[Esip-preserve] [esip-semanticweb] Identifiers for people?

Bruce Barkstrom brbarkstrom at gmail.com
Wed Mar 7 22:26:01 EST 2012


As an additional issue on this subject, we probably need
to identify how people are connected with and disconnected
from organizations.  The ISO 19115 standard has a structure
identified as a "Responsible_Party" (or some name like that).
If you go into it, it includes things like e-mail addresses and
telephone numbers.  That's probably subject to the pleasantries
of various kinds of policy admonitions from CIO's and their
privacy policies.  As I recall, the 19115 metadata on this
subject doesn't allow a start time and an end time for a person
attaching to an organization.  [As an example of the practical
implications, I recently found I was identified as the person to
contact on certain data collections - even though I'd left the
organization more than half a decade ago.]  Still we might
want to find out whether it's necessary to engage in a research
activity to deal with what would seem to be a pretty common
association (many-to-many field relationships in databases aren't
that hard to deal with now are they?).

Bruce B.

On Wed, Mar 7, 2012 at 10:18 PM, Bruce Barkstrom <brbarkstrom at gmail.com> wrote:
> I wonder if we aren't trying to do something that existed a decade
> ago.  When I was head of the LaRC DAAC, we had talked with
> the LaRC technical library staff - and one of them had put together
> an interface that had authors from journal articles, references to
> articles, and if you were interested would produce a link to our
> ordering interface without human intervention.  The expensive part
> of what we'd have had to do was to pay for an abstracting service
> that would extract existing abstracts from articles ($25 k at that point
> in time).  As I recall, it took this library staff member less than three
> days to create a serviceable web interface that landed directly on
> the appropriate point of our ordering interface.
>
> My impression is that this problem was solved by inexpensive
> tools the librarians already knew about and could adapt immediately.
> Why are we trying to reinvent square wheels when there's probably
> an inexpensive circular wheel already available - and one that even
> a decade ago was pretty serviceable?  [I don't know the tool or even
> the name of the librarian - but I don't think all of the references to
> researchers indicate a very serious attempt at due diligence on
> what might be already available.]
>
> Bruce B.
>
> On Wed, Mar 7, 2012 at 5:09 PM, Matt Mayernik <mayernik at ucar.edu> wrote:
>> In library world, this issue is called "authority control". Sky mentioned
>> the Library of Congress name authority IDs; that is the standard tool used
>> by library catalogers (at least in the US) to do name disambiguation. The
>> LoC keeps authority files for authors, which are exactly what Bruce
>> described: a list of possible names for the same person, with one preferred
>> name provided. There is also an international initiative along these lines
>> to create a Virtual International Authority File (VIAF), http://viaf.org/.
>> VIAF is available as RDF, and thus linked-data friendly, etc. The LoC
>> authority files might be as well, I'm not sure.
>>
>> For this debate though, the big issue with both of these is that they are
>> based on book authors. So journal authors who haven't written books wouldn't
>> be in there, and of course the same for the K-12 educators, community
>> college teachers, etc. that Christopher mentioned.
>>
>> I guess the take away is that these library-based "people" vocabs should
>> definitely be part of the discussion, but as another player in the game,
>> they aren't comprehensive in-and-of themselves.
>>
>> Matt
>>
>> ---
>> Matthew Mayernik
>> Research Data Services Specialist
>> NCAR Library
>> National Center for Atmospheric Research (NCAR)
>> Boulder, CO, 80307-3000
>>
>>
>>
>> On 3/7/2012 2:25 PM, Bruce Barkstrom wrote:
>>>
>>> Find somebody who has a copy of AARC2 and see
>>> what librarians have recommended before assuming
>>> that the whole world didn't exist before the Web.
>>>
>>> Bruce B.
>>>
>>> On Wed, Mar 7, 2012 at 4:23 PM, Curt Tilmes<Curt.Tilmes at nasa.gov>  wrote:
>>>>
>>>> On 03/07/2012 01:56 PM, Sky Bristol wrote:
>>>>>
>>>>> USGS has been working on this issue as well for some time. We've
>>>>> done it as part of what we've used "master data management" for in a
>>>>> project called ScienceBase. I just cross-posted a blog I'd written a
>>>>> while back to a public location that provides an overview on this
>>>>> work.
>>>>>
>>>>> https://my.usgs.gov/confluence/x/dQCfCQ
>>>>
>>>>
>>>> Very interesting -- thanks so much for posting that.  A lot of
>>>> interesting work going on there.  (BTW -- Tom Armstrong says Hi!)
>>>>
>>>>
>>>>> One of the biggest challenges we are having in a government world
>>>>> that we still haven't resolved to satisfaction is dealing with
>>>>> Privacy Act concerns. There's all kinds of things that come into
>>>>> play when a government agency starts aggregating information on
>>>>> people, even when many of them work for us. In that respect, doing
>>>>> this under the auspices of DataONE would be a little smoother. :-)
>>>>
>>>>
>>>> Hmm.. I'll probably have to look into that some more..
>>>>
>>>>
>>>>> The blog post doesn't talk about IDs specifically, but we've ended
>>>>> up storing any number of IDs with people records in
>>>>> ScienceBase. We've got Library of Congress name authority IDs, local
>>>>> data system IDs, and a variety of others. We do as much automated
>>>>> disambiguation every time we encounter a new contact as we can and
>>>>> are still working on developing the data librarian practice to fill
>>>>> in the human bits.
>>>>
>>>>
>>>> So anyway, Peter Schoephoester seems like a really popular person at
>>>> USGS, (associated with 346,675 items, wow! busy guy!)
>>>>
>>>> His "Person Profile Page" in sciencebase refers to him simply by
>>>> number (his "Party Id"):
>>>>
>>>>  http://my.usgs.gov/catalog/catalogParty/show/15322
>>>>
>>>> so it looks like you simply mint a new Party Id for each disambiguated
>>>> person/organization you add to the system, then just always use id
>>>> 15322 whenever you want to refer to that person?
>>>>
>>>>
>>>> I'll definitely be taking a closer look at sciencebase..
>>>>
>>>>
>>>> --
>>>> Curt Tilmes, Ph.D.
>>>> U.S. Global Change Research Program
>>>> 1717 Pennsylvania Avenue NW, Suite 250
>>>> Washington, D.C. 20006, USA
>>>>
>>>> +1 202-419-3479 (office)
>>>> +1 443-987-6228 (cell)
>>>> http://globalchange.gov
>>>>
>>>> _______________________________________________
>>>> Esip-preserve mailing list
>>>> Esip-preserve at lists.esipfed.org
>>>> http://www.lists.esipfed.org/mailman/listinfo/esip-preserve
>>>
>>> _______________________________________________
>>> Esip-preserve mailing list
>>> Esip-preserve at lists.esipfed.org
>>> http://www.lists.esipfed.org/mailman/listinfo/esip-preserve
>>
>> _______________________________________________
>> Esip-preserve mailing list
>> Esip-preserve at lists.esipfed.org
>> http://www.lists.esipfed.org/mailman/listinfo/esip-preserve


More information about the Esip-preserve mailing list