Metadata requirements

[AgnesBelt]

List of requirements:

1. An output classification
2. Licensing info
3. Authors (identified by their ORCiD)
4. Author contribution (e.g. CRediT author statement), plus any important specifics (e.g. Jo Blogs may be author no 5, but wrote section 2 which forms part of his PhD. Stating that clearly will be helpful for him later when justifying its use in his thesis).
5. Geographical coverage (country / region)
6. 2/3 sentences description that an FCDO country officer (NOT TECHNICAL) will understand.
7. Links to dependent/subsequent/associated work
8. Association with which WS (more than one?)

[AgnesBelt]

Pending issues left to be addressed:

• What unique identifier should be used for:
  • Web page (URL?)
  • Social media drop (URL?)
  • A picture (e.g. following what needed to be done for Open Learn?)
  • An 'activity' - i.e. summer school / webinar etc.

[AgnesBelt]

Suggestion: Should we create a metadata form to be filled in for each output type?
See:

• How to add custom google forms to your static website
• Google form customization for static websites

[FraGard]

On the open issues, I'd suggest:

• Web page: URL + access date (like when referring websites in papers; date is important, since content changes);
• Social media drop: here, I'd imagine URL is enough
• Picture: for the courses in open learn, we had (e.g. on the ppt): 1) author/source, 2) link, 3) license (with link to license)
• Activity: URL of the event page

[FraGard]

On the list of metadata:

• I'd add also acknowledgments & funding sources
• points 2,3,4,5,6 are standard info for published papers (where 6 is the abstract + highlights) and we can make them standard for any upload to ResearchSquare (for written content), Zenodo (for datasets), GitHub (for code), etc. By creating a metadata form like suggested by @AgnesBelt or simply listing in the guidelines what needs to be there
• points 1,7,8 are CCG-specific. These would probably have to sit in our CCG research index somehow?

[willu47]

I believe most of these can be satisfied with the metadata representation allowed by Zenodo, even 5. Author contribution is partly covered by the contributor field, although the vocabulary doesn’t match 1:1. 8. is probably trickier to enforce (and could be done in Zenodo through establishing a separate deposit to represent a workstream, and then using related_identifiers to group collected works.) 7. is also covered by the related_identifiers field which allows a relation to be defined such as isReferencedBy or isDerivedFrom or reviews.

[willu47]

One perspective on "web archiving"

[willu47]

Another on social media archiving