Hi, If a field always has a numeric value, does boosting it make any difference? I never tested this explicitly, but assuming there is a single numerical token in a field, I'd think that field can either match or not match, so its contribution to the score will always have one of 2 possible values, depending on whether it matched or not. If this is correct, then I'm not sure normalization matters here.
Otis ---- Sematext :: http://sematext.com/ :: Solr - Lucene - Nutch Lucene ecosystem search :: http://search-lucene.com/ ----- Original Message ---- > From: Gastone Penzo <gastone.pe...@gmail.com> > To: solr-user@lucene.apache.org > Sent: Wed, November 24, 2010 5:28:17 AM > Subject: boosting > > Hi, > i have a problem/question to explain: > > for example i have 3 documents with the same fields value, so the total > score is the same for all docs when i do a search > that match. > i have also some boosts set in bf parameter (boost functions in dismax). > > bf = field1^1.0 field2^0.8 field3^0.6 > > all the fields, except for 1, have values between 1 and 2 so i can control > the result score, > but the field1 has unknown value (it can be 1 or 100000). > It is a problem cause i would like to normalize this in a range between 1 > and 2 to compare score with the others. > > is it possibile? if not is there another way to do this?? > > Thank you > > -- > Gastone Penzo >
