Examples with RDFTerm org.apache.clerezza.commons.rdf.RDFTerm used on opensource projects

Example 26 with RDFTerm

use of org.apache.clerezza.commons.rdf.RDFTerm in project stanbol by apache.

the class CeliLemmatizerEnhancementEngineTest method testEngine.

@Test
public void testEngine() throws Exception {
    ContentItem ci = wrapAsContentItem(TEXT);
    //add a simple triple to statically define the language of the test
    //content
    ci.getMetadata().add(new TripleImpl(ci.getUri(), DC_LANGUAGE, new PlainLiteralImpl("it")));
    //unit test should not depend on each other (if possible)
    //CeliLanguageIdentifierEnhancementEngineTest.addEnanchements(ci);
    CeliLemmatizerEnhancementEngine morphoAnalysisEngine = initEngine(false);
    try {
        morphoAnalysisEngine.computeEnhancements(ci);
    } catch (EngineException e) {
        RemoteServiceHelper.checkServiceUnavailable(e);
        return;
    }
    TestUtils.logEnhancements(ci);
    //validate enhancement
    HashMap<IRI, RDFTerm> expectedValues = new HashMap<IRI, RDFTerm>();
    expectedValues.put(Properties.ENHANCER_EXTRACTED_FROM, ci.getUri());
    expectedValues.put(Properties.DC_CREATOR, LiteralFactory.getInstance().createTypedLiteral(morphoAnalysisEngine.getClass().getName()));
    Iterator<Triple> lemmaTextAnnotationIterator = ci.getMetadata().filter(null, RDF_TYPE, ENHANCER_TEXTANNOTATION);
    assertTrue("A TextAnnotation is expected by this Test", lemmaTextAnnotationIterator.hasNext());
    BlankNodeOrIRI lemmaTextAnnotation = lemmaTextAnnotationIterator.next().getSubject();
    assertTrue("TextAnnoations MUST BE IRIs!", lemmaTextAnnotation instanceof IRI);
    assertFalse("Only a single TextAnnotation is expected by this Test", lemmaTextAnnotationIterator.hasNext());
    //validate the enhancement metadata
    validateEnhancement(ci.getMetadata(), (IRI) lemmaTextAnnotation, expectedValues);
    //validate the lemma form TextAnnotation
    int lemmaForms = validateLemmaFormProperty(ci.getMetadata(), lemmaTextAnnotation, "it");
    assertTrue("Only a single LemmaForm property is expected if '" + MORPHOLOGICAL_ANALYSIS + "=false'", lemmaForms == 1);
    shutdownEngine(morphoAnalysisEngine);
}

Example 27 with RDFTerm

use of org.apache.clerezza.commons.rdf.RDFTerm in project stanbol by apache.

the class CeliNamedEntityExtractionEnhancementEngineTest method testInput.

private void testInput(String txt, String lang) throws EngineException, IOException {
    ContentItem ci = wrapAsContentItem(txt);
    try {
        //add a simple triple to statically define the language of the test content
        ci.getMetadata().add(new TripleImpl(ci.getUri(), DC_LANGUAGE, new PlainLiteralImpl(lang)));
        nerEngine.computeEnhancements(ci);
        TestUtils.logEnhancements(ci);
        HashMap<IRI, RDFTerm> expectedValues = new HashMap<IRI, RDFTerm>();
        expectedValues.put(Properties.ENHANCER_EXTRACTED_FROM, ci.getUri());
        expectedValues.put(Properties.DC_CREATOR, LiteralFactory.getInstance().createTypedLiteral(nerEngine.getClass().getName()));
        int textAnnoNum = validateAllTextAnnotations(ci.getMetadata(), txt, expectedValues);
        log.info(textAnnoNum + " TextAnnotations found ...");
        int entityAnnoNum = EnhancementStructureHelper.validateAllEntityAnnotations(ci.getMetadata(), expectedValues);
        log.info(entityAnnoNum + " EntityAnnotations found ...");
    } catch (EngineException e) {
        RemoteServiceHelper.checkServiceUnavailable(e);
    }
}

Example 28 with RDFTerm

use of org.apache.clerezza.commons.rdf.RDFTerm in project stanbol by apache.

the class DBPSpotlightDisambiguateEnhancementEngine method createEnhancements.

/**
	 * The method adds the returned DBpedia Spotlight annotations to the content
	 * item's metadata. For each DBpedia resource an EntityAnnotation is created
	 * and linked to the according TextAnnotation.
	 * 
	 * @param occs
	 *            a Collection of entity information
	 * @param ci
	 *            the content item
	 */
public void createEnhancements(Collection<Annotation> occs, ContentItem ci, Language language) {
    HashMap<RDFTerm, IRI> entityAnnotationMap = new HashMap<RDFTerm, IRI>();
    for (Annotation occ : occs) {
        if (textAnnotationsMap.get(occ.surfaceForm) != null) {
            IRI textAnnotation = textAnnotationsMap.get(occ.surfaceForm);
            Graph model = ci.getMetadata();
            IRI entityAnnotation = EnhancementEngineHelper.createEntityEnhancement(ci, this);
            entityAnnotationMap.put(occ.uri, entityAnnotation);
            Literal label = new PlainLiteralImpl(occ.surfaceForm.name, language);
            model.add(new TripleImpl(entityAnnotation, DC_RELATION, textAnnotation));
            model.add(new TripleImpl(entityAnnotation, ENHANCER_ENTITY_LABEL, label));
            Collection<String> t = occ.getTypeNames();
            if (t != null) {
                Iterator<String> it = t.iterator();
                while (it.hasNext()) model.add(new TripleImpl(entityAnnotation, ENHANCER_ENTITY_TYPE, new IRI(it.next())));
            }
            model.add(new TripleImpl(entityAnnotation, ENHANCER_ENTITY_REFERENCE, occ.uri));
        }
    }
}

Example 29 with RDFTerm

use of org.apache.clerezza.commons.rdf.RDFTerm in project stanbol by apache.

the class TestNamedEntityExtractionEnhancementEngine method testCustomModel.

@Test
public void testCustomModel() throws EngineException, IOException {
    ContentItem ci = wrapAsContentItem("urn:test:content-item:single:sentence", EHEALTH, "en");
    //this test does not use default models
    nerEngine.config.getDefaultModelTypes().clear();
    //but instead a custom model provided by the test data
    nerEngine.config.addCustomNameFinderModel("en", "bionlp2004-DNA-en.bin");
    nerEngine.config.setMappedType("DNA", new IRI("http://www.bootstrep.eu/ontology/GRO#DNA"));
    nerEngine.computeEnhancements(ci);
    Map<IRI, RDFTerm> expectedValues = new HashMap<IRI, RDFTerm>();
    expectedValues.put(Properties.ENHANCER_EXTRACTED_FROM, ci.getUri());
    expectedValues.put(Properties.DC_CREATOR, LiteralFactory.getInstance().createTypedLiteral(nerEngine.getClass().getName()));
    //adding null as expected for confidence makes it a required property
    expectedValues.put(Properties.ENHANCER_CONFIDENCE, null);
    //and dc:type values MUST be the URI set as mapped type
    expectedValues.put(Properties.DC_TYPE, new IRI("http://www.bootstrep.eu/ontology/GRO#DNA"));
    Graph g = ci.getMetadata();
    int textAnnotationCount = validateAllTextAnnotations(g, EHEALTH, expectedValues);
    assertEquals(7, textAnnotationCount);
}

Example 30 with RDFTerm

use of org.apache.clerezza.commons.rdf.RDFTerm in project stanbol by apache.

the class TestMetaxaCore method printTriples.

/**
     * This prints out the Stanbol Enhancer triples that would be created for the metadata
     * contained in the given model.
     *
     * @param m a {@link Model}
     *
     * @return an {@code int} with the number of added triples
     */
private int printTriples(Model m) {
    int tripleCounter = 0;
    HashMap<BlankNode, BlankNode> blankNodeMap = new HashMap<BlankNode, BlankNode>();
    ClosableIterator<Statement> it = m.iterator();
    while (it.hasNext()) {
        Statement oneStmt = it.next();
        BlankNodeOrIRI subject = (BlankNodeOrIRI) MetaxaEngine.asClerezzaResource(oneStmt.getSubject(), blankNodeMap);
        IRI predicate = (IRI) MetaxaEngine.asClerezzaResource(oneStmt.getPredicate(), blankNodeMap);
        RDFTerm object = MetaxaEngine.asClerezzaResource(oneStmt.getObject(), blankNodeMap);
        if (null != subject && null != predicate && null != object) {
            Triple t = new TripleImpl(subject, predicate, object);
            LOG.debug("adding " + t);
            tripleCounter++;
        } else {
            LOG.debug("skipped " + oneStmt.toString());
        }
    }
    it.close();
    return tripleCounter;
}