[1046] in Public-Access_Computer_Systems_Forum
Re: Subject Headings for Gopher Virtual Library
daemon@ATHENA.MIT.EDU (Tony Barry)
Tue Aug 25 13:13:37 1992
Date: Tue, 25 Aug 1992 11:57:31 CDT
Reply-To: Public-Access Computer Systems Forum <PACS-L%UHUPVM1.BITNET@ricevm1.rice.edu>
From: Tony Barry <tony@library.anu.edu.au>
To: Multiple recipients of list PACS-L <PACS-L%UHUPVM1.BITNET@ricevm1.rice.edu>
----------------------------Original message----------------------------
In article <billy.713927567@sol.acs.unt.edu>, Billy Barron - VAX/UNIX
Systems Manager <billy@sol.acs.unt.edu> writes:
> Joel Cooper of Notre Dame and I are leading an effort to turn GopherSpace
> into a virtual library system.
I was starting to do that myself (gopher to info.anu.edu.au port 70) and ran
into the problems which this posting and subsequent responses refered to. So I
stopped until I had a better idea of which way to jump. At the moment I
favour a classified approach.
As Librarians our mind set is still gripped by the need to put
books in a linear sequence on a shelf and map a classification scheme
into that linear arrangement. Because it doesn't work we then
devise subject headings to give us random access to the books that
the classification scatters rather than bringing together.
The gopher however offers us a tree structue to pointers to information
with the ability to included multiple copies of the pointers in any
place that is relevent. All classifications are flawed and by their nature
must be. With gopher however we can put in pointers in more
than one place and get arround these flaws. LC headings however do
not map as gracefully into a tree structure as does a classification.
The various levels of the menus would need to refer to alphabetical
ranges rather than concepts.
There is also the problem of specificity. Most classification and
LC headings are designed for collections of millions. Initially at
least we will be dealing with collections of thousands. Classifications
are much easier to truncate and expand later if needed whereas
heading systems are not.
Currently I would think it best to organise material in a classified
hierachy with no more than a screen of 'leaves' at each level and
subdivide when the number of items gets too large. Put multiple
entries in for items which may be 'about' more than one classification.
______________________________________________________________________
Tony Barry. Head, Centre for Scholarly Information Access and Networks
Australian National University Library | fone +61 6 249 4632
GPO Box 4 Canberra City A.C.T. 2601, AUSTRALIA | phax +61 6 249 4063
tony@info.anu.edu.au, Tony_Barry@anu.edu.au | telex 10717252
