[EP-underground] Re: Interoperability - subject classification/terminology

[Thread: http://www.ecs.soton.ac.uk/~harnad/Hypermail/Amsci/2384.html ]

I agree 100% with the point made by the commentator below:
Institutional Eprint Archives for refereed research papers do *not*
require an elaborate classification system (such as Library of
Congress). These are not books. And the OAI harvesters and search
engines will be the real, cross-archive search tools; elaborate
pre-classification is not needed just for searching within one's own
university's local research output (and creating such an elaborate
classification system is, in my opinion, a waste of time). (And in any
case, I would put my money on boolean inverted full-text search, with
scientometric impact ranking, over any prefabricated human taxonomy in
this online age.)

Reply to comment below: Just pick in one default subject and forget
about the rest.

Stevan Harnad


 On Fri, 7 Mar 2003, W F Clocksin wrote:

> Hi. I am a beginning user of Eprints, and am entering metadata on the 
> default test archive interface. It is a real nuisance to have to 
> specify the Subject (which uses the Library of Congress system).  For 
> books this makes sense because the catalog information is in the front 
> matter of the book, but it is unclear to me why I should have to do 
> this for journal articles. For multidisciplinary articles, it might 
> mean specifying a number of Subjects using the scrolling textbox, which 
> would take longer than copy/pasting the rest of the metadata. I would 
> rather just leave out the Subject. To what extent is a required Subject 
> built into ePrints, or is it simply feature of the test interface that 
> I could omit from a custom interface?
> 
> William Clocksin
> www.clocksin.com