 |
 |
If there is really an appetite to continue DAIs going forward, the wikipedia support for identifiers is modula and there's no reason not to add more identifiers. cheers stuart On 06/05/2014 11:06 PM, Ben Companjen wrote: > Hi, > > Of course there are more identifier systems (or domains, if you will). > > Most/many authors in The Netherlands have a Digital Author Identifier > (DAI), which is the record number in the GGC (Gemeenschappelijk > Geautomatiseerd Catalogiseersysteem), or Shared Automated Catalogue > system. > The DAIs are assigned by (university) libraries and in the case of > university libraries assigning/finding DAIs for their researchers, the DAI > is usually linked to the employee in the repository. Following > "EduStandaard" agreements [0] among all Dutch universities and some > service providers like my employer DANS and the National Library of the > Netherlands (KB), we can harvest the IRs and link publications to > researcher profiles and show them in NARCIS [1]. > > [0]: > http://www.edustandaard.nl/afspraken%20en%20architectuur/beheerde-afspraken > / > [1]: http://www.narcis.nl > > Setup as a service by a company called Pica, the GGC is now hosted by OCLC > after Pica merged into OCLC [2]. The authority files for authors together > are called the NTA ([Dutch Thesaurus Author names]). > > [2]: http://www.oclc.org/nl-NL/ggc.html > > OCLC is also hosting the ISNI database and VIAF (of course). VIAF, as you > know, was setup as a crosswalk of authority files (including the NTA). > OCLC are working on crosswalking identifiers, AFAIK. > > Please be aware that ISNI is a /name/ identifier. Pseudonyms and birth > names for the same person (should) get different ISNIs. And, as said > before, not only people can get ISNIs. Also, the business models for ORCID > and ISNI are different. > > > As a Linked Data aside, Eric, be aware of what an identifier identifies - > and then how you make assertions using them. > For example, ORCID doesn't use the hash or 303 pattern, so if you resolve > http://orcid.org/0000-0002-9952-7800 you get a webpage, i.e. > http://orcid.org/0000-0002-9952-7800 identifies a webpage (the same goes > for DOIs, btw). That is why I say about myself (in Turtle): > > <http://companjen.name/id/BC> dct:identifier > "http://orcid.org/0000-0002-7023-9047" . > > instead of > > <http://companjen.name/id/BC> owl:sameAs > <http://orcid.org/0000-0002-7023-9047> . > > > … for I am not a website. > > Linking me to things I make is done like so (Qualified DC): > > <#thing> dct:creator <http://companjen.name/id/BC> . > > > In your example you used the identifiers as names for the creator(s); it > is as meaningful as saying (in unqualified/simple DC): > > <#thing> dc:creator "Eric Lease Morgan" . > > Hope this helps :) > > Groeten van Ben > > > On 05-06-14 00:14, "Stuart Yeates" <[log in to unmask]> wrote: > >> Others have made excellent contributions to this thread, which I won't >> repeat, but I feel it's worth asking the question: >> >> Who is systematically cross walking these identifiers? >> >> The only party I'm aware of doing this in a large-scale fashion is >> Wikipedia, via https://en.wikipedia.org/wiki/Template:Authority_control >> >> cheers >> stuart >> >> >> On 06/05/2014 06:34 AM, Eric Lease Morgan wrote: >>> ORDID and ResearcherID and Scopus, oh my! >>> >>> It is just me, or are there an increasing number of unique identifiers >>> popping up in Library Land? A person can now be identified with any one >>> of a number of URIs such as: >>> >>> * ORCID - http://orcid.org/0000-0002-9952-7800 >>> * ResearcherID - http://www.researcherid.com/rid/F-2062-2014 >>> * Scopus - >>> http://www.scopus.com/authid/detail.url?authorId=25944695600 >>> * VIAF - http://viaf.org/viaf/26290254 >>> * LC - http://id.loc.gov/authorities/names/n94036700 >>> * ISNI - http://isni.org/isni/0000000035290715 >>> >>> At least these identifiers are (for the most part) “cool”. >>> >>> I have a new-to-me hammer, and these identifiers can play a nice role >>> in linked data. For example: >>> >>> @prefix dc: <http://purl.org/dc/elements/1.1/> . >>> <http://dx.doi.org/10.1108/07378831211213201> dc:creator >>> "http://orcid.org/0000-0002-9952-7800" , >>> "http://id.loc.gov/authorities/names/n94036700" , >>> "http://isni.org/isni/0000000035290715" , >>> "http://viaf.org/viaf/26290254" . >>> >>> How have any of y’all used theses sorts of identifiers, and what >>> problems do you think you will be able to solve by doing so? For >>> example, I know of a couple of instances where these sort of identifiers >>> are being put into MARC records. >>> >>> — >>> Eric Morgan >>> >