On Fri, Mar 27, 2015 at 12:46 PM, Kelley McGrath <[log in to unmask]> wrote:
I strongly disagree with the statement that  someone who understands MARCXML can do whatever they want with the data. I think I have a pretty good grasp of MARC and I have spent countless frustrating hours trying to get information out of MARC records (and not always succeeding).
Granted, I'm not a developer, but I've worked with people who are good developers so I don't think that's the bottleneck.

I'll stick to LCSH and related topics in this note, since that is the area I have studied the most, and thus one about which I know the least 

* Sometimes the data isn't in the MARC XML record.  

For example: 
  • LC subjects files do not record NT relationships.  This information may be inferred.    
  • Sub-divided headings do not necessarily carry BT relationships to the implied broader terms (which may not established).
  • Headings, Inverted is not related to Headings. 
* Sometimes extracting information from the data  requires deep common-sense reasoning to extract relationships/patterns. e.g.
  • Headings are established as compound noun phrases.  Ask a linguist about the semantics of English compound noun phrases... 
* Sometimes processing requires knowledge from the SCM *as it existed at the time a heading was created/applied

I will ignore the problematic  fact that the structure of LCSH has been seriously corrupted for getting on 30 years (since the conversion to BT/NT/RT). 

[I will also ignore the fact that as of 2011 there was no peer-reviewed literature that demonstrated that subdivided subject headings could be said to have meaning.  Granted, at that time there had only ever been one study, and that with problematic methodology.]

BibFrame solves all of this  by  +++ATH