Proposal: do not distinguish creator from contributor

[cmungall]

From #60

The proposal is to not distinguish contributor from creator, specifically, do not use dcterms:creator, only dcterms:contributor in released files.

Arguments for the proposal

@matentzn

dc:contributor should be the default, and, realistically given that ontologies are always a massively collaborative effort, I would even agree to a motion that gets rid of dc:creator altogether.

And from slack:

My argument is this: nearly all term additions are collaborative efforts. People that add a term merely hit the send button, there have been issue discussions, there are PR reviewers. Why create a low-signal difference between some creator and contributor? > You can say contribute a term.
LSCS: I see little value in the differentiation, and a big downside: we cant settle on one simple way to measure contribution. I think its wrong to say creator is somehow more worthy than contributor, given all the amount of work done reviewing definitions etc.

@bpeters42:

would add to it that essentially dc:creator is_a dc:contributor. And the way we have been using 'term editor' is essentially what dc:contributor is. Furthermore, it can be hard / unfair to try to distinguish who is the creator, in so far as sometimes a term gets added to an ontology with placeholder (or empty) definitions etc. by person A, and person B puts in a lot more effort providing those. So I would favor just sticking to dc:contributor by default

Arguments against

(me)

it's a standard, it's a simple clear distinction, it's easily implementable by ontology editing tools, recording as full a historic context of terms as possible is incredibly useful for ontologies over the span of decades. Is it the best method for giving credit for contribution to ontologies? no, but perfection should not be the enemy of the good.

[matentzn]

Ok, you started this :P

it's a standard,

In what way standard? Some ontologies use it one way, others in another.

it's a simple clear distinction, it's easily implementable by ontology editing tools,

Not really, if we we want to easily quantify contribution - we now need to distinguish two things. Its simpler and easier to have one contribution property.

recording as full a historic context of terms as possible is incredibly useful for ontologies over the span of decades.

This is probably an ok argument (provenance) but I doubt the "incredibly", and I still don't see how the person than added the ID and the label in 4 minutes is different from the one reviewing the PR for 20 and perfecting the definition.

Is it the best method for giving credit for contribution to ontologies? no, but perfection should not be the enemy of the good.

Can you clarify this point - your proposal is marginally better for giving credit then the alternative (simplicifaction). The question here is not what is "better" (in the sense of more fine grained and informative) - the question is if more fine grained and informative is marginally better than simple.

If you want to convince me I would like to hear an argument that clarifies to me how the marginal gain in information content justifies the dubbling in complexity.

[matentzn]

BTW, if we introduced quadratic voting in OBO, I would be only weakly in favour of my position. If the whole community can get behind a two-property solution (terms:creator, terms:contributor) I will 100% back this up! I am much more passionate about #119.

[hlapp]

Perhaps it's useful to compare to the attribution roles for R packages?

[matentzn]

Or Contributor Role Ontology in OBO..

My point is though that we should keep it simple - make the contribution model deliberately trivial to be able to get people to implement it more easily.

What I am still missing is the true marginal gain for creating a more powerful attribution model (contrasted with the added complexity).

[cmungall]

What I am still missing is the true marginal gain for creating a more powerful attribution model (contrasted with the added complexity).

I am always in favor of making things simple. We often make simple things very complicated in OBO with very little gain.

However, I am not sure I see the complexity here. We just have one predicate per contributor role type. Individual ontologies can use the subset that is relevant for them. Rolling up for summary statistics purposes is easy.

As to the gain - I don't think it is forgone that it is marginal. Outside the OBO bubble, the way I see the community going is to separating contributor roles.

This is probably an ok argument (provenance) but I doubt the "incredibly", and I still don't see how the person than added the ID and the label in 4 minutes is different from the one reviewing the PR for 20 and perfecting the definition

I am not sure we can capture the duration of contribution effort, but we can extract specific roles like definition contributor and synonym contribution or relationship contributor from axiom annotations

[bpeters42]

I don't think the distinction is at all clear. Is Rebecca the creator of all COB terms, because she was the one typing them? What is creation? But I could be convinced that some ontologies that want to subtype contributions are allowed to do so. As long as we can agree that all of them are subtypes of 'contributor', and it would be valid to display them as such and not make the distinction if you don't want to.

…

On Mon, Dec 5, 2022 at 11:33 AM Chris Mungall ***@***.***> wrote: What I am still missing is the true marginal gain for creating a more powerful attribution model (contrasted with the added complexity). I am always in favor of making things simple. We often make simple things very complicated in OBO with very little gain. However, I am not sure I see the complexity here. We just have one predicate per contributor role type. Individual ontologies can use the subset that is relevant for them. Rolling up for summary statistics purposes is easy. As to the gain - I don't think it is forgone that it is marginal. Outside the OBO bubble, the way I see the community going is to separating contributor roles. This is probably an ok argument (provenance) but I doubt the "incredibly", and I still don't see how the person than added the ID and the label in 4 minutes is different from the one reviewing the PR for 20 and perfecting the definition I am not sure we can capture the duration of contribution effort, but we can extract specific roles like definition contributor and synonym contribution or relationship contributor from axiom annotations — Reply to this email directly, view it on GitHub <#120 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADJX2IUJR2YTUXIH3WWE6BLWLY7Q5ANCNFSM6AAAAAASSOVRAM> . You are receiving this because you were mentioned.Message ID: <information-artifact-ontology/ontology-metadata/issues/120/1338036024@ github.com>

-- Bjoern Peters Professor La Jolla Institute for Immunology 9420 Athena Circle La Jolla, CA 92037, USA Tel: 858/752-6914 Fax: 858/752-6987 http://www.liai.org/pages/faculty-peters

[matentzn]

I think I could get behind @bpeters42 suggestion, although adding a sub-property axiom between two externally defined properties will cause our friends from the logic department to riot..

[cmungall]

I don't think the distinction is at all clear. Is Rebecca the creator of
all COB terms, because she was the one typing them? What is creation?

Yes, she was the creator, and if we had switched on certain Protege defaults, she would have been tagged as such.

Is this a perfect way of capturing the individual intellectual and at-the-coal-face contributions of the many COB contributors? Definitely not. But it is metadata that accurately reflects what actually happened.

Is this a typical situation (one "Protege driver" many people backseat driving)? I don't think so.

I think I could get behind @bpeters42 suggestion, although adding a sub-property axiom between two externally defined properties will cause our friends from the logic department to riot..

I don't think it's a logic problem at all - it's a problem of imposing a meaning on an external vocabulary that goes beyond or even conflicts with what vocabulary says.

I know dcterms is frustratingly loose and vague, but this is the tradeoff with standardization.

I don't really see the need to inject additional axioms between dcterms to satisfy this use case. if we want to do a query where we bundle contributor and creator we do that.

[cmungall]

I think this is a good time to bring this to the broader community.

One final argument against my proposal:

if we want to truly capture accurate contributor role, then the most simple, flexible, and easily interoperable system is:

• use dcterms:contributor universally as the predicate connecting entity to person
• optionally include axiom annotations on this triple that link to a more granular vocabulary for describing precise contributions; we don't need to decide on the predicate for the axiom annotation or the vocabulary here but it could be
  • CREDIT
  • https://obofoundry.org/ontology/cro
  • datacite roles

[matentzn]

I really like your last proposal. Just saying. Even more then my oversimplified one!

[zhengj2007]

I agree that it is hard to distinguish creator from contributor and creator is a kind of contributor. Would prefer to use dcterms:contributor that proposed here.