Set up a workflow for adding zbMATH document IDs from MaRDI to Wikidata

[Daniel-Mietchen]

New feature description in words:

MaRDI is more complete in terms of zbMATH document IDs than Wikidata, so the latter would benefit from being synced to the former.
This complements #580 .

TODOS:

• query MaRDI for items with zbMATH document IDs whose Wikidata counterpart does not have zbMATH document ID statements
• convert query results to edits. This query yields corresponding QuickStatements commands, which can be uploaded and run automatically on the Wikidata side. Depending on the scale at which we want to do this, a bot permission might be needed. Perhaps we could also use the existing Research Bot account.
• do some quality checks (e.g. whether the respective zbMATH doc IDs resolve) and log the results

Considered Alternatives:

• The sample query above uses the Wikidata endpoint for the full graph. In the long term, we should probably replace that with the one for the scholarly graph.
• Instead of SPARQL, we might want to use the REST API, which should be faster and not time out so easily.

Checklist for this issue:

• Participants in discussion have been invited as assignees
• All fields of the issue have been filled
• Example fields have been removed
• The main MaRDI project has been assigned to this issue

[physikerwelt]

I was looking into this. This should be set up on the Wikidata end and controllable from there.
For the SPARQL query, I found this entry suspicious:

Q114056795 P894 "arXiv:2006.14963"

The others seem to be fine.
We could develop a tool running to tool-labs based on https://github.com/cthoyt/quickstatements_client that runs the SPARQL query and inserts it into wiki data.
I once created the https://en.wikipedia.org/wiki/User:ZentralBot for such edits, but we can also go for a new username. However, I suggest closing this ticket here and continuing the discussion on phab.