Package org.apache.lucene.search

Class BayesianScoreEstimator

java.lang.Object

org.apache.lucene.search.BayesianScoreEstimator

public class BayesianScoreEstimator
extends Object

Estimates BayesianScoreQuery parameters (alpha, beta, base rate) from corpus statistics via pseudo-query sampling.

The estimation algorithm:

1. Reservoir-sample terms from the target field's indexed vocabulary
2. Partition the sampled terms into pseudo-queries
3. Run each pseudo-query via BM25 and collect the score distribution
4. Estimate: beta = median(scores), alpha = 1 / std(scores)
5. Estimate base rate: mean fraction of documents scoring above the 95th percentile

WARNING: This API is experimental and might change in incompatible ways in the next release.

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static final record	BayesianScoreEstimator.Parameters	Estimated parameters for BayesianScoreQuery.

Method Summary

Modifier and Type	Method	Description
static BayesianScoreEstimator.Parameters	estimate(IndexSearcher searcher, String field)	Estimates parameters with default settings (50 samples, 5 tokens per query, seed 42).
static BayesianScoreEstimator.Parameters	estimate(IndexSearcher searcher, String field, int nSamples, int tokensPerQuery, long seed)	Estimates BayesianScoreQuery parameters from the given index.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Method Details

estimate

public static BayesianScoreEstimator.Parameters estimate(IndexSearcher searcher,
 String field,
 int nSamples,
 int tokensPerQuery,
 long seed)
                                                  throws IOException

Estimates BayesianScoreQuery parameters from the given index.

Parameters:

searcher - the index searcher to sample from

field - the indexed text field to create pseudo-queries for

nSamples - number of pseudo-queries to sample (default 50)

tokensPerQuery - number of indexed terms per pseudo-query (default 5)

seed - random seed for reproducible sampling

Returns:

estimated alpha, beta, and base rate

Throws:

IOException - if an I/O error occurs reading the index

estimate

public static BayesianScoreEstimator.Parameters estimate(IndexSearcher searcher,
 String field)
                                                  throws IOException

Estimates parameters with default settings (50 samples, 5 tokens per query, seed 42).

Parameters:

searcher - the index searcher

field - the text field

Returns:

estimated parameters

Throws:

IOException - if an I/O error occurs