I wouldn't be surprised if there exists a clear (if not widely adopted) terminology that represents the relationship and distinction between a country and an indigenous nation. For example, in our usual contexts, country is an entity that has a ISO 3166-1 alpha-2 country code. Anyway, that's out of my depth.

MathewBiddle commented 2 years ago

@emiliom and @mwengren thank you both for the context and discussion points.

As I can't seem to find other examples of these changes I think it would be worthwhile for the IOOS community to discuss and explore the options. From what I can see now, here are the action items to be discussed:

mwengren commented 2 years ago

@MathewBiddle Thinking of options to resolve the sector vocabulary question: what if we add more terms to the corresponding vocabulary, rather than changing the existing tribal term? See: https://mmisw.org/ont/ioos/sector. Would that impact the Asset Inventory if indigenous or first_nation were both added?

For the creator_country and publisher_country attributes in the IOOS Metadata Profile, I think I agree with @emiliom that those were intended to be countries recognized in the ISO standard he cited, so they could be mapped to a traditional contact info XML blurb such as the example below from the IOOS SOS Guidelines:

<sml:ResponsibleParty>
  <sml:organizationName>PNW Buoys</sml:organizationName>
  <sml:contactInfo>
    <sml:address>
      <sml:deliveryPoint>1007 Balch Blvd.</sml:deliveryPoint>
      <!-- Optional: City; but strongly encouraged  -->
      <sml:city>Fremont</sml:city>
      <sml:administrativeArea>WA</sml:administrativeArea>
      <sml:postalCode>98195</sml:postalCode>
      <!-- Required: country [Values: USA|<COUNTRY NAME>|NON-USA] -->
      <sml:country>USA</sml:country>
      <!-- Required: electronicMailAddress -->
      <sml:electronicMailAddress>contact@buoys.com</sml:electronicMailAddress>
    </sml:address>
    <!-- Optional: onlineResource; but strongly encouraged for operator -->
    <sml:onlineResource xlink:href="http://pnw.buoyoperator.org"/>
  </sml:contactInfo>
</sml:ResponsibleParty>

Having said that though, we don't mention that vocabulary restriction anywhere in the attribute descriptions like you mentioned, so that was probably just assumed. Technically speaking, anything could be entered in those attributes and they would be 'compliant' with our profile.

Not sure how to handle that one. Maybe we can leave it open ended as it is and as long as data providers understand the purpose of the attribute is to create a mailing address-like metadata collection, they can include any country identifier they prefer?

Any other ideas out there?

MathewBiddle commented 2 years ago

@mwengren I haven't used Operator sector for a request on the asset inventories. That doesn't mean we don't need it, but we probably are okay with adding in new terms - as long as we're not deprecating or changing the meaning of the existing one.

For reference, here are the unique Operator_sector terms we currently have in the inventory (2019-2021) http://erddap.ioos.us/erddap/tabledap/raw_asset_inventory.htmlTable?Operator_Sector&distinct()&orderBy(%22Operator_Sector%22)

@kbailey-noaa do you have any insights into how we use the Operator_sector for the Asset Inventory?

MathewBiddle commented 1 year ago

@kbailey-noaa is okay with making these adjustments.

mwengren commented 1 year ago

@emiliom @fgayanilo I was going to update the terms in https://mmisw.org/ont/ioos/sector as described here, but I need permission to edit the vocabularies owned by 'ioos' in MMISW. How does one get access to this? My newly-created account ID on MMISW is: mwengren.

@emiliom It appears you are the author of the sector vocabulary, that's why I included you in my request.

TIA!

MathewBiddle commented 1 year ago

Could you add me too? username: mathewbiddle

fgayanilo commented 1 year ago

there is an account for user= ioos with admin access for the collection. You can reset the password -- I think the account is registered to Derrick (@carueda I lost my "Admin" tab)

carueda commented 1 year ago

@fgayanilo – We (I and @graybeal) would be happy to assist with https://mmisw.org/ont stuff but I'm afraid I don't understand what the specific need is.

fgayanilo commented 1 year ago

@carueda NOAA IOOS wants to update the vocabulary they maintain (IOOS collections). Matt Biddle (username=mathewbiddle) and Micah Wengren (username=mwengren) want to update their collection. We created an account with username=ioos (if my memory serves me right) for this purpose, but I cannot remember the profile for that user. Perhaps, you can help reset the password for that account.

carueda commented 1 year ago

Thanks @fgayanilo ...

I just added those users as members to the ioos organization.

image

graybeal commented 1 year ago

That's amazing. I was just thinking during ESIP SemTech meeting yesterday about IOOS as an example of a group that published their vocabularies (some)!

Would be great to have that content updated.

John

On Nov 9, 2022, at 9:25 AM, Felimon Gayanilo @.**@.>> wrote:

NOAA IOOS wants to update the vocabulary they maintain (IOOS collections). Matt Biddle (username=mathewbiddle) and Micah Wengren (username=mwengren) want to update their collection. We created an account with username=ioos (if my memory serves me right) for this purpose, but I cannot remember the profile for that user. Perhaps, you can help reset the password for that account.

— Reply to this email directly, view it on GitHubhttps://github.com/ioos/ioos-metadata/issues/35#issuecomment-1309092545, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AAJVJUASJQKZWOA343QDUETWHPNABANCNFSM5YOWQFIQ. You are receiving this because you were mentioned.Message ID: @.***>

======================== John Graybeal Technical Program Manager Center for Expanded Data Annotation and Retrieval /+/ NCBO BioPortal Stanford Center for Biomedical Informatics Research 650-736-1632 | ORCID 0000-0001-6875-5360

mwengren commented 1 year ago

Thanks, I see the 'Edit new version' button now. This should do the trick. I'm not sure how actively we plan to maintain all of these vocabularies, most aren't in active use at this point, but we do need to make a few changes to this particular one.

emiliom commented 1 year ago

I'm glad I saw this after it had already been resolved :smile: !

Brings back memories.

mwengren commented 1 year ago

Thanks all! The https://mmisw.org/ont/ioos/sector vocabulary has now been updated to include the terms indigenous and first_nation.

Reading the vocabulary description more closely, I realized we're going against our own recommendations there (to use gov_state for example to represent Canadian provinces), but I think it's best just to go ahead with the new terms as we decided.

I will leave this open so we can remember (hopefully) to add a vocabulary reference for the _country attributes in the IOOS Metadata Profile as referenced here.

MathewBiddle commented 1 month ago

Note to myself to do some provenance tracking to see where this sector element appears in the following portal metadata records (if it does):

Need to find a dataset that uses this attribute though.

mwengren commented 1 month ago

Hey @MathewBiddle I can give some background about this to save you some time.

AFAIK, the reason the sector attribute appears in the IOOS Metadata Profile is that it was needed so that the netCDF files RAs were creating in the early days to emulate SOS services with the THREDDS ncSOS plugin, xref: https://github.com/ioos/ioos-metadata/issues/35#issuecomment-1152727865.

There isn't any representation of the IOOS sector attribute in ISO XML, that I know of, so you won't see it in data.gov or OneStop, because those are reading our ISO XML files.

However, the IOOS Catalog does pull additional metadata from IMP 1.2 for ERDDAP datasets only (we probably don't have this documented anywhere, but we should - I know where to find it on GitHub, however: https://github.com/ioos/ckanext-ioos-theme/issues/208).

It was added to be able to query by these attributes in the CKAN API and use in other downstream metadata formats, other than ISO, such as Schema.org JSON-LD.

So, essentially, the metadata is there at the IOOS Catalog level, but for ERDDAP datasets only, and may or may not be carried further by downstream harvesters if they choose to use it. I don't think we use sector anywhere in the Schema.org JSON-LD, but we do use some of the other IMP 1.2 attributes.

Here's an example dataset from GCOOS (click on the IOOS Metadata 1.2 tab at the bottom) showing various IMP 1.2 attributes, incl sector FWIW: https://data.ioos.us/dataset/fugro-cruise-c16215-line-1443-75-khz-vmadcp1.

MathewBiddle commented 1 month ago

Thanks @mwengren!

I couldn't find Fugro Cruise C16215 Line 1443, 75 kHz VMADCP on data.gov (search): image

But, I did find another record that I could trace through all the harvesters (Fugro Cruise C16215 Line 1442, 75 kHz VMADCP). GCOOS Record:

At the end of the day, I've confirmed @mwengren comments. Now I am wondering if this conversation should be elevated or if we're stuck using the existing framework and can't add additional items to be harvested upstream?

fgayanilo commented 1 month ago

Let me know if we need to adjust our data profile.