On 7/14/06, Marvin Humphrey <[EMAIL PROTECTED]> wrote:
Yonik, I disagree on one point. I recommend against omitting norms for title fields.
Well, yes, I should have said "sometimes", when you don't need or want length normalization. The scenarios where you don't want/need length normalization in full-text fields is typically with fields that are restricted to being short (like title or name). It's definitely corpus dependent though.
KinoSearch adopted a default tf() truncation scheme where all fields were normalized as if they contained a minimum of 100 tokens.
The kind of "title" fields I was thinking of were definitely less than 100 tokens, so it amounts to the same thing (but my advice should have been clearer). -Yonik http://incubator.apache.org/solr Solr, the open-source Lucene search server --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
