Authority Control: Are You Who We Say You Are?

My understanding is that there's no way to prevent someone for creating an ORCID ID for a fake identity as an author or reviewer. If this is correct, are you thinking of any way to actually validate the identity of the person creating the ORCID ID?

(Simeon Warner): There isn't a mechanism to stop creation of an ORCID iD for a fake identity. However, there is a dispute resolution process (see <http://orcid.org/orcid-dispute-procedures>) to deal with cases of misuse. I note that this is the same situation as the current publication world where articles and books are published under alternate identities which may or may not be readily connected to the real person. A key benefit of ORCID in even these situations is re-use of the identity for multiple works, grants, etc. and connections to other systems that might validate or explain that identity.

(BTW, my favorite example a fake identity is "John Rainwater" who never existed but has a fine publication record in mathematics. John's publications were actually written by a number of different people (see <https://en.wikipedia.org/wiki/John_Rainwater>). I note that John is well described in VIAF <http://viaf.org/processed/LC%7Cno2002106847> based on the LC record <http://id.loc.gov/authorities/names/no2002106847>.)

Can institutions capture all the ORCIDS for researchers at an institution?

(SW): Yes, by two means: As a member organization helping to create and link researchers to their ORCID identities, the organization will get validated ORCID identities via linking process (using oauth, see <http://support.orcid.org/knowledgebase/articles/119676-tokens-through-3-legged-oauth-authorization>). The second method is to extract from ORCID data all entries with the institution names or Ringgold identities the represent your organization (they are quite granular, there are about a dozen identities for different parts of Cornell for example).

What happens to the ID when people pass away?

(SW): If an ID owner passes away the ORCID record will persist and the ID stays "out there" in whatever systems have recorded it. The ID can be used to link to the researcher identity in just the same way as when they were alive. The record in the ORCID registry will remain and any existing permissions for updates will too. Thus, if the record holder granted permission to a publisher or institution to update their record then that can be used to add works that are perhaps published posthumously.

Can ORCID be searched in a database?

(SW): ORCID can be searched directly from the web, and from the public and member APIs (see <http://support.orcid.org/knowledgebase/articles/132354-searching-with-the-public-api>). There are also periodic public data dumps that can be loaded into a local database for query and processing (see <http://support.orcid.org/knowledgebase/articles/223698-how-do-i-get-a-public-data-file->).

What can members of ORCID and ISNI do to help promote convergence of the two identifiers? What are the barriers being encountered in this discussion?

(Laura Dawson): Assignment criteria is the main hurdle. To obtain an ORCID, you only need an email address. To obtain an ISNI, you need to be associated with some "work" - that's a precondition for assignment - so essentially, you have to be "published" in some way.

(SW): I think Laura's point is important, and illustrates different applications for ISNI and ORCID. For ORCID to be integrated into workflows there needs to be the ability to get an ORCID before publication. ORCID iDs are not designed to handle the kinds of rights management uses-cases as ISNI ids are, and the scope of ORCID is smaller. However, there is obvious overlap between ORCID and ISNI, and the two organizations continue to work together on things like the ISNI2ORCID wizard.

We use NACO to authorize name headings in the bibliographic records we create. What benefits can you see to include ORCI/ISNI/VIAF identifiers as well as NACO control numbers in bibliographic records?

(LD): My own opinion is that bibliographic records should contain as many identifiers as possible for Linked Data to function to its maximum potential.

Is there a diagram available that defines the relationship amongst ISNI, ORCID, and VIAF?

(SW): I don't think that a realistic, simple and tidy diagram is possible. Perhaps the best starting point is the diagram on slide 35 of the presentation by Micah Altman and Karen Smith-Yoshimura from the Spring 2014 CNI meeting: <http://www.slideshare.net/oclcr/integrating-researcher-identifiers-into-university-and-library-systems>. The presentation is also an excellent survey of the space.