7

I want to make my geophysical dataset available for all humanity for all future. Until now, I'd put the data on our institutes' FTP server and write the link in the academic article I'm publishing. This is not optimal. Links may change, servers may disappear, universities may change names or decide to restructure everything, and if someone reads the article 10 years from now, there's a good chance the data won't be available at the same address. What I need is a permanent identifier for the dataset, like a DOI.

I've found DataCite, but the only DataCite member in my country is the Swedish National Data Service which deals with humanities, social sciences and health sciences. My geophysical dataset does not fit. Is there another alternative for persistently publishing my dataset? Either including hosting, or just a persistent link that redirects to a website, where I could update the redirect should the location of the data change for some reason.

What alternatives do I have for persistently publishing my data?

5
  • related : opendata.stackexchange.com/q/827/263
    – Joe
    Commented Aug 8, 2013 at 20:20
  • @Joe Yes, that's actually how I found DataCite, I probably should have linked the question.
    – gerrit
    Commented Aug 8, 2013 at 20:29
  • Perhaps the World Data Center can help me. I've contacted them for more information.
    – gerrit
    Commented Aug 9, 2013 at 15:16
  • 1
    The World Data Centers are being replaced by the World Data System. I believe they all have to re-apply, and I don't know what the schedule is for all of that. (it might've already happened .. I remember hearing about it a year or two ago)
    – Joe
    Commented Aug 11, 2013 at 13:49
  • Also related : opendata.stackexchange.com/q/768/263
    – Joe
    Commented Aug 11, 2013 at 13:50

4 Answers 4

5

For either users or producers of open data, the Register for Research data Repositories (RE3) can be a useful resource. In their own words:

The goal of re3data.org is to create a global registry of research data repositories. The registry will cover research data repositories from different academic disciplines. re3data.org will present repositories for the permanent storage and access of data sets to researchers, funding bodies, publishers and scholarly institutions. In the course of this mission re3data.org aims to promote a culture of sharing, increased access and better visibility of research data.

Among other information, the registry specifies what repositories provide the ability to identify particular datasets through DOI, URN, or otherwise. It allows to search by discpline. For example, searching exclusively for open access atmospheric science repositories gets 37 results. Limiting it to those that have persistent identifiers reduces the number of results to 7, such as the World Data Center for Climate.

The registry aims at both producers and users of open data.

4
  • I think the 'open access' in this case means that they allow people to use their data ... you want 'open deposit'. (which is what they call it for other types of repositories ... I've seen the concept talked about in cases like Dryad, but I don't know that I've seen that phrase used specifically in dealing w/ data.)
    – Joe
    Commented Aug 11, 2013 at 13:42
  • 1
    @Joe True. A subset of the repositories linked by RE3 permit arbitrary scientists to deposit their data. I contacted the World Data Center for Climate, and there is no fee, but depending on the size it should meet some quality requirements. My initial 60 GB dataset I can deposit if I send an example file, paper draft, and usage documentation, but my total dataset will be 1 TB, and for this one they want proof of significant value to the larger scientific community.
    – gerrit
    Commented Aug 12, 2013 at 10:05
  • I don't know what 'significant value' means in their case, but for our group, it's mostly an issue of how unique the data is (eg, was there other similar data collected by other groups?), and how likely we think the data would be re-used by other groups (does it help to complete the historic record for some observation type? Is it a long-duration study, or short-lived?). We're also starting to look at documentation standards (can't be re-used if it's not documented well).
    – Joe
    Commented Aug 12, 2013 at 12:16
  • 1
    Significant value were my words, their formulation is that, for 1 TB, it depends on the data quality metrics and feedback from scientific community.
    – gerrit
    Commented Aug 12, 2013 at 12:32
3

The University of Minnesota has a pretty good guide on the topic. In case of link rot, I've copied the summary:

Sharing your data and making it open available is easy. You can do any of the following:

If you're attached to a university or similar, and they have an Intitutional Repository, you can talk to them to see if they'd accept data. (not all do, because of size consideratons, and not all can mint DOIs ... it'd be worth telling them about EZID) if they can't.)

If you don't have an IR (instutional repo.) or a discipline repository that will accept it, and you don't want to go the journal publishing route, you still have more options that will take it and assign an ID:

  • Dryad will accept submissions, with a small fee attached. Most of their data is ecological & environmental in nature (they were built to support DataONE.
  • FigShare will take data for free provided you make it available to the public.
  • Zenodo will take data for free, under different licensing, and can be embargoed.
2

Try talking to archive.org. It doesn't seem like an ordinary person can upload into the dataset collection, but they might be able to arrange something; this feels to me like the right place to put it.

2

In addition to re3data, try Databib. Check with the librarian at your institution who is specifically assigned to your department also.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.