Example usage for opennlp.tools.languagemodel NGramLanguageModel predictNextTokens

Introduction

In this page you can find the example usage for opennlp.tools.languagemodel NGramLanguageModel predictNextTokens.

Prototype

@Override
    public String[] predictNextTokens(String... tokens) 

Source Link

Usage

From source file:opennlp.tools.languagemodel.NgramLanguageModelTest.java

@Test
public void testNgramModel() throws Exception {
    NGramLanguageModel model = new NGramLanguageModel(4);
    model.add(new StringList("I", "saw", "the", "fox"), 1, 4);
    model.add(new StringList("the", "red", "house"), 1, 4);
    model.add(new StringList("I", "saw", "something", "nice"), 1, 2);
    double probability = model.calculateProbability(new StringList("I", "saw", "the", "red", "house"));
    Assert.assertTrue("a probability measure should be between 0 and 1 [was " + probability + "]",
            probability >= 0 && probability <= 1);

    StringList tokens = model.predictNextTokens(new StringList("I", "saw"));
    Assert.assertNotNull(tokens);// w  w  w . j av a 2 s  .co  m
    Assert.assertEquals(new StringList("the", "fox"), tokens);
}

---

From source file:opennlp.tools.languagemodel.NgramLanguageModelTest.java

@Test
public void testTrigram() throws Exception {
    NGramLanguageModel model = new NGramLanguageModel(3);
    model.add(new StringList("I", "see", "the", "fox"), 2, 3);
    model.add(new StringList("the", "red", "house"), 2, 3);
    model.add(new StringList("I", "saw", "something", "nice"), 2, 3);
    double probability = model.calculateProbability(new StringList("I", "saw", "the", "red", "house"));
    Assert.assertTrue("a probability measure should be between 0 and 1 [was " + probability + "]",
            probability >= 0 && probability <= 1);

    StringList tokens = model.predictNextTokens(new StringList("I", "saw"));
    Assert.assertNotNull(tokens);/*w ww .j a  va2 s  .c  om*/
    Assert.assertEquals(new StringList("something", "nice"), tokens);
}

---

From source file:opennlp.tools.languagemodel.NgramLanguageModelTest.java

@Test
public void testBigram() throws Exception {
    NGramLanguageModel model = new NGramLanguageModel(2);
    model.add(new StringList("I", "see", "the", "fox"), 1, 2);
    model.add(new StringList("the", "red", "house"), 1, 2);
    model.add(new StringList("I", "saw", "something", "nice"), 1, 2);
    double probability = model.calculateProbability(new StringList("I", "saw", "the", "red", "house"));
    Assert.assertTrue("a probability measure should be between 0 and 1 [was " + probability + "]",
            probability >= 0 && probability <= 1);

    StringList tokens = model.predictNextTokens(new StringList("I", "saw"));
    Assert.assertNotNull(tokens);/* w w  w .  ja  v a  2 s .c  o m*/
    Assert.assertEquals(new StringList("something"), tokens);
}

---

From source file:opennlp.tools.languagemodel.NgramLanguageModelTest.java

@Test
public void testSerializedNGramLanguageModel() throws Exception {
    NGramLanguageModel languageModel = new NGramLanguageModel(
            getClass().getResourceAsStream("/opennlp/tools/ngram/ngram-model.xml"), 3);
    double probability = languageModel.calculateProbability(new StringList("The", "brown", "fox", "jumped"));
    Assert.assertTrue("a probability measure should be between 0 and 1 [was " + probability + "]",
            probability >= 0 && probability <= 1);
    StringList tokens = languageModel.predictNextTokens(new StringList("fox"));
    Assert.assertNotNull(tokens);//w ww . ja  v a 2s  .c  o  m
    Assert.assertEquals(new StringList("jumped"), tokens);
}

---

From source file:opennlp.tools.languagemodel.NgramLanguageModelTest.java

@Test
public void testTrigramLanguageModelCreationFromText() throws Exception {
    int ngramSize = 3;
    NGramLanguageModel languageModel = new NGramLanguageModel(ngramSize);
    InputStream stream = getClass().getResourceAsStream("/opennlp/tools/languagemodel/sentences.txt");
    for (String line : IOUtils.readLines(stream)) {
        String[] array = line.split(" ");
        List<String> split = Arrays.asList(array);
        List<String> generatedStrings = NGramGenerator.generate(split, ngramSize, " ");
        for (String generatedString : generatedStrings) {
            String[] tokens = generatedString.split(" ");
            if (tokens.length > 0) {
                languageModel.add(new StringList(tokens), 1, ngramSize);
            }//from   ww w  .j  a va2  s  .  c  om
        }
    }
    StringList tokens = languageModel.predictNextTokens(new StringList("neural", "network", "language"));
    Assert.assertNotNull(tokens);
    Assert.assertEquals(new StringList("models"), tokens);
    double p1 = languageModel.calculateProbability(new StringList("neural", "network", "language", "models"));
    double p2 = languageModel.calculateProbability(new StringList("neural", "network", "language", "model"));
    Assert.assertTrue(p1 > p2);
}

---