In English, fieldNorm essentially means: "give term hits in shorter fields more weight/importance than those in longer fields".
I believe the implementation is 1/sqrt(number of terms in field). Keep in mind that index-time boost is calculated into the field norm. Otis ----- Original Message ---- From: Seeta Somagani <[EMAIL PROTECTED]> To: java-user@lucene.apache.org Sent: Thursday, June 8, 2006 8:00:22 PM Subject: fieldNorm Hi, My simple search application is returning hits ordered by the default relevance. But, I noticed that when for example searched for "what", the first hit has considerably lesser occurrences of the term "what" than those that follow. When I viewed the explanation for each document, all the values made sense except the fieldNorm which is comparatively very high for the topmost hit. Can someone please explain how the fieldNorm factor is calculated? Thanks so much. Seeta --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
