Yes, this is probably the right way, but the problem is how. The method lengthNorm(String fieldName, int numTerms) of the Similarity Class takes as input parameter only the number of terms of the current doc being indexed. Instead, I need the evaluate the sum of the squared of the distinct term frequencies.
Claudio -----Messaggio originale----- Da: Grant Ingersoll [mailto:gsing...@apache.org] Inviato: lunedì 10 agosto 2009 15.18 A: java-user@lucene.apache.org Oggetto: Re: sumOfSquaredWeights for lengthNorm You can override the Similarity class and set it on both the IndexWriter and the IndexReader. Is that your question? On Aug 10, 2009, at 3:55 AM, Claudio Gennaro wrote: > Dear all, > > > > I read a very old message on the list about the use of > sumOfSquaredWeights > in lengthNorm (Mon, 06 Mar 2006 19:06:28 GMT from Eugene). I just > wanted to > do the same, i.e., to override the Similarity class of Lucene in > order to > normalize the term weights of a doc with the sum of the squares of > doc term > frequencies in the doc. > --------------------------------------------------------------------- To unsubscribe, e-mail: java-user-unsubscr...@lucene.apache.org For additional commands, e-mail: java-user-h...@lucene.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: java-user-unsubscr...@lucene.apache.org For additional commands, e-mail: java-user-h...@lucene.apache.org
