Skip to main content.
home | support | download

Back to List Archive

Re: Swish-e Ranking for Search within XML files

From: Tomek Noiński <noix(at)not-real.sphere.pl>
Date: Mon Sep 13 2004 - 15:37:28 GMT
Dnia Mon, 13 Sep 2004 07:15:44 -0700 (PDT)
Peter Karman <karman@cray.com> napisał:

> If you have the latest cvs build, try using the -R 1 option (or 
> RankScheme( 1 ) if using SWISH::API) and see if that makes a
> difference.

Thanks, I use SWISH::API and I tried the option, along with
"IgnoreTotalWordCountWhenRanking no".
It did help a little, but it's still not perfect.

I think something might not work well with MetaNamesRank.

I tried searching for something like "test or pcode=test or pres=test"
(there were other meta tags in the query, but not containing "test"
string).

With options:

MetaNamesRank 1000 pcode
MetaNamesRank 3 pres

.record with pcode="test" and no "test" in body or pres came before
another, with pcode="large" and "test" string in pres (one instance).

However, with options:

MetaNamesRank 1000 pcode
MetaNamesRank 4 pres

."large" came before "test".

Is it a bug or could it be a proper behaviour?
It doesn't seem intuitive, but I haven't studied the algorithms you use
:)

> You might try searching for 'test AND code=test' to see if that helps 
> make your results smaller (though that may not be what you want).

I don't want to burden users with a complicated interface and I form the
query string from what user types in a text box, like"(q) or meta1=(q)
or meta2=(q)...".
I believe it is a good way to do this?

-- 
Noix
Received on Mon Sep 13 08:37:40 2004