Java LastCovered类代码示例

import org.cleartk.ml.feature.extractor.CleartkExtractor.LastCovered; //导入依赖的package包/类
public static XStream createXStream() {
	//define alias so the xml file can be read easier
	XStream xstream = new XStream();
	// org.cleartk.classifier.feature.*
	xstream.alias("TypePathExtractor", TypePathExtractor.class);
	xstream.alias("FeatureCollection", FeatureCollection.class);

	// org.cleartk.ml.feature.extractor.*
	xstream.alias("CleartkExtractor", CleartkExtractor.class);
	xstream.alias("CombinedExtractor1", CombinedExtractor1.class);
	xstream.alias("CoveredTextExtractor", CoveredTextExtractor.class);
	xstream.alias("DirectedDistanceExtractor", DirectedDistanceExtractor.class);
       xstream.alias("DistanceExtractor", DistanceExtractor.class);
       xstream.alias("FeatureExtractor1", FeatureExtractor1.class);
       xstream.alias("FeatureExtractor2", FeatureExtractor2.class);
       xstream.alias("NamedFeatureExtractor1", NamedFeatureExtractor1.class);
       xstream.alias("NamingExtractor1", NamingExtractor1.class);
       xstream.alias("RelativePositionExtractor", RelativePositionExtractor.class);
       xstream.alias("WhiteSpaceExtractor", WhiteSpaceExtractor.class);


	// within CleartkExtractor
	xstream.alias("Bag", Bag.class);
	xstream.alias("Preceding", Preceding.class);
	xstream.alias("Following", Following.class);
	xstream.alias("Covered", Covered.class);
	xstream.alias("FirstCovered", FirstCovered.class);
	xstream.alias("LastCovered", LastCovered.class);
	xstream.alias("Ngram", Ngram.class);

	xstream.alias("list", ArrayList.class);
	return xstream;
}
 

import org.cleartk.ml.feature.extractor.CleartkExtractor.LastCovered; //导入依赖的package包/类
public static XStream createXStream()
{
    // define alias so the xml file can be read easier
    XStream xstream = new XStream();
    xstream.alias("list", ArrayList.class);

    xstream.alias("TypePathExtractor", TypePathExtractor.class);
    xstream.alias("FeatureCollection", FeatureCollection.class);


    xstream.alias("Bag", Bag.class);
    xstream.alias("Preceding", Preceding.class);
    xstream.alias("Following", Following.class);
    xstream.alias("Covered", Covered.class);
    xstream.alias("FirstCovered", FirstCovered.class);
    xstream.alias("LastCovered", LastCovered.class);
    xstream.alias("Ngram", Ngram.class);

    xstream.alias("CleartkExtractor", CleartkExtractor.class);
    xstream.alias("Covered", Covered.class);
    xstream.alias("Following", Following.class);
    xstream.alias("Preceding", Preceding.class);
    xstream.alias("CoveredTextExtractor", CoveredTextExtractor.class);
    xstream.alias("FeatureExtractor1", FeatureExtractor1.class);
    xstream.alias("TypePathExtractor", TypePathExtractor.class);
    xstream.alias("CapitalTypeFeatureFunction", CapitalTypeFeatureFunction.class);
    xstream.alias("CharacterNgramFeatureFunction", CharacterNgramFeatureFunction.class);
    xstream.alias("FeatureFunctionExtractor", FeatureFunctionExtractor.class);
    xstream.alias("LowerCaseFeatureFunction", LowerCaseFeatureFunction.class);
    xstream.alias("NumericTypeFeatureFunction", NumericTypeFeatureFunction.class);

    return xstream;
}
 

import org.cleartk.ml.feature.extractor.CleartkExtractor.LastCovered; //导入依赖的package包/类
@Test
public void testBasic() throws Exception {
  CleartkExtractor<Chunk, Token> extractor = new CleartkExtractor<Chunk, Token>(
      Token.class,
      new CoveredTextExtractor<Token>(),
      new Preceding(2),
      new Preceding(3, 6),
      new Covered(),
      new FirstCovered(1),
      new FirstCovered(1, 3),
      new LastCovered(1),
      new LastCovered(1, 3),
      new Following(1, 3),
      new Following(3, 5),
      new Preceding(5, 6),
      new Following(5, 6));

  this.tokenBuilder.buildTokens(
      this.jCas,
      "The quick brown fox jumped over the lazy dog.",
      "The quick brown fox jumped over the lazy dog .",
      "DT JJ JJ NN VBD IN DT JJ NN .");
  Chunk chunk = new Chunk(this.jCas, 20, 31);
  chunk.addToIndexes();
  Assert.assertEquals("jumped over", chunk.getCoveredText());

  List<Feature> features = extractor.extract(this.jCas, chunk);
  Assert.assertEquals(19, features.size());
  Iterator<Feature> iter = features.iterator();
  this.assertFeature("Preceding_0_2_1", "brown", iter.next());
  this.assertFeature("Preceding_0_2_0", "fox", iter.next());
  this.assertFeature("Preceding_3_6_5", "OOB2", iter.next());
  this.assertFeature("Preceding_3_6_4", "OOB1", iter.next());
  this.assertFeature("Preceding_3_6_3", "The", iter.next());
  this.assertFeature("Covered_0", "jumped", iter.next());
  this.assertFeature("Covered_1", "over", iter.next());
  this.assertFeature("FirstCovered_0_1_0", "jumped", iter.next());
  this.assertFeature("FirstCovered_1_3_1", "over", iter.next());
  this.assertFeature("FirstCovered_1_3_2", "OOB1", iter.next());
  this.assertFeature("LastCovered_0_1_0", "over", iter.next());
  this.assertFeature("LastCovered_1_3_2", "OOB1", iter.next());
  this.assertFeature("LastCovered_1_3_1", "jumped", iter.next());
  this.assertFeature("Following_1_3_1", "lazy", iter.next());
  this.assertFeature("Following_1_3_2", "dog", iter.next());
  this.assertFeature("Following_3_5_3", ".", iter.next());
  this.assertFeature("Following_3_5_4", "OOB1", iter.next());
  this.assertFeature("Preceding_5_6_5", "OOB2", iter.next());
  this.assertFeature("Following_5_6_5", "OOB2", iter.next());
}
 

import org.cleartk.ml.feature.extractor.CleartkExtractor.LastCovered; //导入依赖的package包/类
@Test
public void testBag() throws Exception {
  CleartkExtractor<Chunk, Token> extractor = new CleartkExtractor<Chunk, Token>(
      Token.class,
      new TypePathExtractor<Token>(Token.class, "pos"),
      new Bag(new Preceding(2)),
      new Bag(new Preceding(3, 6)),
      new Bag(new FirstCovered(1), new LastCovered(1)),
      new Bag(new Following(1, 3)),
      new Bag(new Following(3, 5)),
      new Bag(new Preceding(1), new Following(1)),
      new Bag(new Bag(new Bag(new Following(1, 3)))));

  this.tokenBuilder.buildTokens(
      this.jCas,
      "The quick brown fox jumped over the lazy dog.",
      "The quick brown fox jumped over the lazy dog .",
      "DT JJ JJ NN VBD IN DT JJ NN .");
  Chunk chunk = new Chunk(this.jCas, 20, 31);
  chunk.addToIndexes();
  Assert.assertEquals("jumped over", chunk.getCoveredText());

  List<Feature> features = extractor.extract(this.jCas, chunk);
  Assert.assertEquals(15, features.size());
  Iterator<Feature> iter = features.iterator();
  this.assertFeature("Bag_Preceding_0_2_TypePath(Pos)", "JJ", iter.next());
  this.assertFeature("Bag_Preceding_0_2_TypePath(Pos)", "NN", iter.next());
  this.assertFeature("Bag_Preceding_3_6_TypePath(Pos)", "OOB2", iter.next());
  this.assertFeature("Bag_Preceding_3_6_TypePath(Pos)", "OOB1", iter.next());
  this.assertFeature("Bag_Preceding_3_6_TypePath(Pos)", "DT", iter.next());
  this.assertFeature("Bag_FirstCovered_0_1_LastCovered_0_1_TypePath(Pos)", "VBD", iter.next());
  this.assertFeature("Bag_FirstCovered_0_1_LastCovered_0_1_TypePath(Pos)", "IN", iter.next());
  this.assertFeature("Bag_Following_1_3_TypePath(Pos)", "JJ", iter.next());
  this.assertFeature("Bag_Following_1_3_TypePath(Pos)", "NN", iter.next());
  this.assertFeature("Bag_Following_3_5_TypePath(Pos)", ".", iter.next());
  this.assertFeature("Bag_Following_3_5_TypePath(Pos)", "OOB1", iter.next());
  this.assertFeature("Bag_Preceding_0_1_Following_0_1_TypePath(Pos)", "NN", iter.next());
  this.assertFeature("Bag_Preceding_0_1_Following_0_1_TypePath(Pos)", "DT", iter.next());
  this.assertFeature("Bag_Bag_Bag_Following_1_3_TypePath(Pos)", "JJ", iter.next());
  this.assertFeature("Bag_Bag_Bag_Following_1_3_TypePath(Pos)", "NN", iter.next());
}
 

import org.cleartk.ml.feature.extractor.CleartkExtractor.LastCovered; //导入依赖的package包/类
@Test
public void testNgram() throws Exception {
  CleartkExtractor<Chunk, Token> extractor = new CleartkExtractor<Chunk, Token>(
      Token.class,
      new CoveredTextExtractor<Token>(),
      new Ngram(new Preceding(2)),
      new Ngram(new Preceding(3, 6)),
      new Ngram(new Preceding(1), new FirstCovered(1), new LastCovered(1)),
      new Ngram(new Following(1, 3)),
      new Ngram(new Following(3, 5)),
      new Ngram(new Preceding(2), new Following(1, 2)));

  this.tokenBuilder.buildTokens(
      this.jCas,
      "The quick brown fox jumped over the lazy dog.",
      "The quick brown fox jumped over the lazy dog .",
      "DT JJ JJ NN VBD IN DT JJ NN .");
  Chunk chunk = new Chunk(this.jCas, 20, 31);
  chunk.addToIndexes();
  Assert.assertEquals("jumped over", chunk.getCoveredText());

  List<Feature> features = extractor.extract(this.jCas, chunk);
  Assert.assertEquals(6, features.size());
  Iterator<Feature> iter = features.iterator();
  this.assertFeature("Ngram_Preceding_0_2", "brown_fox", iter.next());
  this.assertFeature("Ngram_Preceding_3_6", "OOB2_OOB1_The", iter.next());
  this.assertFeature(
      "Ngram_Preceding_0_1_FirstCovered_0_1_LastCovered_0_1",
      "fox_jumped_over",
      iter.next());
  this.assertFeature("Ngram_Following_1_3", "lazy_dog", iter.next());
  this.assertFeature("Ngram_Following_3_5", "._OOB1", iter.next());
  this.assertFeature("Ngram_Preceding_0_2_Following_1_2", "brown_fox_lazy", iter.next());

  extractor = new CleartkExtractor<Chunk, Token>(
      Token.class,
      new CoveredTextExtractor<Token>(),
      new Ngram(new Preceding(2), new Following(2)));

  jCas = JCasFactory.createJCas();
  this.tokenBuilder.buildTokens(this.jCas, "A B C D E");
  chunk = new Chunk(this.jCas, 4, 5);
  chunk.addToIndexes();
  Assert.assertEquals("C", chunk.getCoveredText());
  features = extractor.extract(this.jCas, chunk);
  Assert.assertEquals(1, features.size());
  this.assertFeature("Ngram_Preceding_0_2_Following_0_2", "A_B_D_E", features.get(0));

}
 

import org.cleartk.ml.feature.extractor.CleartkExtractor.LastCovered; //导入依赖的package包/类
@Test
public void testBounds() throws Exception {
  CleartkExtractor<Chunk, Token> extractor = new CleartkExtractor<Chunk, Token>(
      Token.class,
      new CoveredTextExtractor<Token>(),
      new Preceding(2),
      new LastCovered(1),
      new Following(3));

  this.tokenBuilder.buildTokens(
      this.jCas,
      "She bought milk.\nHe sold oranges.",
      "She bought milk .\nHe sold oranges .");
  Chunk boughMilk = new Chunk(this.jCas, 4, 15);
  boughMilk.addToIndexes();
  Assert.assertEquals("bought milk", boughMilk.getCoveredText());
  Chunk soldOranges = new Chunk(this.jCas, 20, 32);
  soldOranges.addToIndexes();
  Assert.assertEquals("sold oranges", soldOranges.getCoveredText());
  Collection<Sentence> sentences = JCasUtil.select(jCas, Sentence.class);
  Assert.assertEquals(2, sentences.size());
  Iterator<Sentence> sentIter = sentences.iterator();
  Sentence sent1 = sentIter.next();
  Sentence sent2 = sentIter.next();

  List<Feature> features = extractor.extractWithin(this.jCas, boughMilk, sent1);
  Assert.assertEquals(6, features.size());
  Iterator<Feature> iter = features.iterator();
  this.assertFeature("Preceding_0_2_1", "OOB1", iter.next());
  this.assertFeature("Preceding_0_2_0", "She", iter.next());
  this.assertFeature("LastCovered_0_1_0", "milk", iter.next());
  this.assertFeature("Following_0_3_0", ".", iter.next());
  this.assertFeature("Following_0_3_1", "OOB1", iter.next());
  this.assertFeature("Following_0_3_2", "OOB2", iter.next());

  features = extractor.extractWithin(this.jCas, boughMilk, sent2);
  Assert.assertEquals(6, features.size());
  iter = features.iterator();
  this.assertFeature("Preceding_0_2_1", "OOB2", iter.next());
  this.assertFeature("Preceding_0_2_0", "OOB1", iter.next());
  this.assertFeature("LastCovered_0_1_0", "OOB1", iter.next());
  this.assertFeature("Following_0_3_0", "OOB1", iter.next());
  this.assertFeature("Following_0_3_1", "He", iter.next());
  this.assertFeature("Following_0_3_2", "sold", iter.next());

  features = extractor.extractWithin(this.jCas, soldOranges, sent2);
  Assert.assertEquals(6, features.size());
  iter = features.iterator();
  this.assertFeature("Preceding_0_2_1", "OOB1", iter.next());
  this.assertFeature("Preceding_0_2_0", "He", iter.next());
  this.assertFeature("LastCovered_0_1_0", "oranges", iter.next());
  this.assertFeature("Following_0_3_0", ".", iter.next());
  this.assertFeature("Following_0_3_1", "OOB1", iter.next());
  this.assertFeature("Following_0_3_2", "OOB2", iter.next());
}