[open-bibliography] ORCID SURVEY

Mon Oct 11 14:24:39 UTC 2010

Thomas Krichel <krichel at openlib.org> wrote:

> > Possibly if its open data, open API. 
>
>   For it to be open data, you need to be able to get a copy
>   of the data. Then an API does not matter. 

Not true. Users of the data do not want to deal with a file containing 100M or more authorship events.
They want to have search and browse facilities over the data. If ORCID does not provide an API, then some
other agent or agents will have to. And what is the business  model to support the activity of those agents for
the benefit of the community? ORCID will be perfectly positioned to support an API, and if well constituted policitally it 
should be able to provide a good open one. That would be by far the best outcome. If we dont push for it, we wont get it.

> >  If its a closed membership club like Xref, count on me and others in the open biblio community 
> > http://wiki.okfn.org/wg/bibliography
> > to create and maintain a competing and I expect ultimately more successful distributed system for purposes of academic research, based on open data principles.

>   I am already maintaining AuthorClaim as a free system for authors claiming papers. Close to 100,000,000 authorships can be claimed there. 

Right. This is a foundation of what the open biblio community  could support, but we need much more than that: there needs to be an incentive
for authors to participate. Possibly if these 100,000,000 authorships were embedded in ORCID data and accessible via API, then
the incentive would be there for authors to curate their bibliographies by claiming items and upgrading the data in those items to a point
where it was useful to them e.g. in compiling CVs, displaying their data on the web, etc.

> > On the question of what data I think ORCID should keep. As long as it is open, then as much as possible, subject only to privacy and copyright laws

>   In AuthorClaim, that is only the password of a user for the system, and the email address. The author may decide to make her email public, but private 
> is the default setting.

It is a vexing issue who or what agent is responsible for security of a massive system of user/passwds and email addresses. I think this
should not be entrusted to a central agent. Rather we should leverage the existing system of trusted agents for holding user/passwd info, let this be a 
distributed system, one node for each ORCID participant, and then let these particiants exchange and match data through public identity assertions, each 
with an owner  (Agent A asserts person A:1234 is sameAs B:3456 ). Let the chips fall where they may with regard to the accuracy of the assertions. 
It should be fairly easy for interested agents to detect inconsistencies and dubious assertions, and suggest corrections.  
This may not be perfect, neither is democracy, but its better than a dictatorship.
--Jim

----------------------------------------------
Jim Pitman
Director, Bibliographic Knowledge Network Project
http://www.bibkn.org/

Professor of Statistics and Mathematics
University of California
367 Evans Hall # 3860
Berkeley, CA 94720-3860

ph: 510-642-9970  fax: 510-642-7892
e-mail: pitman at stat.berkeley.edu
URL: http://www.stat.berkeley.edu/users/pitman