lucene-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Chris Hostetter <>
Subject RE: Low hits
Date Tue, 23 Jan 2007 19:43:57 GMT

: When I index the whole website, then when I type a title of a document I
: have like 60 to 70 % as score.
: When I index only one page, then when I type the title I have like 2% as
: score.

I don't know what Regain is ... but this sounds like some issue between
how it reports the scores Lucene generates.  Lucene scores are not
percentages, they are either a "raw score" which is an absolute
calculation absed on the query boosts and constants in your Similarity
class -- raw scores can be any floating point number; or they are a "Hits
semi-nomalized score" which are garunteed to be between 0 and 1, but the
"best" matching document is only garunteed to have a score no greater then
1 -- it is not garunteed to have a score that *is* 1.

perhaps Regain is translating the Hits score as a percentage -- in which
case there is nothing wrong with the top matching document haveing a score
of "2%" ... it's still the top matching document.

I would suggest you start by asking some questions of the Regain user
community or Regain is a commercial product, ask their customer support
people aboutthe scores.


To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message