Print

Print


No. I think, and always have thought, that DOIs are much over-rated. The
only reason that I ever seen given for them is a social one; scientists
think they are cool. 

What I think would be much nicer would be a URI with a defined and
associated set of standards; so, for example, a simple XML schema
accessible by content negotiation or equivalent; some way of actually
getting the data back from the URI, rather than a landing page; add in a
set of tools to test whether implementations are actually correct. This
sort of standard would then be inclusive; anyone could adopt it without
having to become a member of some registration authority.

And all without users having to remember that the URI they see in the
browser, or that google got them to is not the right thing to pass
around; instead you need an unrelated DOI that may or may not be on the
landing page. 

I don't have the task of managing research data, of course. I just want
to use it; so, it's possible I am missing something. No one has told me
what it is yet, though, and I have asked. 

Phil

Robin Rice <[log in to unmask]> writes:
> Am I the only one that questions the appropriateness of DOIs for institutional
> data repositories?
>
> It seems on the whole that these (our) repositories are fairly immature to be
> adopting DOIs (unlike UKDA and NERC, for example).
>
> What is the driver for emerging institutional data repositories to adopt the
> DataCite system as opposed to other (more lightweight?) systems of persistent
> identification? Is it really a top priority for making our repositories
> robust, discoverable, well-used, cited, sustainable, and trusted, with the
> limited resources we have? Our repository has been up since 2009 (using
> handles) and we've never had a depositor ask for DOIs.
>
> Also, speaking of drivers - I see that Web of Knowledge is now listing its
> repositories to be included in the Data Citation Index - I don't think all of
> these repositories use DOIs exclusively for persistent identification.
> http://wokinfo.com//products_tools/multidisciplinary/dci/repositories/
>
> Thomson Reuters certainly don't mention DOIs in their criteria for inclusion
> (though they do talk about persistence and stability) -
> http://science.thomsonreuters.com/now/elqNow/elqRedir.htm?ref=http://wokinfo.com/media/pdf/DCI_selection_essay.pdf
>
> In short guess I'd support a workshop that considered some of these broader
> questions rather than taking one particular answer as a given.
>
> cheers,

-- 
Phillip Lord,                           Phone: +44 (0) 191 222 7827
Lecturer in Bioinformatics,             Email: [log in to unmask]
School of Computing Science,            http://homepages.cs.ncl.ac.uk/phillip.lord
Room 914 Claremont Tower,               skype: russet_apples
Newcastle University,                   msn: [log in to unmask]
NE1 7RU                                 twitter: phillord

########################################################################
To unsubscribe from the JISCMRD list go to
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=JISCMRD