[
https://issues.apache.org/jira/browse/LUCENE-10146?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17431363#comment-17431363
]
ASF subversion and git services commented on LUCENE-10146:
----------------------------------------------------------
Commit 6bb2bbcd6ab2e07a646c17351437ea5210b08004 in lucene's branch
refs/heads/main from Julie Tibshirani
[ https://gitbox.apache.org/repos/asf?p=lucene.git;h=6bb2bbc ]
LUCENE-10146: Add note that dot product is preferred over cosine (#400)
While VectorSimilarityFunction#COSINE is helpful when you need to preserve the
original vectors, it is significantly slower than DOT_PRODUCT. This commit adds
javadocs to COSINE explaining that dot product is the fastest option.
> Add VectorSimilarityFunction.COSINE
> -----------------------------------
>
> Key: LUCENE-10146
> URL: https://issues.apache.org/jira/browse/LUCENE-10146
> Project: Lucene - Core
> Issue Type: Improvement
> Reporter: Julie Tibshirani
> Priority: Major
> Fix For: main (9.0)
>
> Time Spent: 2h 20m
> Remaining Estimate: 0h
>
> To perform ANN search with cosine similarity, users are expected to normalize
> the document and query vectors to unit length, then use
> {{VectorSimilarityFunction.DOT_PRODUCT}}. I think it would be good to also
> support cosine similarity directly through
> {{VectorSimilarityFunction.COSINE}}. This would allow users to perform ANN
> based on cosine similarity, while retaining access to the original vectors
> through {{VectorValues}}. That way they can use the original vectors in a
> reranking step or return them to the application for further processing.
> It looks like nmslib and hnswlib support cosine similarity. On the other
> hand, FAISS only supports dot product and suggests users normalize the
> vectors to perform cosine similarity
> (https://github.com/facebookresearch/faiss/issues/95). To me adding this one
> additional similarity is worth it in terms of what it lets users accomplish.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]