Isn't the point of BM25 to use variable document length normalization? It works when used on the entire index/corpus. It is meant to influence the TF values.
Comparing results between Lucene queries is not advisable. Why did you switch to BM25? Do you field lengths vary much?
Hi,I'm wondering is there away to calculate BM25 score for the query itself (query-against-query)?Adding it to index seems to be invalid solution as it will influent the tf-idf of the index and make it corrupted.Thank you,Bohdan
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email].
To view this discussion on the web visit https://groups.google.com/d/msgid/elasticsearch/2183d2cc-a997-4397-aa0b-c5ac2fcedd71%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.