Examples with Feature org.dkpro.tc.api.features.Feature used on opensource projects

Example 81 with Feature

use of org.dkpro.tc.api.features.Feature in project dkpro-tc by dkpro.

the class LuceneNGramCPFE method extract.

@Override
public Set<Feature> extract(JCas view1, JCas view2) throws TextClassificationException {
    FrequencyDistribution<String> view1Ngrams = null;
    FrequencyDistribution<String> view2Ngrams = null;
    TextClassificationTarget aTarget1 = JCasUtil.selectSingle(view1, TextClassificationTarget.class);
    TextClassificationTarget aTarget2 = JCasUtil.selectSingle(view2, TextClassificationTarget.class);
    view1Ngrams = NGramUtils.getDocumentNgrams(view1, aTarget1, ngramLowerCase, filterPartialStopwordMatches, ngramMinN1, ngramMaxN1, stopwords, Token.class);
    view2Ngrams = NGramUtils.getDocumentNgrams(view2, aTarget2, ngramLowerCase, filterPartialStopwordMatches, ngramMinN2, ngramMaxN2, stopwords, Token.class);
    FrequencyDistribution<String> documentComboNgrams = ComboUtils.getCombinedNgrams(view1Ngrams, view2Ngrams, ngramMinNCombo, ngramMaxNCombo, ngramUseSymmetricalCombos);
    Set<Feature> features = new HashSet<Feature>();
    prefix = "comboNG";
    features = addToFeatureArray(documentComboNgrams, topKSetCombo, features);
    return features;
}

Example 82 with Feature

use of org.dkpro.tc.api.features.Feature in project dkpro-tc by dkpro.

the class DiffNounChunkTokenLengthTest method extractTest1.

@Test
public void extractTest1() throws Exception {
    Chunk chunk1 = new Chunk(jcas1, 0, 4);
    chunk1.addToIndexes();
    Chunk chunk2 = new Chunk(jcas2, 0, 4);
    chunk2.addToIndexes();
    DiffNounChunkTokenLength extractor = new DiffNounChunkTokenLength();
    Set<Feature> features = extractor.extract(jcas1, jcas2);
    assertEquals(1, features.size());
    for (Feature feature : features) {
        assertFeature("DiffNounPhraseTokenLength", 0.0, feature, 0.0001);
    }
}

Example 83 with Feature

use of org.dkpro.tc.api.features.Feature in project dkpro-tc by dkpro.

the class DiffNounChunkTokenLengthTest method extractTest2.

@Test
public void extractTest2() throws Exception {
    Chunk chunk1 = new Chunk(jcas1, 0, 4);
    chunk1.addToIndexes();
    Chunk chunk2 = new Chunk(jcas2, 0, 7);
    chunk2.addToIndexes();
    DiffNounChunkTokenLength extractor = new DiffNounChunkTokenLength();
    Set<Feature> features = extractor.extract(jcas1, jcas2);
    assertEquals(1, features.size());
    for (Feature feature : features) {
        assertFeature("DiffNounPhraseTokenLength", -1.0, feature, 0.0001);
    }
}

Example 84 with Feature

use of org.dkpro.tc.api.features.Feature in project dkpro-tc by dkpro.

the class DiffNrOfTokensPairFeatureExtractorTest method testExtract.

@Test
public void testExtract() throws ResourceInitializationException, AnalysisEngineProcessException, TextClassificationException {
    AnalysisEngineDescription desc = createEngineDescription(BreakIteratorSegmenter.class);
    AnalysisEngine engine = createEngine(desc);
    JCas jcas1 = engine.newJCas();
    jcas1.setDocumentLanguage("en");
    jcas1.setDocumentText("This is the text of view 1. And some more.");
    engine.process(jcas1);
    JCas jcas2 = engine.newJCas();
    jcas2.setDocumentLanguage("en");
    jcas2.setDocumentText("This is the text of view 2.");
    engine.process(jcas2);
    DiffNrOfTokensPairFeatureExtractor extractor = new DiffNrOfTokensPairFeatureExtractor();
    Set<Feature> features = extractor.extract(jcas1, jcas2);
    assertEquals(1, features.size());
    for (Feature feature : features) {
        assertFeature("DiffNrOfTokens", 4, feature);
    }
}

Example 85 with Feature

use of org.dkpro.tc.api.features.Feature in project dkpro-tc by dkpro.

the class NGramCPPipelineTest method testNonBinaryFeatureValues.

// TODO: Write a symmetry test. Note that features will be the same. Needs different dataset.
@Test
public void testNonBinaryFeatureValues() throws Exception {
    NGramCPPipelineTest test = new NGramCPPipelineTest();
    test.initialize();
    test.parameters = new Object[] { LuceneNGramCPFE.PARAM_UNIQUE_EXTRACTOR_NAME, "123", LuceneNGramCPFE.PARAM_USE_VIEW1_NGRAMS_AS_FEATURES, false, LuceneNGramCPFE.PARAM_USE_VIEW2_NGRAMS_AS_FEATURES, false, LuceneNGramCPFE.PARAM_USE_VIEWBLIND_NGRAMS_AS_FEATURES, false, LuceneNGramCPFE.PARAM_SOURCE_LOCATION, test.lucenePath, LuceneNGramPMetaCollector.PARAM_TARGET_LOCATION, test.lucenePath, LuceneNGramCPFE.PARAM_NGRAM_BINARY_FEATURE_VALUES_COMBO, false, LuceneNGramCPFE.PARAM_NGRAM_MAX_N_COMBO, 2, LuceneNGramCPFE.PARAM_NGRAM_SYMMETRY_COMBO, true };
    test.runPipeline();
    int two = 0;
    int one = 0;
    int zero = 0;
    for (Feature feature : test.instanceList.get(0).getFeatures()) {
        Integer value = ((Double) feature.getValue()).intValue();
        if (new Integer(value) == 2) {
            two++;
        }
        if (new Integer(value) == 1) {
            one++;
        }
        if (new Integer(value) == 0) {
            zero++;
        }
    }
    assertEquals(two, 4);
    assertEquals(one, 12);
    assertEquals(zero, 0);
}