org.terrier.matching.models

Class BM25

java.lang.Object

org.terrier.matching.models.WeightingModel

org.terrier.matching.models.BM25

All Implemented Interfaces:

Serializable, Cloneable, Model

public class BM25
extends WeightingModel

This class implements the Okapi BM25 weighting model. The default parameters used are:
k_1 = 1.2d
k_3 = 8d
b = 0.75d
The b parameter can be altered by using the setParameter method.

Author:

Gianni Amati, Ben He, Vassilis Plachouras

See Also:

Serialized Form

Field Summary

Fields inherited from class org.terrier.matching.models.WeightingModel

averageDocumentLength, c, cs, documentFrequency, es, i, keyFrequency, numberOfDocuments, numberOfPointers, numberOfTokens, numberOfUniqueTerms, rq, termFrequency

Constructor Summary

Constructors

Constructor and Description
BM25() A default constructor.

Method Summary

Methods

Modifier and Type	Method and Description
String	getInfo() Returns the name of the model.
double	getParameter() Returns the b parameter to the BM25 ranking formula as set by setParameter()
double	score(double tf, double docLength) Uses BM25 to compute a weight for a term in a document.
double	score(double tf, double docLength, double n_t, double F_t, double keyFrequency) Deprecated.
void	setParameter(double _b) Sets the b parameter to BM25 ranking formula

Methods inherited from class org.terrier.matching.models.WeightingModel

clone, getOverflowed, prepare, score, setCollectionStatistics, setEntryStatistics, setKeyFrequency, setRequest

Methods inherited from class java.lang.Object

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

Constructor Detail

BM25

public BM25()

A default constructor.

Method Detail

getInfo

public final String getInfo()

Returns the name of the model.

Specified by:

getInfo in interface Model

Specified by:

getInfo in class WeightingModel

Returns:

the name of the model

score

public double score(double tf,
           double docLength)

Uses BM25 to compute a weight for a term in a document.

Specified by:

score in class WeightingModel

Parameters:

tf - The term frequency in the document

docLength - the document's length

Returns:

the score assigned to a document with the given tf and docLength, and other preset parameters

score

@Deprecated
public double score(double tf,
                      double docLength,
                      double n_t,
                      double F_t,
                      double keyFrequency)

Deprecated.

This method provides the contract for implementing weighting models. As of Terrier 3.6, the 5-parameter score method is being deprecated since it is not used. The two parameter score method should be used instead. Tagged for removal in a later version.

Specified by:

score in class WeightingModel

Parameters:

tf - The term frequency in the document

docLength - the document's length

n_t - The document frequency of the term

F_t - the term frequency in the collection

keyFrequency - the term frequency in the query

Returns:

the score returned by the implemented weighting model.

setParameter

public void setParameter(double _b)

Sets the b parameter to BM25 ranking formula

Specified by:

setParameter in interface Model

Overrides:

setParameter in class WeightingModel

Parameters:

_b - the b parameter value to use.

getParameter

public double getParameter()

Returns the b parameter to the BM25 ranking formula as set by setParameter()

Specified by:

getParameter in interface Model

Overrides:

getParameter in class WeightingModel