Java Target2LabelSequence类代码示例

import cc.mallet.pipe.Target2LabelSequence; //导入依赖的package包/类
private Pipe makePipe() {
    Alphabet alpha = new Alphabet();
    Target2LabelSequence labelPipe = new Target2LabelSequence();
    LabelAlphabet labelAlpha = (LabelAlphabet) labelPipe.getTargetAlphabet();

    return new SerialPipes(ImmutableList.of(
        new SWordConverterPipe(),
        new StringListToTokenSequence(alpha, labelAlpha),   // convert to token sequence
        new TokenSequenceLowercase(),                       // make all lowercase
        new PhoneNeighborPipe(true, makeNeighbors()),         // grab neighboring graphemes
        new PhoneClassPipe(true, makeClassNeighbors()),
        new VowelNeighborPipe(),
//          new SurroundingTokenFeature(false),
//          new SurroundingTokenFeature(true),
//          new NeighborShapeFeature(true, makeShapeNeighs()),
        new IsFirstPipe(),
        new ThisPhoneClassPipe(),
//        new AppendEndPipe(), // right before TS2F to get text set, last not to mess w neighbors
        new TokenSequenceToFeature(),                       // convert the strings in the text to features
        new TokenSequence2FeatureVectorSequence(alpha, true, false),
        labelPipe
    ));
  }
 

import cc.mallet.pipe.Target2LabelSequence; //导入依赖的package包/类
private Pipe makePipe() {
    Alphabet alpha = new Alphabet();
    Target2LabelSequence labelPipe = new Target2LabelSequence();
    LabelAlphabet labelAlpha = (LabelAlphabet) labelPipe.getTargetAlphabet();

    return new SerialPipes(ImmutableList.of(
        new StringListToTokenSequence(alpha, labelAlpha),   // convert to token sequence
        new TokenSequenceLowercase(),                       // make all lowercase
        new NeighborTokenFeature(true, makeNeighbors()),         // grab neighboring graphemes
        new SurroundingTokenFeature(false),
//        new SurroundingTokenFeature(true),
        new NeighborShapeFeature(true, makeShapeNeighs()),
        new LeadingTrailingFeature(),
        new TokenSequenceToFeature(),                       // convert the strings in the text to features
        new TokenSequence2FeatureVectorSequence(alpha, true, true),
        labelPipe
    ));
  }
 

import cc.mallet.pipe.Target2LabelSequence; //导入依赖的package包/类
private Pipe makePipe() {
  Alphabet alpha = new Alphabet();
  Target2LabelSequence labelPipe = new Target2LabelSequence();
  LabelAlphabet labelAlpha = (LabelAlphabet) labelPipe.getTargetAlphabet();

  return new SerialPipes(ImmutableList.of(
      new StringListToTokenSequence(alpha, labelAlpha),   // convert to token sequence
      new TokenSequenceLowercase(),                       // make all lowercase
      new NeighborTokenFeature(true, makeNeighbors()),         // grab neighboring graphemes
      new SurroundingTokenFeature(false),
      new SurroundingTokenFeature(true),
      new NeighborShapeFeature(true, makeShapeNeighs()),
      new LeadingTrailingFeature(),
      new TokenSequenceToFeature(),                       // convert the strings in the text to features
      new TokenSequence2FeatureVectorSequence(alpha, true, false),
      labelPipe
  ));
}
 

import cc.mallet.pipe.Target2LabelSequence; //导入依赖的package包/类
private static Pipe makePipe() {
  Alphabet alpha = new Alphabet();
  Target2LabelSequence labelPipe = new Target2LabelSequence();
  LabelAlphabet labelAlpha = (LabelAlphabet) labelPipe.getTargetAlphabet();

  return new SerialPipes(ImmutableList.of(
      new StringListToTokenSequence(alpha, labelAlpha),   // convert to token sequence
      new TokenSequenceLowercase(),                       // make all lowercase
      new NeighborTokenFeature(true, makeNeighbors()),         // grab neighboring graphemes
      new NeighborShapeFeature(true, makeShapeNeighs()),
      new TokenSequenceToFeature(),                       // convert the strings in the text to features
      new TokenSequence2FeatureVectorSequence(alpha, true, true),
      labelPipe,
      new LabelSequenceToLabelsAssignment(alpha, labelAlpha)
  ));
}
 

import cc.mallet.pipe.Target2LabelSequence; //导入依赖的package包/类
private SerialPipes buildSerialPipes(List<String> featureNames, List<String> replacements,
        List<String> conjunctions) throws LangDetectException, IOException {
    ArrayList<Pipe> pipes = new ArrayList<Pipe>();
    pipes.add(new LineGroupString2TokenSequence());
    pipes.add(new AddTargetToLinePipe(6));
    pipes.add(new LineToTargetTextPipe());
    pipes.add(new TargetReplacementPipe(replacements));

    FeaturePipeProvider featurePipeProvider = new FeaturePipeProvider();
    for (String featureName : featureNames) {
        pipes.add(featurePipeProvider.getPipe(featureName));
    }

    int[][] offsetConjunctions = new int[conjunctions.size()][];
    for (int i = 0; i < conjunctions.size(); i++) {
        String conjunction = conjunctions.get(i).replaceAll("min", "-");
        String[] conjunctionElements = conjunction.split(";");
        int[] conjunctionArray = new int[conjunctionElements.length];
        for (int j = 0; j < conjunctionElements.length; j++) {
            conjunctionArray[j] = Integer.parseInt(conjunctionElements[j]);
        }
        offsetConjunctions[i] = conjunctionArray;
    }
    pipes.add(new OffsetConjunctions(offsetConjunctions));

    pipes.add(new TokenSequence2FeatureVectorSequence(false, false));
    pipes.add(new Target2LabelSequence());

    // pipes.add(new PrintInputAndTarget());

    return new SerialPipes(pipes);

}
 

import cc.mallet.pipe.Target2LabelSequence; //导入依赖的package包/类
private SerialPipes makePipe(Alphabet alpha) {
    Target2LabelSequence labelPipe = new Target2LabelSequence();
    LabelAlphabet labelAlpha = (LabelAlphabet) labelPipe.getTargetAlphabet();

    return new SerialPipes(ImmutableList.of(
        new AlignmentToTokenSequence(alpha, labelAlpha, true, true, false),   // convert to token sequence
        new TokenSequenceLowercase(),                       // make all lowercase
        new NeighborTokenFeature(true, makeNeighbors()),         // grab neighboring graphemes
        new NeighborShapeFeature(true, makeShapeNeighs()),
//        new WindowFeature(false, 4),
//        new WindowFeature(true, 6),
        new NeighborSyllableFeature(-2, -1, 1, 2),
        new SyllCountingFeature(),
        new SyllCharRoleFeature(),
//        new NearSyllFeature(true),
//        new NearSyllFeature(false),
//        new SyllMarkingFeature(),
//        new SyllSequenceFeature(),
//        new SyllRelativeMarkFeature(),
        new EndingVowelFeature(),
        //new SonorityFeature2(true),
        //new SonorityFeature2(false),
//        new WindowFeature(false, 4),
        new VowelWindowFeature(2, 1, "PRESYL_", -1, false),
        new VowelWindowFeature(2, 1, "PSTSYL_", 1, false),
//        new VowelWindowFeature(3, 2, "LSTSYL_", 0, true),
        new SurroundingTokenFeature2(false, 1, 1),
//        new SurroundingTokenFeature2(true, 1, 1),
        new SurroundingTokenFeature2(false, 2, 2),
//        new SurroundingTokenFeature2(false, 3, 2),
        new SurroundingTokenFeature2(true, 3, 3),
//        new SurroundingTokenFeature2(true, 4, 4),
//        new LeadingTrailingFeature(),
        new TokenSequenceToFeature(),                       // convert the strings in the text to features
        new TokenSequence2FeatureVectorSequence(alpha, true, false),
        labelPipe
    ));
  }
 

import cc.mallet.pipe.Target2LabelSequence; //导入依赖的package包/类
public static List<Pipe> getPipes() throws Exception {

        List<Pipe> pipes = newArrayList();

        pipes.add(new Jcas2TokenSequence());
        pipes.add(new Target2LabelSequence());

        // more piiiiipes
        addAllGoodPipes(pipes);

        pipes.add(new FeatureWindow(window, window));
        // for debugging pipes.add(new PrintInputAndTarget());
        pipes.add(new TokenSequence2FeatureVectorSequence());
        return pipes;
    }