info.debatty.java.stringsimilarity

Class Cosine

java.lang.Object

info.debatty.java.stringsimilarity.Cosine

All Implemented Interfaces:

NormalizedStringDistance, NormalizedStringSimilarity, StringDistance, StringSimilarity, Serializable

@Immutable
public class Cosine
extends Object
implements NormalizedStringDistance, NormalizedStringSimilarity

The similarity between the two strings is the cosine of the angle between these two vectors representation. It is computed as V1 . V2 / (|V1| * |V2|) The cosine distance is computed as 1 - cosine similarity.

Author:

Thibault Debatty

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
Cosine() Implements Cosine Similarity between strings.
Cosine(int k) Implements Cosine Similarity between strings.

Method Summary

Modifier and Type	Method and Description
double	distance(String s1, String s2) Return 1.0 - similarity.
int	getK() Return k, the length of k-shingles (aka n-grams).
Map<String,Integer>	getProfile(String string) Compute and return the profile of s, as defined by Ukkonen "Approximate string-matching with q-grams and maximal matches".
double	similarity(Map<String,Integer> profile1, Map<String,Integer> profile2)
double	similarity(String s1, String s2) Compute the cosine similarity between strings.

Methods inherited from class java.lang.Object

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

Constructor Detail

Cosine

public Cosine(int k)

Implements Cosine Similarity between strings. The strings are first transformed in vectors of occurrences of k-shingles (sequences of k characters). In this n-dimensional space, the similarity between the two strings is the cosine of their respective vectors.

Parameters:

k -

Cosine

public Cosine()

Implements Cosine Similarity between strings. The strings are first transformed in vectors of occurrences of k-shingles (sequences of k characters). In this n-dimensional space, the similarity between the two strings is the cosine of their respective vectors. Default k is 3.

Method Detail

similarity

public final double similarity(String s1,
                               String s2)

Compute the cosine similarity between strings.

Specified by:

similarity in interface StringSimilarity

Parameters:

s1 - The first string to compare.

s2 - The second string to compare.

Returns:

The cosine similarity in the range [0, 1]

Throws:

NullPointerException - if s1 or s2 is null.

distance

public final double distance(String s1,
                             String s2)

Return 1.0 - similarity.

Specified by:

distance in interface StringDistance

Parameters:

s1 - The first string to compare.

s2 - The second string to compare.

Returns:

1.0 - the cosine similarity in the range [0, 1]

Throws:

NullPointerException - if s1 or s2 is null.

similarity

public final double similarity(Map<String,Integer> profile1,
                               Map<String,Integer> profile2)

Parameters:

profile1 -

profile2 -

Returns:

getK

public int getK()

Return k, the length of k-shingles (aka n-grams).

Returns:

The length of k-shingles.

getProfile

public final Map<String,Integer> getProfile(String string)

Compute and return the profile of s, as defined by Ukkonen "Approximate string-matching with q-grams and maximal matches". https://www.cs.helsinki.fi/u/ukkonen/TCS92.pdf The profile is the number of occurrences of k-shingles, and is used to compute q-gram similarity, Jaccard index, etc. Pay attention: the memory requirement of the profile can be up to k * size of the string

Parameters:

string -

Returns:

the profile of this string, as an unmodifiable Map