Feature 109 funding

[mbjones]

These are the proposed changes for Funding to use the @reverse keyword on fundedItem. It includes both updates to the Dataset guidance docs, a new example, a new figure, and an ADR describing the impact.

Discussion of these features was in issue #109

[mbjones]

@datadavev can you give this a quick review so I can merge it into develop if there is nothing controversial? I think you already looked over the changes based on your comments in issue #109 . Thanks.

[datadavev]

Overall this seems great to me as a mechanism to describe funding sources of a dataset. There are a couple minor typos, and a suggestion to add a reference to the spec for @reverse. Otherwise this is good to go.

The only issues I have are in the broader context of assuming that there's some special characteristic of the JSON structure of the JSON-LD block that confers meaning, when in fact, there is none. The logical representation of the content using @reverse or @graph with appropriate targets is the same, and the RDF representation of each is identical. Hence, my impression is that emphasis should be more on the correct representation of the relations between entities rather than promoting one particular construct (i.e. @reverse) when there is no logical difference.

[mbjones]

Yeah, makes sense. How about if I add a sentence at the end explaining that and alternative would be to use the @graph representation?

[datadavev]

Yeah, makes sense. How about if I add a sentence at the end explaining that and alternative would be to use the @graph representation?

Kind of feels more like there should be a general guideline on JSON-LD, but adding a sentence here could be helpful. Perhaps something like:

Note that using @reverse is a convenience that can simplify the construction of JSON-LD for describing datasets since it places the funders as objects within the SO:Dataset object rather than composing multiple objects within a @graph list which may be more complicated to generate. Logically there is no difference. In either case, consumers of JSON-LD should take appropriate steps to process the content to achieve an expected JSON structure (e.g. with framing) or treat the JSON-LD as an RDF graph and query accordingly.

[mbjones]

Thanks @datadavev I incorporated your feedback into the branch and PR.

[smrgeoinfo]

In the first paragraph on Funding, the Crossref Funder ID is (I think...) referred to with three different names.

...institutional identifier such as a ROR, or ISNI identifier, or a more detailed Crossref Funder ID. The ROR for the National Scien....ifiers as well. The Funder ID has the advan.... When possible, providing both a ROR and FunderId is helpful. ...

It would be much clearer if the same name was used, maybe 'Funder ID' and link to Crossref (https://www.crossref.org/services/funder-registry/) the first time its used?

[mbjones]

Thanks, @smrgeoinfo, good catch. I just harmonized the language used to "Funder ID" and added a link in sha e87b2e0. Does that address your feedback?

[smrgeoinfo]

looks good