I've gone on and on on various lists about trying to analyze MARC as data elements, to the point that I'm sure many people just wish I'd shut up. The best way to shut me up is for me to either finish or at least take the analysis as far as I can. To that end, I now have a wiki page that gives my analysis of the MARC fixed fields (007-008). (Home page for project.)
These fields are pretty straight-forward since by their definition they are already defined as discrete data elements. The only tricky bit is that each data element is tied to a particular resource format (e.g. text, map, video).
I have captured the link between the elements as elements and MARC by creating a key that includes the tag, the format, and the position within the tag:
007microform05
For the actual values (and my analysis includes the values, although I skipped things like "no attempt to code" -- those could be added in), I follow the same format:
007microform05a
There is an example of a single element in Turtle and RDF/XML on the wiki page. There is also a link to the whole in Turtle and RDF/XML, which I might as well link to here:
Note that these files, in fact the whole project, is basically an attempt at a proof of concept. Treat it as totally BETA and make any comments, suggestions, etc. that you want. If you see this as being useful, feel free to volunteer some time to make it better.
Special thanks to Gordon Dunsire for turning out code from my text.
Note:
I've added a link to an HTML display of the first analysis of the 0XX fields to that page. The trick here is to figure out which subfields can stand alone as data elements, and which have dependencies (like "source of code") that require them to be compound elements.
No comments:
Post a Comment