Repositories for datasets

[tkuhn]

Should we give some more specific guidelines of where authors should/have to deposit their datasets? Zenodo, figshare, datahub? Do they need to be archival in the sense that authors cannot later delete the content (unlike GitHub)?

[micheldumontier]

That's a good suggestion. They should also investigate institutional libraries.

[LEHunter]

And, perhaps to be repetitive, it might be good to align with the NIH Data Science Commons Digital Object Compliance statements: https://datascience.nih.gov/commons

[tkuhn]

Probably a good starting point: http://journals.plos.org/plosone/s/data-availability

The recommended repositories section is particularly helpful: http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories

[kverspoor]

The recommended repositories there look very good. Key is getting a DOI I think.

[tkuhn]

We can borrow the PLOS guidelines for now, and maybe come up with our own version in the future. I included that in the guidelines: http://datasciencehub.net/guidelines.html