Question: what is a LivePaper?

[stevewds]

Suppose I want to indicate that a dataset is associated with a publication that is in some stage of production (perhaps a preprint, perhaps accepted to a journal). Would I create a LivePaper object or a ScholarlyArticle?

[lzehl]

I would leave it to @apdavison to respond in more detail but here my understanding:

Preprints should have a DOI and can be associated to a DatasetVersion under relatedPublication by linking to a respective DOI object.

For publications that do not yet have a DOI I'm not a 100% sure if a link under relatedPublication to a ScholarlyArticle is correct (since this would still require that the article is already published). @apdavison what do you think ? do we need a modification / extension here for "submitted"/"accepted" publications?

[apdavison]

I think ScholarlyArticle is fine for preprints that don't have a DOI. You would just leave the "isPartOf" field empty, or identify the preprint server using Periodical. If accepted, then you would set "isPartOf" to a PublicationVolume with "volumeNumber" set to " accepted" or "in press".

(to answer the question in the title, a "live paper" is an interactive, web-based document that embeds research products such as datasets, models, code within it - see https://www.ebrains.eu/data/live-papers/live-papers for an example of how these are used in EBRAINS)

[stevewds]

Thanks.

It seems that the "isPartOf" field is required, so one wouldn't be able to leave it empty, right? I wonder if it make sense to allow "isPartOf" to be optional?

While I'm making suggestions :-) , I hope it's okay to make another. Right now, to specify a ScholarlyArticle, one has to specify a volume or issue object and then the periodical. It seems burdensome even when the information is available, and sometimes the information isn't available for some periodicals that aren't organized in volumes and issues (like some electronic publications, like biorxiv or eLife). What about having a "Periodical" field with volume or issue parameters as optional for the ScholarlyArticle field?

Steve

[lzehl]

@stevewds Thanks a lot for this feedback. I think we will not change the requirements for ScholarlyArticle. But your needs trigger for sure a reevaluation of the metadata model.

For the preprint / "isPartOf" issue: A preprint that does not have a DOI and is not submitted / belongs to any journal / preprint-archive should then probably just be registered as a WebResource (if it is accessible through a URL) or registered entirely differently. @apdavison maybe a specific schema for preprints ?

For the periodical issue: I think volumes and issues are important metadata for ScholarlyArticles. I would not like to introduce short cuts into the metadata model that allow just to skip this information. From a curation point of view I would always request to provide the full information (especially they are typically easy to access through the citation tools / options from the periodicals. However we need to consider all use cases. Even though eLife articles do not have a publication issue, they are associated to a publication volume (as far as I know). For biorxiv you are fully correct as far as I know they do not have issues or volumes. This brings us back to the question if preprints should be handled in separate schemas. Another option a bit different to your approach: we add Periodical as other linked type to the property "isPartOf" in ScholarlyArticle.

@apdavison what do you think?