qmul.align.SentenceSyntacticSimilarityMeasure.java Source code

Introduction

Here is the source code for qmul.align.SentenceSyntacticSimilarityMeasure.java

Source

/*******************************************************************************
 * Copyright (c) 2013, 2014 Matthew Purver, Queen Mary University of London.
 * All rights reserved. This program and the accompanying materials
 * are made available under the terms of the GNU Public License v3.0
 * which accompanies this distribution, and is available at
 * http://www.gnu.org/licenses/gpl.html
 * 
 * Contributors:
 *     Matthew Purver, Queen Mary University of London - initial API and implementation
 ******************************************************************************/
package qmul.align;

import java.util.HashMap;

import qmul.corpus.DialogueSentence;
import qmul.corpus.DialogueTurn;
import qmul.util.parse.PennTreebankTokenizer;
import qmul.util.parse.StanfordParser;
import qmul.util.similarity.SimilarityMeasure;
import qmul.util.treekernel.TreeKernel;
import edu.stanford.nlp.trees.Tree;

public class SentenceSyntacticSimilarityMeasure implements SimilarityMeasure<DialogueSentence> {

    private int kernelType = TreeKernel.SYN_TREES;

    /**
     * use the default {@link TreeKernel} kernel type (syn rule subtrees)
     */
    public SentenceSyntacticSimilarityMeasure() {
        super();
        reset();
    }

    /**
     * @param kernelType
     *            specify the {@link TreeKernel} kernel type (subtrees/subset trees/syn rules)
     */
    public SentenceSyntacticSimilarityMeasure(int kernelType) {
        this();
        this.kernelType = kernelType;
    }

    /*
     * (non-Javadoc)
     * 
     * @see qmul.util.similarity.SimilarityMeasure#similarity(java.lang.Object, java.lang.Object)
     */
    @Override
    public double similarity(DialogueSentence a, DialogueSentence b) {
        Tree t1 = a.getSyntax();
        Tree t2 = b.getSyntax();
        if ((t1 == null) || (t2 == null)) {
            return 0.0;
        }
        return TreeKernel.resetAndCompute(t1, t2, kernelType);
    }

    /*
     * (non-Javadoc)
     * 
     * @see qmul.util.similarity.SimilarityMeasure#reset()
     */
    @Override
    public void reset() {
        //      TreeKernel.clearAllowedProductions();
        //      TreeKernel.clearBannedProductions();
    }

    /*
     * (non-Javadoc)
     * 
     * @see qmul.util.similarity.SimilarityMeasure#rawCountsA()
     */
    @Override
    public HashMap<? extends Object, Integer> rawCountsA() {
        return TreeKernel.getL1counts();
    }

    /*
     * (non-Javadoc)
     * 
     * @see qmul.util.similarity.SimilarityMeasure#rawCountsB()
     */
    @Override
    public HashMap<? extends Object, Integer> rawCountsB() {
        return TreeKernel.getL2counts();
    }

    /*
     * (non-Javadoc)
     * 
     * @see qmul.util.similarity.SimilarityMeasure#rawCountsAB()
     */
    @Override
    public HashMap<? extends Object, Integer> rawCountsAB() {
        return TreeKernel.getCommonCounts();
    }

    public static void main(String[] args) {
        SentenceSyntacticSimilarityMeasure sm = new SentenceSyntacticSimilarityMeasure();
        PennTreebankTokenizer tok = new PennTreebankTokenizer(true);
        StanfordParser p = new StanfordParser();
        // TreeParser p = new RASPParser();
        // String s1 = "Hello , how are you ?";
        // String s2 = "UNCLEAR what can I do for this lady today ?";
        String s1 = "You're full of catarrh.";
        String s2 = "Lot of wax in it, right enough.";
        p.parse(tok.getWordsFromString(s1));
        Tree t1 = p.getBestParse();
        p.parse(tok.getWordsFromString(s2));
        Tree t2 = p.getBestParse();
        System.out.println(s1);
        System.out.println(t1.pennString());
        System.out.println(s2);
        System.out.println(t2.pennString());
        System.out.println(TreeKernel.resetAndCompute(t1, t2, TreeKernel.SYN_TREES));
        System.out.println(TreeKernel.resetAndCompute(t1, t2, TreeKernel.SUB_TREES));
        System.out.println(TreeKernel.resetAndCompute(t1, t2, TreeKernel.SUBSET_TREES));
        // System.exit(0);

        DialogueTurn t = new DialogueTurn("t", 1, null, null);
        DialogueSentence a = new DialogueSentence("a", 1, t, "ok");
        DialogueSentence b = new DialogueSentence("b", 1, t, "ok");
        a.setSyntax(p.parse(a.getTranscription()));
        b.setSyntax(p.parse(b.getTranscription()));
        System.out.println("" + a + " " + a.getSyntax() + "\n" + b + " " + b.getSyntax() + "\n" + "sim = "
                + sm.similarity(a, b));
        a.setTranscription("ok ok ok ok");
        a.setSyntax(p.parse(a.getTranscription()));
        System.out.println("" + a + " " + a.getSyntax() + "\n" + b + " " + b.getSyntax() + "\n" + "sim = "
                + sm.similarity(a, b));
        a.setTranscription("that's really not ok");
        a.setSyntax(p.parse(a.getTranscription()));
        System.out.println("" + a + " " + a.getSyntax() + "\n" + b + " " + b.getSyntax() + "\n" + "sim = "
                + sm.similarity(a, b));
        b.setTranscription("that's really not ok");
        b.setSyntax(p.parse(b.getTranscription()));
        System.out.println("" + a + " " + a.getSyntax() + "\n" + b + " " + b.getSyntax() + "\n" + "sim = "
                + sm.similarity(a, b));
        a.setTranscription("john likes the small bear");
        a.setSyntax(p.parse(a.getTranscription()));
        b.setTranscription("jim hates the big rabbit");
        b.setSyntax(p.parse(b.getTranscription()));
        // TreeKernel.setIncludeWords(true);
        System.out.println("" + a + " " + a.getSyntax() + "\n" + b + " " + b.getSyntax() + "\n" + "sim = "
                + sm.similarity(a, b) + "\n" + sm.rawCountsA() + "\n" + sm.rawCountsB() + "\n" + sm.rawCountsAB());
        a.setTranscription("the man likes the small bear");
        a.setSyntax(p.parse(a.getTranscription()));
        // TreeKernel.setIncludeWords(true);
        System.out.println("" + a + " " + a.getSyntax() + "\n" + b + " " + b.getSyntax() + "\n" + "sim = "
                + sm.similarity(a, b) + "\n" + sm.rawCountsA() + "\n" + sm.rawCountsB() + "\n" + sm.rawCountsAB());
    }

}