org.grouplens.lenskit.vectors.similarity

Class SignificanceWeightedVectorSimilarity

java.lang.Object

org.grouplens.lenskit.vectors.similarity.SignificanceWeightedVectorSimilarity

All Implemented Interfaces:

Serializable, VectorSimilarity

@Shareable
public class SignificanceWeightedVectorSimilarity
extends Object
implements VectorSimilarity, Serializable

Apply significance weighting to a similarity function. The threshold is configured with the SigWeightThreshold parameter.

Significance weighting decreases the similarity between two vectors when the number of common entities between the two vectors is low. For a threshold \(S\) and key sets \(K_1\) and \(K_2\), the similarity is multipled by \[\frac{|K_1 \cap K_2|}{\mathrm{max}(|K_1 \cap K_2|, S)}\]

• Herlocker, J., Konstan, J.A., and Riedl, J. An Empirical Analysis of Design Choices in Neighborhood-Based Collaborative Filtering Algorithms. Information Retrieval Vol. 5 Issue 4 (October 2002), pp. 287-310.

See Also:

SigWeightThreshold, Serialized Form

Constructor Summary

Constructors

Constructor and Description
SignificanceWeightedVectorSimilarity(int thresh, VectorSimilarity sim)

Method Summary

Modifier and Type	Method and Description
VectorSimilarity	getDelegate() Get the underlying similarity (for debuggin purposes).
boolean	isSparse() Query whether this similarity function is sparse (returns 0 for vectors with disjoint key sets).
boolean	isSymmetric() Query whether this similarity function is symmetric.
double	similarity(SparseVector vec1, SparseVector vec2) Compute the similarity between two vectors.
String	toString()

Methods inherited from class java.lang.Object

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

Constructor Detail

SignificanceWeightedVectorSimilarity

@Inject
public SignificanceWeightedVectorSimilarity(@SigWeightThreshold
                                                    int thresh,
                                                    VectorSimilarity sim)

Method Detail

getDelegate

public VectorSimilarity getDelegate()

Get the underlying similarity (for debuggin purposes).

Returns:

The wrapped vector similarity.

similarity

public double similarity(SparseVector vec1,
                         SparseVector vec2)

Description copied from interface: VectorSimilarity

Compute the similarity between two vectors.

Specified by:

similarity in interface VectorSimilarity

Parameters:

vec1 - The left vector to compare.

vec2 - The right vector to compare.

Returns:

The similarity, in the range [-1,1].

isSparse

public boolean isSparse()

Description copied from interface: VectorSimilarity

Query whether this similarity function is sparse (returns 0 for vectors with disjoint key sets).

Specified by:

isSparse in interface VectorSimilarity

Returns:

true iff VectorSimilarity.similarity(SparseVector, SparseVector) will always return true when applied to two vectors with no keys in common.

isSymmetric

public boolean isSymmetric()

Description copied from interface: VectorSimilarity

Query whether this similarity function is symmetric. Symmetric similarity functions return the same result when called on (A,B) and (B,A).

Specified by:

isSymmetric in interface VectorSimilarity

Returns:

true if the function is symmetric.

toString

public String toString()

Overrides:

toString in class Object