Better alignment with ENVO

[jannahastings]

Description of the issue

We need to align the OEO better with ENVO. See EnvironmentOntology/envo#1053

Ideas of solution

We should compare their definitions with ours for overlapping terms and note similarities and divergences. At the minimum we need to include cross-references.

Workflow checklist

• I am aware of the workflow for this repository

[akleinau]

I'll make a sheet

[akleinau]

https://docs.google.com/spreadsheets/d/1EJ_c_t1WQhi_hLvAe8RIhUqZ0tdhKpKfce53fwXX44A/edit?usp=sharing
this is a sheet with all maybe related and equivalent terms I could find

[stap-m]

Is this a good issue for a dev meeting?

[jannahastings]

yes, I think so. I could lead the discussion on this point in the dev meeting.

[l-emele]

For documentation: In the OEO Developer Meeting 15 we decided:

• main outcomes: keep our own definition and make connection to ENVO clear (sameAs-relation)
• todo's:
  • revisit and refine i.a. coal, (solar energy), oil and petroleum products definition
  • JH creates mapping until next time

Next discussion on this topic postponed from to OEO Dev Meeting 17.

[sfluegel05]

I added some classes and comments to the mapping. Also, I marked the pairs of classes which are not the same in yellow.

[sfluegel05]

In dev meeting 19 we discussed the following options for annotation properties:

1. rdfs:seeAlso
2. owl:equivalentClass in extra module (oeo-envo-mapping) that imports ENVO and OEO
3. owl:equivalentClass with literals linking to ENVO classes instead of IRIs (no automatic reasoning possible, but doesn't interfere with our class structure)

We chose the third option, because it doesn't require a new module, but is still provides all the necessary information (i.e. the link to ENVO). It is also extensible (e.g. to the second option) if this is later needed.
An example of the third option has been implemented in PR #765

[sfluegel05]

In dev meeting 22, we discussed the remaining unclear cases. We decided to open two new issues for oil and petroleum products and pollution and implement the rest.
For documentation purporses, here is a PDF version of our collaborative sheet: envo-oeo-relations-table-21-07-29.pdf

[cmungall]

Hi @jannahastings et al, just catching up on this. Great to see action on this!

For those ending up at this issue:

• it looks like the (unusual) decision to pun owl:equivalentClass as an annotation property was revoked and the (imprecise and non-standard) IAO:0006011 is now used
• both this issue and the reciprocal issue on the ENVO were closed but the mappings are not particularly transparent

I would like to reopen this and propose

• skos predicates are used following SSSOM conventions
• the mappings are surfaced more on both the ENVO and OEO side
• we have some loose coordinating to avoid duplicating work

Maybe some kind of zoom call between ENVO and OEO developers would be useful?

[cmungall]

It also looks like the mappings provided in OEO are quite incomplete

e.g of

• [] OEO:00000331 ! portion of matter
  • [i] OEO:00000441 ! water
    • [i] OEO:00110001 ! steam
    • [i] OEO:00110000 ! liquid water
      • [i] OEO:00010104 ! water body
        • [i] OEO:00010105 ! ocean
        • [i] OEO:00010093 ! river
      • [i] OEO:00000345 ! pumped water

and

• [] ENVO:01001476 ! body of liquid
  • [i] ENVO:01000685 ! water mass
    • [i] ENVO:00000063 ! water body
      • [i] ENVO:01000617 ! lentic water body
        • [i] ENVO:00001999 ! marine water body
          • [i] ENVO:00000015 ! ocean
      • [i] ENVO:01001319 ! saline water body
        • [i] ENVO:00000015 ! ocean
      • [i] ENVO:01000618 ! lotic water body
        • [i] ENVO:00000029 ! watercourse
          • [i] ENVO:00000023 ! stream
            • [i] ENVO:00000022 ! river

only "river" is mapped

Also, OEO maps from http://openenergy-platform.org/ontology/oeo/OEO_00000441 water to CHEBI "water". However, all OEO terms like water, methane etc actually have better mappings in ENVO under material, e.g.

[stap-m]

• it looks like the (unusual) decision to pun owl:equivalentClass as an annotation property was revoked and the (imprecise and non-standard) IAO:0006011 is now used

Discussion in oeo-dev-79:

• discuss an alternative for IAO:0006011 in OEO:
  • skos:relatedMatch
  • skos:close match
  • skos:exactMatch
• become SSSOM conform: https://mapping-commons.github.io/sssom/about/
• decision: for now replace may be identical to/ IAO:0006011 with skos:closeMatch, import skos:relationMappings hierarchy