1 /*
2 * Licensed to the Apache Software Foundation (ASF) under one or more
3 * contributor license agreements. See the NOTICE file distributed with
4 * this work for additional information regarding copyright ownership.
5 * The ASF licenses this file to You under the Apache License, Version 2.0
6 * (the "License"); you may not use this file except in compliance with
7 * the License. You may obtain a copy of the License at
8 *
9 * http://www.apache.org/licenses/LICENSE-2.0
10 *
11 * Unless required by applicable law or agreed to in writing, software
12 * distributed under the License is distributed on an "AS IS" BASIS,
13 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
14 * See the License for the specific language governing permissions and
15 * limitations under the License.
16 */
17 package org.apache.commons.text.similarity;
18
19 import java.util.HashSet;
20 import java.util.Map;
21 import java.util.Set;
22
23 /**
24 * Measures the Cosine similarity of two vectors of an inner product space and
25 * compares the angle between them.
26 *
27 * <p>
28 * For further explanation about the Cosine Similarity, refer to
29 * http://en.wikipedia.org/wiki/Cosine_similarity.
30 * </p>
31 *
32 * @since 1.0
33 */
34 public class CosineSimilarity {
35
36 /**
37 * Calculates the cosine similarity for two given vectors.
38 *
39 * @param leftVector left vector
40 * @param rightVector right vector
41 * @return cosine similarity between the two vectors
42 */
43 public Double cosineSimilarity(final Map<CharSequence, Integer> leftVector, final Map<CharSequence, Integer> rightVector) {
44 if (leftVector == null || rightVector == null) {
45 throw new IllegalArgumentException("Vectors must not be null");
46 }
47
48 final Set<CharSequence> intersection = getIntersection(leftVector, rightVector);
49
50 final double dotProduct = dot(leftVector, rightVector, intersection);
51 double d1 = 0.0d;
52 for (final Integer value : leftVector.values()) {
53 d1 += Math.pow(value, 2);
54 }
55 double d2 = 0.0d;
56 for (final Integer value : rightVector.values()) {
57 d2 += Math.pow(value, 2);
58 }
59 double cosineSimilarity;
60 if (d1 <= 0.0 || d2 <= 0.0) {
61 cosineSimilarity = 0.0;
62 } else {
63 cosineSimilarity = (double) (dotProduct / (double) (Math.sqrt(d1) * Math.sqrt(d2)));
64 }
65 return cosineSimilarity;
66 }
67
68 /**
69 * Returns a set with strings common to the two given maps.
70 *
71 * @param leftVector left vector map
72 * @param rightVector right vector map
73 * @return common strings
74 */
75 private Set<CharSequence> getIntersection(final Map<CharSequence, Integer> leftVector,
76 final Map<CharSequence, Integer> rightVector) {
77 final Set<CharSequence> intersection = new HashSet<>(leftVector.keySet());
78 intersection.retainAll(rightVector.keySet());
79 return intersection;
80 }
81
82 /**
83 * Computes the dot product of two vectors. It ignores remaining elements. It means
84 * that if a vector is longer than other, then a smaller part of it will be used to compute
85 * the dot product.
86 *
87 * @param leftVector left vector
88 * @param rightVector right vector
89 * @param intersection common elements
90 * @return the dot product
91 */
92 private double dot(final Map<CharSequence, Integer> leftVector, final Map<CharSequence, Integer> rightVector,
93 final Set<CharSequence> intersection) {
94 long dotProduct = 0;
95 for (final CharSequence key : intersection) {
96 dotProduct += leftVector.get(key) * rightVector.get(key);
97 }
98 return dotProduct;
99 }
100
101 }