Modern Library Data

This blog post, on the list for next week’s readings, actually provides a link to an audio file of Karen Coyle speaking about the future of library data.  I found it quite informative; Coyle is well-spoken and makes some core concepts clear in regards to the future of library data in the Semantic Web.  It’s not too long and is definitely worth a listen.  I’ve listed a few highlights below.

  • Among other things, she discusses how “in a world of linked data, library data tends to be a nice, standardized document that describes the resources of the library.”  She talks about they way we shouldn’t try to turn MaRC records into linked data, because it would be the same mistake we made when moving from card catalogs to machine readable data–old traditions and formats simply won’t work in the new world of linked data.
  • An identifier that changes isn’t a good identifier.  In the card world, we used the display form (ie, Smith, John), but modern libraries look to use URIs (uniform resources identifiers).  A URI is similar to a URL, but it is a “nonsense string” that represents the author and always will, regardless of display form.
  • URLs CAN BE used as identifiers, but don’t have to—kind of blurs the line between identifying and locating.
  • Problems: There is currently no standard way to make a URI for an ISBN.