com.graphaware.nlp.ml.similarity.CosineSimilarity Maven / Gradle / Ivy

Go to download
Show more of this group Show more artifacts with this name
Show all versions of nlp Show documentation
GraphAware Framework Module for adding NLP capabilities to Neo4j.
The newest version!
/*
 * Copyright (c) 2013-2018 GraphAware
 *
 * This file is part of the GraphAware Framework.
 *
 * GraphAware Framework is free software: you can redistribute it and/or modify it under the terms of
 * the GNU General Public License as published by the Free Software Foundation, either
 * version 3 of the License, or (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY;
 * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
 * See the GNU General Public License for more details. You should have received a copy of
 * the GNU General Public License along with this program.  If not, see
 * .
 */
package com.graphaware.nlp.ml.similarity;

import java.util.List;
import java.util.Map;
import java.util.TreeSet;
import java.util.concurrent.atomic.AtomicReference;
import java.util.stream.IntStream;

public class CosineSimilarity implements Similarity {

    @Override
    public float getSimilarity(Map xVector, Map yVector) {
        float a = getDotProduct(xVector, yVector);
        float b = getNorm(xVector) * getNorm(yVector);

        if (b > 0) {
            return a / b;
        } else {
            return 0;
        }
    }
    //improved version for user functions
    public double getSimilarity(List xVector, List yVector) {
        double a = getDotProduct(xVector, yVector);
        double b = getNorm(xVector) * getNorm(yVector);

        if (b > 0) {
            return a / b;
        } else {
            return 0;
        }
    }

    public double getSimilarity(float[] xVector, float[] yVector) {
        double a = getDotProduct(xVector, yVector);
        double b = getNorm(xVector) * getNorm(yVector);

        if (b > 0) {
            return a / b;
        } else {
            return 0;
        }
    }

    private float getDotProduct(final Map xVector, final Map yVector) {

        final AtomicReference sum = new AtomicReference<>(0f);
        TreeSet keys = new TreeSet<>(xVector.keySet());
        keys.addAll(yVector.keySet());
        keys.stream().forEach((key) -> {
            if (xVector.containsKey(key) && yVector.containsKey(key)) {
                float curValue = sum.get();
                curValue += xVector.get(key) * yVector.get(key);
                sum.set(curValue);
            }
        });
        return sum.get();
    }

    private double getDotProduct(final List xVector, final List yVector) {

        final AtomicReference sum = new AtomicReference<>(0f);
        IntStream.range(0, xVector.size())
                .boxed().forEach((key) -> {
                    float curValue = sum.get();
                    curValue += xVector.get(key) * yVector.get(key);
                    sum.set(curValue);
                });
        return sum.get();
    }


    private double getDotProduct(final float[] xVector, final float[] yVector) {

        final AtomicReference sum = new AtomicReference<>(0f);
        IntStream.range(0, xVector.length)
                .boxed().forEach((key) -> {
            float curValue = sum.get();
            curValue += xVector[key] * yVector[key];
            sum.set(curValue);
        });
        return sum.get();
    }

    private float getNorm(Map xVector) {
        final AtomicReference sum = new AtomicReference<>(0f);
        xVector.values().stream().forEach((value) -> {
            float curValue = sum.get();
            curValue += value * value;
            sum.set(curValue);
        });
        return Double.valueOf(Math.sqrt(sum.get().doubleValue())).floatValue();
    }
    
    private double getNorm(List xVector) {
        final AtomicReference sum = new AtomicReference<>(0f);
        xVector.stream().forEach((value) -> {
            float curValue = sum.get();
            curValue += value * value;
            sum.set(curValue);
        });
        return Math.sqrt(sum.get().doubleValue());
    }

    private double getNorm(float[] xVector) {
        final AtomicReference sum = new AtomicReference<>(0f);
        for (float value: xVector) {
            float curValue = sum.get();
            curValue += value * value;
            sum.set(curValue);
        }
        return Math.sqrt(sum.get().doubleValue());
    }

    public double cosineSimilarity(float[] vectorA, float[] vectorB) {
        double dotProduct = 0.0d;
        double normA = 0.0d;
        double normB = 0.0d;
        for (int i = 0; i < vectorA.length; i++) {
            dotProduct += vectorA[i] * vectorB[i];
            normA += Math.pow(vectorA[i], 2);
            normB += Math.pow(vectorB[i], 2);
        }
        return dotProduct / (Math.sqrt(normA) * Math.sqrt(normB));
    }

}