Re: Slightly OT: MARC

Patrick,

Thank you, I was having a great day.  The only parallel with Wikipedia I was drawing (although there could be others) was that were one to harness crowds to aid the process of brokering of SIs of various kinds, a critical mass might be obtained relatively early (say, three to five years) while brokering would continue indefinitely (say, a thousand years... ;-)

Personally, I would find a system where I can choose whether or not to believe Jo Bloggs' brokering of two SIs, intentional or not (e.g. from DBpedia and Grovemusic - "they're talking about the same Beethoven"), based on a clout rating system (such as Amazon or eBay) preferable to the current subject wilderness.

Cheers,
Antony

http://www.musicdna.info


  ----- Original Message ----- 
  From: Patrick Durusau 
  To: topicmapmail-Zo64W7twoUFWk0Htik3J/[email protected] 
  Sent: Tuesday, July 20, 2010 11:08 AM
  Subject: Re: [topicmapmail] Slightly OT: MARC


               Sender ALLOWED   [ Remove ]  [ Block ]    details  
       
              Vanquish Anti-Spam Control Panel 
       


   
  Anthony,

  On 7/20/2010 3:46 AM, Antony Pitts wrote:
  > Patrick,
  >
  > Yes, in many if not all cases brokering will continue as new sources are
  > taken into account and history reinterpreted, but the bulk of the "work"
  > could/will be done in a relatively short timespan, I would have 
  > thought - a
  > bit like the time it took for Wikipedia to acquire a critical mass.
  >

  I don't see Wikipedia as an analogy for mapping subject identifications. 
  Taking nothing away from its usefulness, but comparing the level of 
  detail in Wikipedia to Corpus Juris Secundum (an encyclopedia of 
  American Jurisprudence) makes the gap in terms of subjects fairly stark. 
  And each of those subjects can have one or more identifications in at 
  least 50 distinct legal jurisdictions, to say nothing of overlapping 
  regulatory bodies and our much beloved federal government.

  To be sure, at a certain level of coarseness, even the SW looks doable, 
  but do you really want to rely on medical advice based on information 
  coarsened to make it tractable to an automated system or easy completion 
  by volunteers?

  Personally I think a fairly coarse topic map for some areas could be 
  very useful and "acquire a critical mass" as you say, but it should not 
  be mistaken as being a source of final answers. Wikipedia is very 
  useful, hell, I use it nearly every day, but only as a starting point. 
  There is a very broad and deep world beyond Wikipedia.

  Hope you are having a great day!

  Patrick



  > Cheers,
  > Antony
  >
  > [I, too, managed not to post my initial reply]
  >
  > ----- Original Message ----- From: Patrick Durusau
  > To: Antony Pitts
  > Cc: [email protected]
  > Sent: Tuesday, July 20, 2010 2:02 AM
  > Subject: Re: [topicmapmail] Slightly OT: MARC
  >
  >
  > Antony,
  >
  > <<
  > On 7/19/2010 7:36 PM, Antony Pitts wrote:
  > Patrick,
  >
  >> From what I understand from various record companies and 
  >> broadcasters, it's
  > a similar story with ISRC codes - the "unique identifiers" for sound
  > recordings.
  >
  > The thing is, humans can spot the differences fairly easily on a
  > case-by-case basis - hence the power of crowd-editing, whereas 
  > computers can
  > deal with vast catalogues that our minds would have immense trouble
  > remembering and sorting.
  >
  > Perhaps the solution (at least this is the way we're heading with the
  > musicDNA project) is community-brokered PSIs, more-or-less.
  >>>
  >
  > +1!
  >
  > Ultimately the question comes down to: Does some user or group of users
  > recognize two or more representatives as representing the same subject?
  >
  > Whatever our machines generate is a suggestion, nothing more.
  >
  > I am reading Berlo's "The Process of Communication" (1960) and he has a
  > fascinating treatment on why meaning is not found in messages but only in
  > people. I won't try to do justice to his argument in an email post but 
  > will
  > see if I can create a summary for my blog.
  >
  > While "community-brokered PSIs" would be a very good thing, I suspect 
  > that
  > over time, even fairly short times, 15 to 20 years (I studied Ancient 
  > Near
  > Eastern texts before I got into the technology stuff so 15 to 20 years 
  > is an
  > eye blink), that you will be able to observe shifting community 
  > usages. And
  > once the social context changes, we will have the record for the usage 
  > but
  > no one will understand it quite the same way.
  >
  > Hope you are having a great day!
  >
  > Patrick
  >
  >
  > <<
  > Antony
  >
  > http://www.musicdna.info
  >>>
  >
  > ----- Original Message ----- From: Patrick Durusau
  > To: topicmapmail-Zo64W7twoUFWk0Htik3J/[email protected]
  > Sent: Tuesday, July 20, 2010 12:01 AM
  > Subject: Re: [topicmapmail] Slightly OT: MARC
  >
  >
  > Sender ALLOWED  [ Remove ]  [ Block ]   details
  >
  > Vanquish Anti-Spam Control Panel
  >
  >
  >
  >
  > Lars,
  >
  > It isn't that every library "does its own thing."
  >
  > ISBNs are actually assigned by publishers.
  >
  > Each publisher is assigned a block of numbers but management of the
  > numbers in that block is left up to them.
  >
  > Bad idea.
  >
  > Some publishers, often after a decent period of time, simply recycle the
  > numbers. :-(
  >
  > Or even give a first and second edition of a book the same number.
  >
  > Or correct their original mistake with the numbers.
  >
  > (I worked for a consortium that wasn't always real particular about its
  > use of ISBNs. Wasn't my department but stories of earlier mis-deeds
  > circulated.)
  >
  > Or do many of the horrors that have been described.
  >
  > But, as Alex suggested, it isn't completely useless, it just isn't
  > failsafe to rely upon.
  >
  > We have some helpful identifiers for books, but perhaps not completely
  > reliable as a global identifier. How's that?
  >
  > Given the amount of time and energy that has been devoted to the sub-set
  > of subjects that are books, does that give a clue as to what more
  > ambitious projects to give all subjects a single unique identifier are
  > going to wind up?
  >
  > Hope you are having a great evening!
  >
  > Patrick
  >
  > On 7/19/2010 6:40 PM, Lars Heuer wrote:
  >> Hi Alex,
  >>
  >> [ISBN]
  >>
  >>> Have you ever searched for a book by its ISBN only to find books
  >>> completely unrelated? Yeah, same ISBNs can be different books in
  >>>
  >> No, I don't I thought the ISBN is a GUUID? It isn't?
  >>
  >>
  >>> But of course we can use both of these to *help* us create an identity
  >>> for a book. However, book identity is complex when you dig into
  >>>
  >> Well, my question was "Do we have something like a GUUID in MARC?"
  >> Your answer and the answer of Maria seem to deny it.
  >>
  >> I wonder how someone in the US could exchange information about the
  >> book "XY" with DE? This seems to be impossible according to your
  >> respond and to the respond of Maria.
  >>
  >> That's a "bit" strange...
  >>
  >> So, the conclusion is: We don't have any global identifier? Each
  >> library does its own thing?
  >>
  >> I can't believe that....
  >>
  >> Best regards,
  >> Lars
  >>
  >

  -- 
  Patrick Durusau
  patrick-Q/[email protected]
  Chair, V1 - US TAG to JTC 1/SC 34
  Convener, JTC 1/SC 34/WG 3 (Topic Maps)
  Editor, OpenDocument Format TC (OASIS), Project Editor ISO/IEC 26300
  Co-Editor, ISO/IEC 13250-1, 13250-5 (Topic Maps)

  Another Word For It (blog): http://tm.durusau.net
  Homepage: http://www.durusau.net
  Twitter: patrickDurusau
  _______________________________________________
  topicmapmail mailing list
  topicmapmail-Zo64W7twoUFWk0Htik3J/[email protected]
  http://www.infoloom.com/mailman/listinfo/topicmapmail



------------------------------------------------------------------------------



  No virus found in this incoming message.
  Checked by AVG - www.avg.com 
  Version: 9.0.839 / Virus Database: 271.1.1/3016 - Release Date: 07/19/10 19:36:00

_______________________________________________
topicmapmail mailing list
topicmapmail-Zo64W7twoUFWk0Htik3J/[email protected]
http://www.infoloom.com/mailman/listinfo/topicmapmail