info.debatty.java.stringsimilarity.Cosine Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of java-string-similarity Show documentation
Show all versions of java-string-similarity Show documentation
Implementation of various string similarity and distance algorithms: Levenshtein, Jaro-winkler, n-Gram, Q-Gram, Jaccard index, Longest Common Subsequence edit distance, cosine similarity...
/*
* The MIT License
*
* Copyright 2015 Thibault Debatty.
*
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this software and associated documentation files (the "Software"), to deal
* in the Software without restriction, including without limitation the rights
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
* copies of the Software, and to permit persons to whom the Software is
* furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
* THE SOFTWARE.
*/
package info.debatty.java.stringsimilarity;
import info.debatty.java.stringsimilarity.interfaces.NormalizedStringSimilarity;
import info.debatty.java.stringsimilarity.interfaces.NormalizedStringDistance;
import net.jcip.annotations.Immutable;
/**
* 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
*/
@Immutable
public class Cosine extends ShingleBased implements
NormalizedStringDistance, NormalizedStringSimilarity {
/**
* 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.
*
* @param k
*/
public Cosine(final int k) {
super(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.
* Default k is 3.
*/
public Cosine() {
super();
}
/**
* Compute the cosine similarity between strings.
* @param s1
* @param s2
* @return
*/
public final double similarity(final String s1, final String s2) {
if (s1.length() < getK() || s2.length() < getK()) {
return 0;
}
KShingling ks = new KShingling(getK());
int[] profile1 = ks.getArrayProfile(s1);
int[] profile2 = ks.getArrayProfile(s2);
return dotProduct(profile1, profile2)
/ (norm(profile1) * norm(profile2));
}
/**
* Compute the norm L2 : sqrt(Sum_i( v_i²)).
*
* @param profile
* @return L2 norm
*/
private static double norm(final int[] profile) {
double agg = 0;
for (int v : profile) {
agg += 1.0 * v * v;
}
return Math.sqrt(agg);
}
private static double dotProduct(
final int[] profile1, final int[] profile2) {
// profiles may not have the same length
int length = Math.min(profile1.length, profile2.length);
double agg = 0;
for (int i = 0; i < length; i++) {
agg += 1.0 * profile1[i] * profile2[i];
}
return agg;
}
/**
* Return 1.0 - similarity.
* @param s1
* @param s2
* @return
*/
public final double distance(final String s1, final String s2) {
return 1.0 - similarity(s1, s2);
}
}