Re: [EP-underground] Full-text search capability for Eprint Archives (fwd)

(apologies if you receive this twice)

On Mon, 16 Jun 2003, Stevan Harnad wrote:
> It needs some thought whether it makes more sense to invert full-text
> at each local archive, or at a harvester level (google-like).

mnoGoSearch will use the db to do word->document indexing. ie. it provides 
the inverted-full-text feature, but uses the sql engine (internal indexes) 
for the actual search of the words (potentially faster and suitable to db 
optimizations/distribution/etc).

One can also partition the search spaces, to a folder in a filesystem, or 
to a website, or according to a specific field in the metadata... 
and support for byte/multibye characters (it's developed by a russian 
team), load distribution,etc

this gives us more details:
http://search.mnogo.ru/features.html

Regards, 
Silvio


--
http://www.iis.ee.ic.ac.uk/smacedo
Intelligent and Interactive Systems, Imperial College London

Non est gravis scire omnium responsum - est gravis scire reperir illum.
- and then there was Google