Background:  "ORCID update", December 2011:

Warming ourselves to the subject, here are some excerpts from relevant recent conversations:

RL"Bob": We thought that a concrete step would be to initiate mace-dir work around defining an "authorid" and a "published name" as eduPerson attributes. The spectre of us doing this might be enough to bring together some of the potential interested parties including some publishers, the national libraries, university libraries, OCLC, etc, to advise and dissent.  Some of we might want to consider whether any of the former incommon-libraries participants could be interested.

Of course in trying to do this we might find that what needs to be done is something entirely different (eg working with the author profile schemas from Catalyst or something), but as the play says: Sometimes a person has to go a very long distance out of his way to come back a short distance correctly."

KJK re issue of MACE-Dir as venue for ORCID attribute mongering: "I didn't mean to imply that mace-dir is not the right spot to discuss these attributes - I think it is. I meant to suggest that we take the opportunity of strong international interest in this to make sure our mace-dir processes respect the thrills of international discussion - perhaps taping a level set call so others can listen later in the process or at another time, leaning heavily on a wiki for continuous but asynchronous discussions, etc. Be more sensitive of internationals (though I do not want to get up at 3 am for a mace-dir phone call :))"

KeithH reply to KJK: "100% agreement from me.  Let's think about a MACE-Dir 'ORCID attributes 101' podcast-y thing.  I will start a wiki page & share the link." (This page is it).

  • No labels


  1. Here is the IRISC 2011 Report that covers the  IRISC2011 workshop on Identity and Researchers that talks about the ORCID.

    A number of topics were covered and it augments the site and provides more depth.

  2. Geoff Bilder's article Disambiguation without de-duplication touches on several key issues that I presume are relevant to any ORCID/IdM discussion.

    Bilder's article is in response to Thorisson's late 2010 summary report on ORCID core system requirements and therefore sandwiched in between Thorisson's older white paper and more recent IRISC 2011 Report.

    As I understand it, the current ORCID platform is really focused on individual research self-assertions and the ideal of one canonical identifier with individual in full control of which profile attributes are public and ability to edit them. By contrast, the IdM enterprise directory is predicated on organizational assertions and organizational control. ORCID alpha/beta assumes a 1:1 relationship between author and author's identifier. Since ORCID mandates, by principle, that the author is in charge of the record, it would seem difficult to reconcile ORCID with an IdM system where the institution is presumed to be an authoritative source.

    The author disambiguation space is pretty volatile right now, with bunches of competing author ids vying to become the de facto standard.

    At the very least, author id and published name would need to be multiple values because in practice, even in an ideal system, there will be many exceptions to any 1:1 assumption. Or as Eric Hellman quipped, "Authors are not people."

  3. Looks like a good way to get a grounding in the current status and discussions around ORCID will be to "attend" this upcoming free ORCID Outreach Meeting on May 17th:

    One needs to pre-register, but there is no cost and there is an option to attend thru WebEx and/or dial-in.