Open fabiobatalha opened 5 years ago
For SciELO it would be useful to relate supplementary material (datasets, spreadsheet, etc), as well as, to publish reviewers report related to an article.
Hi @solange-scielo
as well as, to publish reviewers report related to an article.
Have you seen this paper: https://f1000research.com/articles/7-1655/v1 It has a proposal for tagging for reviewers report related to an article and it would be great if everyone could follow this model. Support for this is being factored into Texture as it's an eLife requirement. (sub articles).
I hope this proposal will become a JATS4R working group soon to iron out some of the issues pointed out by reviewers of the paper and other comments.
relate supplementary material (datasets, spreadsheet, etc)
At eLife we don't use related-object for this, can you share an XML sample?
We've seen related-object being used very similarly to a regular reference. So we don't really see the conceptual difference except for display purposes. It would be good if we could consolidate the use of two concepts into one. We can come up with solutions for the special display requirements.
@Melissa37 @fabiobatalha @solange-scielo what would prevent you from using a reference instead ob related object currently? Thanks for the help!
@FAtherden-eLife is going to search our archive for any use of related-object and we'll add here.
@michael
At first, what made us to take this decision and stick in relate-object to identify a fully reviewed document is like you said for display purpose and also because ÉruditArticle Schema supports this kind of metadata that is identified by a specific element (trefbiblio). The purpose of trefbiblio in ÉruditArticle is similar with the related-object in JATS.
But later, seeing examples in Érudit (https://www.erudit.org/fr/revues/cgq/2013-v57-n161-cgq01387/1024906ar/) we assured it is a practice in the HSS journals for review articles, conference review or report documents.
As you can see in the above example (check also the pdf file). This document does not have the usual title, the title is the reference to the document the report was made for, this document also does not have citations once it is about a report.
In my opinion there is a slightly difference in this kind of "citation" and to correctly identify it in JATS we can not classify it as an ordinary citation once it will misrepresent this metadata.
I would like to see the opinion of the librarians in this topic.
We've seen related-object being used very similarly to a regular reference.
When you use reference tagging, this becomes a citation and affects citation metrics for systems using the XML to pull citations. I am thinking they would not pick up a related object in this way. If a system is pulling references into a reference list then this will be pulled in. There is common usage for related-object in the publishing world.
As we have a real use case provided and it could limit Texture's future use by others, I would support separating this from citations and supporting it
Let's work on a clear specification for the related-object use-case first. Could we adapt this issue to conform to our feature-request template and have user stories and JATS examples provided? I'd like to reach exactly one tagging style, with optional fields. Can we assume there's structured data and related-objects can then be rendered (like a citation style)? Or would we choose to rather have those references unstructured (meaning you enter free form text, that's what Fabio requested).
@JGilbert-eLife @FAtherden-eLife can you look at this and think of user stories and ask any questions you think necessary?
@fabiobatalha - please could you provide some XML examples of your use of related-object
? I think I see what you need here, but it would be useful to look at some real cases before drafting a full set of user stories. Thanks!
User stories
eLife:
Erudit:
XML requirements
Provide samples!