Finding and Using Moving Images In Context

So here we are looking down the barrel of PBCore as our new metadata schema. On July 2 we are contemplating the clip ID that appears under each thumbnail. What’s it to be? A clip title makes sense, and so does a number. In the PBCore schema that would be the element 25.25.1 formatIdentifier, which is in the container pbcoreFormatID.

A very helpful PBCore diagram.
Look to the lower right.

Discussion ensues about numbering among the staff.
David Weiss: Our numbers currently look like this
1178.0043.001
1178 is the accession number
0043 is the item in the accession (reel or tape or story on tape)
001 was usually a new preservation element like negative or answer print

I don’t see why we can’t just keep adding numbers to the last grouping to represent digital instantiations. We kicked around the notion of having a fourth set of numbers … We also considered structuring a code into the third set so you knew it was referring to a digital thing, like D01 or 501 where the D or 5 tells you it’s a digital instantiation, but that feels like a throwback to our original numbering system where WABI catalog numbers starting with 0 were negative and 1 was positive. That didn’t turn out to be very useful. My theory is that the less you ask of your number the better you are–except that it be unique identifier. Also considered making the numbers refer to specific things, like 004 is an mpeg2, 005 is mov file, but again its stupid and other fields can do that.

___________________________________________________
In the PBCore community, Jack Brighton reports

I’m currently using a website CMS to catalog A/V media objects and express them as web pages, RSS feeds, and PBCore-compliant XML. This is done using simple html forms to create MySQL tables, then querying the tables to render the metadata in XML templates. Until recently I was extracting the metadata from standard existing fields (like Title and Subject) needed for web page display, and porting them over to corresponding PBCore elements while hard-coding other values that were always the same for the objects being cataloged. This worked fine and created shareable metadata in PBCore format. Last week Mary Miller from the Peabody Archive worked with me to create a better tool using the same website CMS. We built a pair of cataloging forms, one for the intellectual and administrative metadata, and one for the Instantiation/technical metadata which then gets related to the “Platonic record” (Mary’s term). We then created an XML template to express everything as PBCore records. We could extend this very easily to templates for Dublin Core, MODS, MARC, etc. So metadata gets entered once, and can be rendered to different formats as needed. The CMS we’re using is ExpressionEngine, but the same thing would work with Drupal and probably many others. Here’s an example record (hint: don’t use Safari to view this - I love Safari but it doesn’t know how to render PBCore).

Jack Brighton, Director of Internet Development. WILL Public Media (Urbana IL)
jackb@uiuc.edu