DCP Session minutes (thanks to Tim)

Françoise, Trustworthy Data Repositories

• Certification of data centers.
• CDS got certified.
• Useful because of the feedback from experts.
• CoreTrustSeal preferred (ISO too expensive)
• Strongly feel that certification is really important.

André, Tim: DOI Note status

• ADASS BoF had follow up session on DOIs.
• Had meetings with AAS, Planetary Data and Chandra on DOIs.
• Sent round survey asking about DOI usage.
• Aim to release draft at Groningen meeting.
• DOI best practices in astronomy.

Q: DOIs are primitive, aren’t we really talking about data publication best practices?

A: Survey should determine what we are covering.

Raffaele: Chandra

• Chandra migrating to persistent identifiers.
• All papers using Chandra data should be tied to the original datasets.
• Metrics then used to demonstrate that the archive is being used.
• Were always required to use persistent identifiers but now switching to Datacite DOIs.
• Want to know how much data is used for different types of publications.
• Continually have to follow up on authors to make sure they include identifiers.
• Can track evolution of archive usage.
• Can also try to track when multiple datasets from different missions are combined.
• Collection pages link paper to multiple datasets.
• Detailed metadata plan.
• Using DataCite metadata schema 4.1 but still needs work for astronomy.
• Need to mint 40,000 DOIs to backfill archive.
• Why does obscore mandata IVO identifiers and doesn’t allow use of DOI.

Séverin: CADC Self-serve DOIs

• Wanting to publish data that they have generated for their paper.
• Cross-linking data to a paper.
• Initially minting DOIs manually (on request) since 2013.
• Self-serve since March 2019.
• Publishing of VOSpace areas.
• Files are locked once minted.
• System set up to allow paper reviewers to see the data before minting.
• Built on top of VO services.

Q: Are there constraints on what can be stored for minting?

A: No. Whatever author wants to put in there to be associated with the paper.

Q: How do you track raw data holdings usages?

A: CADC aren’t required to do so. Telescopes themselves care more about data usage.

Q: Where is the code?

A: Should be open source. Will make sure it’s available.

Yihan : PaperData Updates

• China-VO systems for uploading data use for journal publications to get DOIs.
• Issuing IVOID and DOI.
• Requires that the paper has been accepted before a DOI is issued and the metadata is complete and true.
• Once DOI issued everything is locked.
• Versioned DOIs are supported.
• Compliant with AAS Journals rules.

Q: Can a DOI be reserved before acceptance of paper?

A: Not in plan at the moment.

Deborah : ESA Science Archives DOI plans

• ESA can issue DOIs using 10.5270
• Already issue DOIs for missions and space craft for Earth observation.
• Really want to make it easy to link data to papers.
• ESA DOIs refer to current best calibrated version of data. No new DOI if calibration improves.
• DOIs deliberately opaque so that people can’t guess what they mean
• Not even sequential.
• Formal Data releases do get DOIs.
• Each proposal collection gets a DOI — not individual observations.
• Proposing to mint DOIs from selecting observations from archive (as MAST do already).
• DOIs for high level science products uploaded back to ESA.

Q: I don’t understand the single DOI even if recalibrated. This seems to go against the definition of DOI.

A: We don’t have the older calibrations of Herschel for example.

Q: Why not deprecate the original DOI pointing to a new DOI? There is no way to reproduce a paper if the calibrated data have changed.

David: RDA as relevant to DCP

• RDA is a Multi-disciplinary organization on data management infrastructure.
• Helpful to monitor where RDA is going.
• Usually very high level abstractions of topics.
• RDA is a place where you could look for relevant outputs that represent useful thinking on your subject.
• Data citation of evolving data.
• Prefer to archive the query rather than the data itself. The query must be able to reflect the date that the query is relevant for.
• This is not always possible in astronomy.

Q: Zwolf has implemented a version of the evolving data citation system.

Q: In planetary science you only get DOI to full dataset and can’t specify a subset for your paper.

Carlo Maria: RDA Federation Identity Management & usage in VAMDC

• Aim is to remove obstacles to research by dealing with federated identity.
• Try to track an individual even if they move institutions.
• VAMDC integrated with Scholix and Zenodo.

Please respond to the DOI survey.