sector division subclasses

[stap-m]

Description of the issue

The class sector division has only one subclass, which is common reporting format. However, in the list of sector individuals, there are more divisions refered to, i.e. EU emission sector division, KSG sector division, NC/BR sector division. They also rather be subclasses of sector division.

Ideas of solution

Make EU emission sector division, KSG sector division and NC/BR sector division subclasses of sector division.

Workflow checklist

• [ ] I discussed the issue with someone else than me before working on a solution
• [ ] I already read the latest version of the workflow for this repository
• [ ] The goal of this ontology is clear to me

I am aware that

• [ ] every entry in the ontology should have a definition
• [ ] classes should arise from concepts rather than from words

[l-emele]

Why do you think, these should be classes now?

[stap-m]

common reporting format is a class, that's why I think the other sector divisions should be, too. Is there a reason why they are not?

[stap-m]

And the concrete application: for the OEKG, we pick the sector divisions from the OEO and they should be found in the same place.

[l-emele]

I see your point. After thinking a bit, I am okay, with having a few more subclasses of section division. However, I would not turn all your proposed individuals in classes, but instead have the following classes:

• In #1400, I already proposed a class energy balance sector division: An energy balance sector division is a sector division that is defined by an energy balance calculation method.
• CRF-based sector division: A crf-based sector division is a sector division that is derived from a common reporting format. This the can collect KSG sector division and NC/BR sector division.
  • Having NC/BR sector division as a class is in my view okay as this sector division from multiple countries and each my have its slight modifications.
  • However, I think KSG sector division should stay an individual: there is one and only the one sector division that is part of the KSG https://www.gesetze-im-internet.de/englisch_ksg/englisch_ksg.html#p0128 and there are no other KSG sector divisions.
  • CRF-based sector division should be implemented as: 'CRF-based sector division' EquivalentTo: 'sector division' and uses some 'common reporting format' to allow multihierachy.
• EU legislation sector division: An EU legislation sector division is a sector division that is used in EU legislation. @han-f: Is this label fine or any suggestion for a better label?
• Make EU emission sector division a subclass of EU legislation sector division. I think it is fine to have here a class, as there are multiple EU emission sector divisions. E.g. the scope of the EU ETS and ESD/ESR sectors changes over time as legislation changed.
  • GovReg sector division should then be an individual of EU emission sector division. As it has the axiom uses 'CRF sectors (IPCC 2006)' it will be also inferred as individual of CRF-based sector division.
  • Renewable Energy Directive sector division should then be an individual of EU legislation sector division.

In general: In the OEKG it should be possible not only to reference classes but also instances of classes.

[stap-m]

Aha thanks, now it makes sense, that some of them are individuals 👀 I like your proposal.

In general: In the OEKG it should be possible not only to reference classes but also instances of classes.

Of course. In fact, this already works for the sectors and should thus also work for the sector divisions.

[l-emele]

So is this ready for implementation?

[areleu]

Is there any particular reason on why the EU is hardcoded in the proposed classes? Why not add a sub-class that has the possibility to have a jurisdiction associated as the form of a Relation. Otherwise, every time we have new sector divisions associated to different govermental entities we would need to add them.

This happens too with the EU allowance class (the instances can still have EU in their definition, is just having it in the classes seems non sustainable to me)

[l-emele]

What do you mean with hardcoded? Of course, the EU should be part of the definition of these classes as it is an existential part. Further, we cannot axiomatise the relation yet the relation to the EU as we did not yet solved how to include geographical entities from an external geo-ontology.

[areleu]

Is the EU as the emitter of legislation a geographical entity?

[areleu]

I would say we pick this up in the next dev meeting. I have a couple of questions that I think is faster to explain if we are talking directly.

Edit: In short I would argue that CRF is an instance of sector division, not a subclass of it. But I guess there is practial reasons to have it the other way and that I would like to know.

[l-emele]

Can you please at least formulate the question in before as preparation for the meeting?

[areleu]

1. Is there a practical reason on why CRF sector divisionwas considered a class in the first place?
2. Is it sustainable to make a class for sector divisions emitted by different legal powers? (let's say for example that someone wants to include divisions from American Jurisdictions) PS: I just noticed the EPA uses CRF
3. Aren't we missing key entities in the social module like legal organisation? (maybe organisation already covers this)

[l-emele]

At least the first question is easy to answer: The practical reason, is that there are at least two instances of CRF sector divisions, one based on the Revised 1996 IPCC Guidelines for National Greenhouse Gas Inventories and a second based on the 2006 IPCC Guidelines for National Greenhouse Gas Inventories, see the respective individuals of that class.

I am happy to discuss the other two discussions on the next oeo dev meeting and will them on the agenda.

[areleu]

I see that there are 2 CRF sector divisions. And they have three things in common:

• The name
• The fact that they were agreed and emitted by the IPCC
• The format in which they are published

I think the first can't be a definitive differentiation from the parent class for obvious reasons.

The second can't be a definitive differentiation because we would imply that there is a precedent to make a different subclass for every emitting organisation.

The third could be a definitive differentiation but I find it rather weak. What if the IPCC decides to publish the reporting divisions as an Ontology (I wish!) and not as a table of concepts. This would make the "new" sector division non relatable to the other two.

[areleu]

In short I think when implementing https://github.com/OpenEnergyPlatform/ontology/pull/1463 we were going in the right direction for subclasses of sector division as it specifies further on what is the method/reason of producing such division, namely an energy balance. If we were to do something analogous but for the CRF family of sector divisions I think we need to look at the porpouse the IPCC recomments the CRF.

My understanding is that it is used primarily for emission accounting. So the natural approach would be creating the subclass of sector division emission accounting sector division.

This leaves still open a question of how to relate GovReg with the EU. And I think the solution would be, as I mentioned earlier, to add a relation to sector division that points to the organisation behind it (information that I find anyways essential).

[l-emele]

Following up on the discussion at the last oeo dev meeting I opened this issue

• 1489

As most geo-related topics in the oeo are still in its infancy, we probably need a while to implement something like 'German energy balance sector division' 'is defined by' Germany.

But I think, we should at least now implement an axiom for the parent class sector division. Something like: 'sector division' 'is define by' some (organisation or person) Can we agree on this?

[areleu]

@l-emele should we close this?

[l-emele]

The current state of sector division is:

I think, we can safely close this issue.