org.apache.lucene.search.similarities
Class BM25Similarity

java.lang.Object
  extended by org.apache.lucene.search.similarities.Similarity
      extended by org.apache.lucene.search.similarities.BM25Similarity

public class BM25Similarity
extends Similarity

BM25 Similarity. Introduced in Stephen E. Robertson, Steve Walker, Susan Jones, Micheline Hancock-Beaulieu, and Mike Gatford. Okapi at TREC-3. In Proceedings of the Third Text REtrieval Conference (TREC 1994). Gaithersburg, USA, November 1994.

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

Nested Class Summary
 
Nested classes/interfaces inherited from class org.apache.lucene.search.similarities.Similarity
Similarity.SimScorer, Similarity.SimWeight
 
Field Summary
protected  boolean discountOverlaps
          True if overlap tokens (tokens with a position of increment of zero) are discounted from the document's length.
 
Constructor Summary
BM25Similarity()
          BM25 with these default values: k1 = 1.2, b = 0.75.
BM25Similarity(float k1, float b)
          BM25 with the supplied parameter values.
 
Method Summary
protected  float avgFieldLength(CollectionStatistics collectionStats)
          The default implementation computes the average as sumTotalTermFreq / maxDoc, or returns 1 if the index does not store sumTotalTermFreq (Lucene 3.x indexes or any field that omits frequency information).
 long computeNorm(FieldInvertState state)
          Computes the normalization value for a field, given the accumulated state of term processing for this field (see FieldInvertState).
 Similarity.SimWeight computeWeight(float queryBoost, CollectionStatistics collectionStats, TermStatistics... termStats)
          Compute any collection-level weight (e.g.
protected  float decodeNormValue(byte b)
          The default implementation returns 1 / f2 where f is SmallFloat.byte315ToFloat(byte).
protected  byte encodeNormValue(float boost, int fieldLength)
          The default implementation encodes boost / sqrt(length) with SmallFloat.floatToByte315(float).
 float getB()
          Returns the b parameter
 boolean getDiscountOverlaps()
          Returns true if overlap tokens are discounted from the document's length.
 float getK1()
          Returns the k1 parameter
protected  float idf(long docFreq, long numDocs)
          Implemented as log(1 + (numDocs - docFreq + 0.5)/(docFreq + 0.5)).
 Explanation idfExplain(CollectionStatistics collectionStats, TermStatistics termStats)
          Computes a score factor for a simple term and returns an explanation for that score factor.
 Explanation idfExplain(CollectionStatistics collectionStats, TermStatistics[] termStats)
          Computes a score factor for a phrase.
protected  float scorePayload(int doc, int start, int end, BytesRef payload)
          The default implementation returns 1
 void setDiscountOverlaps(boolean v)
          Sets whether overlap tokens (Tokens with 0 position increment) are ignored when computing norm.
 Similarity.SimScorer simScorer(Similarity.SimWeight stats, AtomicReaderContext context)
          Creates a new Similarity.SimScorer to score matching documents from a segment of the inverted index.
protected  float sloppyFreq(int distance)
          Implemented as 1 / (distance + 1).
 String toString()
           
 
Methods inherited from class org.apache.lucene.search.similarities.Similarity
coord, queryNorm
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait
 

Field Detail

discountOverlaps

protected boolean discountOverlaps
True if overlap tokens (tokens with a position of increment of zero) are discounted from the document's length.

Constructor Detail

BM25Similarity

public BM25Similarity(float k1,
                      float b)
BM25 with the supplied parameter values.

Parameters:
k1 - Controls non-linear term frequency normalization (saturation).
b - Controls to what degree document length normalizes tf values.

BM25Similarity

public BM25Similarity()
BM25 with these default values:

Method Detail

idf

protected float idf(long docFreq,
                    long numDocs)
Implemented as log(1 + (numDocs - docFreq + 0.5)/(docFreq + 0.5)).


sloppyFreq

protected float sloppyFreq(int distance)
Implemented as 1 / (distance + 1).


scorePayload

protected float scorePayload(int doc,
                             int start,
                             int end,
                             BytesRef payload)
The default implementation returns 1


avgFieldLength

protected float avgFieldLength(CollectionStatistics collectionStats)
The default implementation computes the average as sumTotalTermFreq / maxDoc, or returns 1 if the index does not store sumTotalTermFreq (Lucene 3.x indexes or any field that omits frequency information).


encodeNormValue

protected byte encodeNormValue(float boost,
                               int fieldLength)
The default implementation encodes boost / sqrt(length) with SmallFloat.floatToByte315(float). This is compatible with Lucene's default implementation. If you change this, then you should change decodeNormValue(byte) to match.


decodeNormValue

protected float decodeNormValue(byte b)
The default implementation returns 1 / f2 where f is SmallFloat.byte315ToFloat(byte).


setDiscountOverlaps

public void setDiscountOverlaps(boolean v)
Sets whether overlap tokens (Tokens with 0 position increment) are ignored when computing norm. By default this is true, meaning overlap tokens do not count when computing norms.


getDiscountOverlaps

public boolean getDiscountOverlaps()
Returns true if overlap tokens are discounted from the document's length.

See Also:
setDiscountOverlaps(boolean)

computeNorm

public final long computeNorm(FieldInvertState state)
Description copied from class: Similarity
Computes the normalization value for a field, given the accumulated state of term processing for this field (see FieldInvertState).

Matches in longer fields are less precise, so implementations of this method usually set smaller values when state.getLength() is large, and larger values when state.getLength() is small.

Specified by:
computeNorm in class Similarity
Parameters:
state - current processing state for this field
Returns:
computed norm value

idfExplain

public Explanation idfExplain(CollectionStatistics collectionStats,
                              TermStatistics termStats)
Computes a score factor for a simple term and returns an explanation for that score factor.

The default implementation uses:

 idf(docFreq, searcher.maxDoc());
 
Note that CollectionStatistics.maxDoc() is used instead of IndexReader#numDocs() because also TermStatistics.docFreq() is used, and when the latter is inaccurate, so is CollectionStatistics.maxDoc(), and in the same direction. In addition, CollectionStatistics.maxDoc() is more efficient to compute

Parameters:
collectionStats - collection-level statistics
termStats - term-level statistics for the term
Returns:
an Explain object that includes both an idf score factor and an explanation for the term.

idfExplain

public Explanation idfExplain(CollectionStatistics collectionStats,
                              TermStatistics[] termStats)
Computes a score factor for a phrase.

The default implementation sums the idf factor for each term in the phrase.

Parameters:
collectionStats - collection-level statistics
termStats - term-level statistics for the terms in the phrase
Returns:
an Explain object that includes both an idf score factor for the phrase and an explanation for each term.

computeWeight

public final Similarity.SimWeight computeWeight(float queryBoost,
                                                CollectionStatistics collectionStats,
                                                TermStatistics... termStats)
Description copied from class: Similarity
Compute any collection-level weight (e.g. IDF, average document length, etc) needed for scoring a query.

Specified by:
computeWeight in class Similarity
Parameters:
queryBoost - the query-time boost.
collectionStats - collection-level statistics, such as the number of tokens in the collection.
termStats - term-level statistics, such as the document frequency of a term across the collection.
Returns:
SimWeight object with the information this Similarity needs to score a query.

simScorer

public final Similarity.SimScorer simScorer(Similarity.SimWeight stats,
                                            AtomicReaderContext context)
                                     throws IOException
Description copied from class: Similarity
Creates a new Similarity.SimScorer to score matching documents from a segment of the inverted index.

Specified by:
simScorer in class Similarity
Parameters:
stats - collection information from Similarity.computeWeight(float, CollectionStatistics, TermStatistics...)
context - segment of the inverted index to be scored.
Returns:
SloppySimScorer for scoring documents across context
Throws:
IOException - if there is a low-level I/O error

toString

public String toString()
Overrides:
toString in class Object

getK1

public float getK1()
Returns the k1 parameter

See Also:
BM25Similarity(float, float)

getB

public float getB()
Returns the b parameter

See Also:
BM25Similarity(float, float)


Copyright © 2000-2013 Apache Software Foundation. All Rights Reserved.