edu.ucla.sspace.bigram.BigramSpace Maven / Gradle / Ivy

Go to download

Show more of this group Show more artifacts with this name
Show all versions of sspace-wordsi Show documentation

The S-Space Package is a collection of algorithms for building Semantic Spaces as well as a highly-scalable library for designing new distributional semantics algorithms. Distributional algorithms process text corpora and represent the semantic for words as high dimensional feature vectors. This package also includes matrices, vectors, and numerous clustering algorithms. These approaches are known by many names, such as word spaces, semantic spaces, or distributed semantics and rest upon the Distributional Hypothesis: words that appear in similar contexts have similar meanings.

The newest version!

/*
 * Copyright (c) 2011, Lawrence Livermore National Security, LLC. Produced at
 * the Lawrence Livermore National Laboratory. Written by Keith Stevens,
 * [email protected] OCEC-10-073 All rights reserved. 
 *
 * This file is part of the S-Space package and is covered under the terms and
 * conditions therein.
 *
 * The S-Space package is free software: you can redistribute it and/or modify
 * it under the terms of the GNU General Public License version 2 as published
 * by the Free Software Foundation and distributed hereunder to you.
 *
 * THIS SOFTWARE IS PROVIDED "AS IS" AND NO REPRESENTATIONS OR WARRANTIES,
 * EXPRESS OR IMPLIED ARE MADE.  BY WAY OF EXAMPLE, BUT NOT LIMITATION, WE MAKE
 * NO REPRESENTATIONS OR WARRANTIES OF MERCHANT- ABILITY OR FITNESS FOR ANY
 * PARTICULAR PURPOSE OR THAT THE USE OF THE LICENSED SOFTWARE OR DOCUMENTATION
 * WILL NOT INFRINGE ANY THIRD PARTY PATENTS, COPYRIGHTS, TRADEMARKS OR OTHER
 * RIGHTS.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program. If not, see .
 */

package edu.ucla.sspace.bigram;

import edu.ucla.sspace.basis.BasisMapping;
import edu.ucla.sspace.basis.StringBasisMapping;
import edu.ucla.sspace.common.SemanticSpace;
import edu.ucla.sspace.matrix.GrowingSparseMatrix;
import edu.ucla.sspace.matrix.BaseTransform;
import edu.ucla.sspace.matrix.FilteredTransform;
import edu.ucla.sspace.matrix.PointWiseMutualInformationTransform;
import edu.ucla.sspace.matrix.SparseMatrix;
import edu.ucla.sspace.matrix.Transform;
import edu.ucla.sspace.text.IteratorFactory;
import edu.ucla.sspace.vector.SparseDoubleVector;

import java.io.BufferedReader;
import java.io.IOException;

import java.util.ArrayDeque;
import java.util.Iterator;
import java.util.Properties;
import java.util.Queue;
import java.util.Set;


/**
 * @author Keith Stevens
 */
public class BigramSpace implements SemanticSpace {

    private final BasisMapping basis;

    private final SparseMatrix bigramMatrix;

    private final int windowSize;

    private final Transform filter;

    public BigramSpace() {
        this(new StringBasisMapping(), 8,
             new PointWiseMutualInformationTransform(), 5);
    }

    public BigramSpace(BasisMapping basis, 
                       int windowSize,
                       BaseTransform base,
                       double minValue) {
        this.basis = basis;
        this.windowSize = windowSize;
        this.filter = new FilteredTransform(base, minValue);
        this.bigramMatrix = new GrowingSparseMatrix();
    }

    /**
     * {@inheritDoc}
     */
    public int getVectorLength() {
        return bigramMatrix.columns();
    }

    /**
     * {@inheritDoc}
     */
    public Set getWords() {
        return basis.keySet();
    }

    /**
     * {@inheritDoc}
     */
    public SparseDoubleVector getVector(String word) {
        int index = basis.getDimension(word);
        return (index < 0) ? null : bigramMatrix.getRowVector(index);
    }

    /**
     * Returns "BigramSpace".
     */
    public String getSpaceName() {
        return "BigramSpace";
    }

    /**
     * {@inheritDoc}
     */
    public void  processDocument(BufferedReader document) throws IOException {
        Queue bigramWindow = new ArrayDeque();

        Iterator documentTokens = IteratorFactory.tokenize(document);

        for (int i = 0; i < windowSize && documentTokens.hasNext(); ++i) {
            String word = documentTokens.next();
            int index = basis.getDimension(word);
            if (index >= 0)
                bigramWindow.offer(word);
        }

        while (!bigramWindow.isEmpty()) {
            if (documentTokens.hasNext()) {
                String word = documentTokens.next();
                int index = basis.getDimension(word);
                if (index >= 0)
                    bigramWindow.offer(word);
            }

            String term = bigramWindow.remove();
            int index1 = basis.getDimension(term);
            if (index1 < 0)
                continue;

            for (String other : bigramWindow) {
                int index2 = basis.getDimension(other);
                if (index2 < 0)
                    continue;
                bigramMatrix.add(index1, index2, 1.0);
            }
        }
    }

    /**
     * {@inheritDoc}
     */
    public void processSpace(Properties props) {
        filter.transform(bigramMatrix, bigramMatrix);
        basis.setReadOnly(true);
    }
}