[WEB4LIB] Re: metadata with Dublin Core

[Tony Barry]

At 1:28 PM -0700 4/6/01, Lee Jaffe wrote:
>I've been wondering about this in relation to availability of texts
>online, thinking that libraries needed to create an equivalent to
>Napster for document.  Perhaps we could put together a server that
>would search for and index documents based on Dublin Core.
>Normally I'd say that it would be better for us to work within
>the Web mainstream (i.e., access via Google searches and Yahoo)
>even though that doesn't give us the precision or authority we'd
>like.  However, the example of Napster shows that it doesn't
>hurt to go through a separate channel within the larger stream.

The future is not in the big search engines. Eventually they won't 
scale and the limitations of free text searching will remain a 
problem as will spamming of indexes.

The future is in selective indexing of selected sites aided by 
quality indexing and Dublin Core http://dublincore.org/ and RDF 
http://www.w3.org/RDF/ provide a framework to do it in. For instance 
you wouldn't search for Australian Government information through a 
general search engine. You would use http://www.fed.gov.au which used 
DC. The best example I know is a pictorial search engine 
PictureAustralia http://www.pictureaustralia.org/about.html which is 
a central index of images mounted at a growing range of major 
cultural institution which uses controlled indexing embedded in 
Dublin Core.

As regards a Napster like solution for libraries p2p networking and 
existing tools already provide the building blocks.

Guntella http://www.gnutella.wego.com/ for instance already is being 
used to swap documents.

The building blocks to collect, serve and describe documents are all 
there. I'm sure something could be put together with -

Ariel http://www.rlg.org/ariel/index.html to capture the data

Prospero http://auto.med.ohio-state.edu/mailman/listinfo/prospero to 
store or disseminate it and unpickle the GEDI formatting

Open source tools could be used http://oss4lib.org/

And possibilities for front ends exist with mylibrary 
http://my.lib.ncsu.edu/about/ Gossamer threads 
http://www.gossamer-threads.com/scripts/ and zope http://www.zope.org/

Of course all this would be illegal is done by libraries because of 
copyright however if done by end users themselves ....

Anyway there is always jake http://jake.med.yale.edu/docs/about.html 
to try and keep within the law.

Tony
-- 
phone  +61 2 6241 7659
mailto:me at Tony-Barry.emu.id.au
http://purl.oclc.org/NET/Tony.Barry