jaepil commented on code in PR #15948: URL: https://github.com/apache/lucene/pull/15948#discussion_r3292058407
########## lucene/core/src/java/org/apache/lucene/search/BayesianScoreEstimator.java: ########## @@ -0,0 +1,228 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +package org.apache.lucene.search; + +import java.io.IOException; +import java.util.ArrayList; +import java.util.Arrays; +import java.util.List; +import java.util.Random; +import org.apache.lucene.index.IndexReader; +import org.apache.lucene.index.StoredFields; +import org.apache.lucene.index.Term; +import org.apache.lucene.util.ArrayUtil; + +/** + * Estimates {@link BayesianScoreQuery} parameters (alpha, beta, base rate) from corpus statistics + * via pseudo-query sampling. + * + * <p>The estimation algorithm: + * + * <ol> + * <li>Sample N documents randomly from the index + * <li>For each document, create a pseudo-query from its first few tokens in the target field + * <li>Run each pseudo-query via BM25 and collect the score distribution + * <li>Estimate: beta = median(scores), alpha = 1 / std(scores) + * <li>Estimate base rate: mean fraction of documents scoring above the 95th percentile + * </ol> + * + * @lucene.experimental + */ +public class BayesianScoreEstimator { Review Comment: I think the right lifecycle is tied to the `IndexReader/Searcher` rather than to each query. If the corpus evolves, a reopened NRT reader or a new commit has different reader-level statistics, so callers should re-run estimation when they move to that new searcher/reader. This PR keeps estimation explicit for that reason: the estimated Parameters are bound to the reader they were estimated from, and applications can choose their refresh cadence based on their searcher lifecycle. I avoided doing this lazily in `rewrite()` because the estimation work is too expensive and would need reader-keyed caching plus careful query identity handling. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
