NASA is using ENVO and other OBO foundry terms for definitions of terms

[rduerr]

Primarily in documents about their United Metadata Model (UMM) which is used in their Common Metadata Repository (CMR).. See https://wiki.earthdata.nasa.gov/download/attachments/49448405/EED2-TP-039_Rev03_UMM-Var.docx?version=1&modificationDate=1597956196634&api=v2

[cmungall]

Great!

Is the action to add this to the website?

I see this on E-54

"MeasurementContextMedium": "atmosphere", "MeasurementContextMediumURI": "http://www.ontobee.org/ontology/ENVO?iri=http://purl.obolibrary.org/obo/ENVO_01000267", "MeasurementObject": "cloud", "MeasurementObjectURI": "http://www.ontobee.org/ontology/ENVO?iri=http://purl.obolibrary.org/obo/ENVO_01000760", "MeasurementQuantities/Value": "albedo ", "MeasurementQuantitiesMeasurementQuantityURI": "http://www.ontobee.org/ontology/ENVO?iri=http://purl.obolibrary.org/obo/PO_0009087".

They should just use the PURLs, e.g. http://purl.obolibrary.org/obo/ENVO_01000760

[dr-shorthair]

They should just use the PURLs

I see this pattern all the time.

ChEBI references are particularly bad (and the ChEBI HTML pages don't help, as you can't find the PURL anywhere.) Or am I mistaken in thinking that http://purl.obolibrary.org/obo/CHEBI_29351 is the proper PURL for Nitrogen?

[cmungall]

This is the correct PURL, many resources are not used to thinking in semantic web terms

[rduerr]

OK, so I submitted a comment on their system....

[rrovetto]

This is concerning for a few reasons... 1) There appears to be a misreading. Based on the document, it looks like it is an example, not a use of those terms. (If correct, I'd recommend renaming the title of the issue so as to prevent misinformation) 2) More importantly, it should not use terms for a few reasons, some of which are ethical.

• it would be biasing and showing preference for a particular product and group.
• Envo currently uses other ontologies, and those should not be forced or imposed on others projects. -- if using a formal logical definition, it also use the more-abstract connected terms from other ontologies, which may not be the intention, desired, etc.
• in doing so it would disadvantage other similar, perhaps competing, ontologies in the global community, as well as other users.
• in doing so it would contribute to unfair advantage or monopolization by those ontologies.

3) Gov't agencies are to aim for neutrality and fair practice. We should be endeavoring for neutrality and fair practices, and not favor a particular group or product.

[kaiiam]

@rrovetto

There appears to be a misreading. Based on the document, it looks like it is an example, not a use of those terms.

"MeasurementIdentifiers": [
      {
        "MeasurementContextMedium": "Atmosphere",
        "MeasurementContextMediumURI": "http://purl.obolibrary.org/obo/ENVO_01000810",
        "MeasurementObject": "Brightness",
        "MeasurementQuantities": [
            {"Value": "Temperature", "MeasurementQuantityURI": "http://www.ontobee.org/ontology/PATO?iri=http://purl.obolibrary.org/obo/PATO_0000146"}
         ]

It appears to be both recommending and using OBO terms.

More importantly, it should not use envo terms for a few reasons, some of which are ethical. it would be biasing and showing preference for a particular group.

and

Gov't agencies are to aim for neutrality and fair practice. We should be endeavoring for neutrality and fair practices, and not favor a particular group or product.

I don't understand this logic, if they were to use any vocab wouldn't doing so be biased toward that vocab? Therefore they should shouldn't use any vocab as not to be biased? That would lead to more unconnected silos no? Better to use something, ideally better yet multiple vocabs if existing.

in doing so it would disadvantage other similar ontologies in the global community, as well as other users. in doing so it would contribute to monopolization by those ontologies.

Please feel free to recommend other appropriate ontology projects to the relevant agencies/bodies, I think in the end of the day they probably just want something that works. My understanding is that OBO is trying to do it's best to provide comprehensive representation of many domains as possibly. I don't think that's the same as trying to monopolize everything. AFAIK (which is why I spend my time contributing to them) OBO ontologies are completely open source and open access, so anyone should be able to contribute as desired.

[rrovetto]

@kaiiam I believe that is incorrect. In general, ignoring or not caring about assumptions and commitments made by an ontology (or any system) could allow improper things to happen, such as non-consensual (or other under the radar) use of an imported ontology, monopolizing effects thereof, etc. That group is for bio, for which there are many semantic products. Trying to cover as many domains, but while having those restrictions, assumptions and commitments to other products/ontologies that would thereby exclude use of similar or competing products, would be unfair advantage for those and thereby on the line of monopolization. obo has not moved away from its upper ontology, and upper ontologies used in those ways can pose such risks, and my understanding is that Cob project is separate from obo, not obo-directed (is that right?). Happy to speak offline. In general, any vested interests in those or related projects/products/ontologies should not cause anyone to push them or aim to have them dominate. Esip should not be platform to have that happen.

[kaiiam]

@rduerr was just pointing out that this particular working document/working group at NASA was already using ENVO. We (the envo team) didn't even know about this let alone push or ask them to use ENVO/PATO.

Like I said @rrovetto I'm certainly not trying to ignore distinctions/ontological commitments, I'm all ears to suggestions about other vocabs which don't have the issues you raise. If there are some other vocabularies you think that NASA group should be considering reach out to them and suggest it to them.