Re: Library Link discussion (fwd)

Subject: Re: Library Link discussion (fwd)
Tony Barry (tonyb@dynamite.com.au)
Date: Sat, 6 Feb 1999 17:23:28 +1100

Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Message-Id: <v02110119b2e189f109fb@[203.37.43.24]>
Date: Sat, 6 Feb 1999 17:23:28 +1100
To: arl-ejournal@arl.org
From: tonyb@dynamite.com.au (Tony Barry)
Subject: Re: Library Link discussion (fwd)

On 5/2/99, Jean-Claude Guedon <guedon@ere.umontreal.ca> wrote:
>
> 2. If commercial indexers do not want to do it, libraries could do it.
> It could become a new cataloguing task of libraries and if they divide
> the task intelligently among themselves, the work could be fairly
> light and routine. Another good example of the implementation of a
> distributed intelligence society.

If you accept that keyword indexes are incredibly cheap for what they
do, Eric Lease Morgan already does this for ejournals in the field of
library and information science at
<http://sunsite.berkeley.edu/~emorgan/morganagus/>. He runs a robot
across their content.

Small scale robots are already in the reach of individuals - for
instance Maxxaum Computer Phantom <http://www.maxum.com/> software.
The ecomomies of print technology created an environment where large
indexing services flourished. The net could well be a place were small,
highly focussed indexing services are more effective. The complexity
of indexing languages such as MESH are a consequence of the large
domain in which they operate. The smaller the domain the better free
text indexing seems to work. Ally this to a cost in millicents per
article rather than dollars or tens of dollars and new economic models
will emerge.

tony

http://purl.oclc.org/NET/Tony.Barry

This archive was generated by hypermail 2a16 : Wed Dec 22 1999 - 09:15:05 EST