use of org.apache.clerezza.commons.rdf.IRI in project stanbol by apache.
the class DBPSpotlightDisambiguateEnhancementEngine method getSpottedXml.
private String getSpottedXml(String text, Graph graph) {
StringBuilder xml = new StringBuilder();
textAnnotationsMap = new Hashtable<String, IRI>();
xml.append(String.format("<annotation text=\"%s\">", text));
try {
for (Iterator<Triple> it = graph.filter(null, RDF_TYPE, TechnicalClasses.ENHANCER_TEXTANNOTATION); it.hasNext(); ) {
// Triple tAnnotation = it.next();
IRI uri = (IRI) it.next().getSubject();
String surfaceForm = EnhancementEngineHelper.getString(graph, uri, ENHANCER_SELECTED_TEXT);
if (surfaceForm != null) {
String offset = EnhancementEngineHelper.getString(graph, uri, ENHANCER_START);
textAnnotationsMap.put(surfaceForm, uri);
xml.append(String.format("<surfaceForm name=\"%s\" offset=\"%s\"/>", surfaceForm, offset));
}
}
} catch (Exception e) {
log.error(e.getMessage());
}
return xml.append("</annotation>").toString();
}
use of org.apache.clerezza.commons.rdf.IRI in project stanbol by apache.
the class CeliSentimentAnalysisEngine method computeEnhancements.
@Override
public void computeEnhancements(ContentItem ci) throws EngineException {
Entry<IRI, Blob> contentPart = ContentItemHelper.getBlob(ci, SUPPORTED_MIMTYPES);
if (contentPart == null) {
throw new IllegalStateException("No ContentPart with Mimetype '" + TEXT_PLAIN_MIMETYPE + "' found for ContentItem " + ci.getUri() + ": This is also checked in the canEnhance method! -> This " + "indicated an Bug in the implementation of the " + "EnhancementJobManager!");
}
String text = "";
try {
text = ContentItemHelper.getText(contentPart.getValue());
} catch (IOException e) {
throw new InvalidContentException(this, ci, e);
}
if (text.trim().length() == 0) {
log.info("No text contained in ContentPart {" + contentPart.getKey() + "} of ContentItem {" + ci.getUri() + "}");
return;
}
String language = EnhancementEngineHelper.getLanguage(ci);
if (language == null) {
throw new IllegalStateException("Unable to extract Language for " + "ContentItem " + ci.getUri() + ": This is also checked in the canEnhance " + "method! -> This indicated an Bug in the implementation of the " + "EnhancementJobManager!");
}
// used for the palin literals in TextAnnotations
Language lang = new Language(language);
try {
List<SentimentExpression> lista = this.client.extractSentimentExpressions(text, language);
LiteralFactory literalFactory = LiteralFactory.getInstance();
Graph g = ci.getMetadata();
for (SentimentExpression se : lista) {
try {
IRI textAnnotation = EnhancementEngineHelper.createTextEnhancement(ci, this);
// add selected text as PlainLiteral in the language extracted from the text
g.add(new TripleImpl(textAnnotation, ENHANCER_SELECTED_TEXT, new PlainLiteralImpl(se.getSnippetStr(), lang)));
g.add(new TripleImpl(textAnnotation, DC_TYPE, CeliConstants.SENTIMENT_EXPRESSION));
if (se.getStartSnippet() != null && se.getEndSnippet() != null) {
g.add(new TripleImpl(textAnnotation, ENHANCER_START, literalFactory.createTypedLiteral(se.getStartSnippet().intValue())));
g.add(new TripleImpl(textAnnotation, ENHANCER_END, literalFactory.createTypedLiteral(se.getEndSnippet().intValue())));
g.add(new TripleImpl(textAnnotation, ENHANCER_SELECTION_CONTEXT, new PlainLiteralImpl(getSelectionContext(text, se.getSnippetStr(), se.getStartSnippet()), lang)));
g.add(new TripleImpl(textAnnotation, CeliConstants.HAS_SENTIMENT_EXPRESSION_POLARITY, literalFactory.createTypedLiteral(se.getSentimentPolarityAsDoubleValue())));
}
} catch (NoConvertorException e) {
log.error(e.getMessage(), e);
}
}
} catch (IOException e) {
throw new EngineException("Error while calling the CELI Sentiment Analysis service (configured URL: " + serviceURL + ")!", e);
} catch (SOAPException e) {
throw new EngineException("Error wile encoding/decoding the request/response to the CELI Sentiment Analysis service!", e);
}
}
use of org.apache.clerezza.commons.rdf.IRI in project stanbol by apache.
the class CeliLanguageIdentifierEnhancementEngineTest method tesetEngine.
@Test
public void tesetEngine() throws Exception {
ContentItem ci = wrapAsContentItem(TEXT);
try {
langIdentifier.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(langIdentifier.getClass().getName()));
int numTextAnnotations = validateAllTextAnnotations(ci.getMetadata(), TEXT, expectedValues);
assertEquals("A single TextAnnotation is expected by this Test", 1, numTextAnnotations);
// even through this tests do not validate service quality but rather
// the correct integration of the CELI service as EnhancementEngine
// we expect the "fr" is detected for the parsed text
assertEquals("The detected language for text '" + TEXT + "' MUST BE 'fr'", "fr", EnhancementEngineHelper.getLanguage(ci));
int entityAnnoNum = validateAllEntityAnnotations(ci.getMetadata(), expectedValues);
assertEquals("No EntityAnnotations are expected", 0, entityAnnoNum);
} catch (EngineException e) {
RemoteServiceHelper.checkServiceUnavailable(e);
}
}
use of org.apache.clerezza.commons.rdf.IRI in project stanbol by apache.
the class CeliLemmatizerEnhancementEngineTest method validateMorphoFeatureProperty.
/**
* [1..*] values of an {@link TypedLiteral} in the form {key=value}
* @param enhancements The graph with the enhancements
* @param textAnnotation the TextAnnotation to check
*/
private void validateMorphoFeatureProperty(Graph enhancements, BlankNodeOrIRI textAnnotation) {
// This taste checks for known morpho features of a given input (constant TERM)
Iterator<Triple> morphoFeatureIterator = enhancements.filter(textAnnotation, RDF_TYPE, null);
assertTrue("No POS Morpho Feature value found for TextAnnotation " + textAnnotation + "!", morphoFeatureIterator.hasNext());
while (morphoFeatureIterator.hasNext()) {
RDFTerm morphoFeature = morphoFeatureIterator.next().getObject();
assertTrue("Morpho Feature value are expected of typed literal", morphoFeature instanceof IRI);
String feature = ((IRI) morphoFeature).getUnicodeString();
assertFalse("Morpho Feature MUST NOT be empty", feature.isEmpty());
if (feature.startsWith(OLIA_NAMESPACE)) {
String key = feature.substring(OLIA_NAMESPACE.length());
LexicalCategory cat = LexicalCategory.valueOf(key);
assertTrue("Part of Speech of " + TERM + " should be " + LexicalCategory.Noun, (cat == LexicalCategory.Noun));
}
}
morphoFeatureIterator = enhancements.filter(textAnnotation, CeliMorphoFeatures.HAS_GENDER, null);
assertTrue("No Gender Morpho Feature value found for TextAnnotation " + textAnnotation + "!", morphoFeatureIterator.hasNext());
if (morphoFeatureIterator.hasNext()) {
RDFTerm morphoFeature = morphoFeatureIterator.next().getObject();
assertTrue("Morpho Feature value are expected of typed literal", morphoFeature instanceof IRI);
String feature = ((IRI) morphoFeature).getUnicodeString();
assertFalse("Morpho Feature MUST NOT be empty", feature.isEmpty());
if (feature.startsWith(OLIA_NAMESPACE)) {
String key = feature.substring(OLIA_NAMESPACE.length());
Gender cat = Gender.valueOf(key);
assertTrue("Gender of " + TERM + " should be " + Gender.Feminine, (cat == Gender.Feminine));
}
}
morphoFeatureIterator = enhancements.filter(textAnnotation, CeliMorphoFeatures.HAS_NUMBER, null);
assertTrue("No Number Morpho Feature value found for TextAnnotation " + textAnnotation + "!", morphoFeatureIterator.hasNext());
if (morphoFeatureIterator.hasNext()) {
RDFTerm morphoFeature = morphoFeatureIterator.next().getObject();
assertTrue("Morpho Feature value are expected of typed literal", morphoFeature instanceof IRI);
String feature = ((IRI) morphoFeature).getUnicodeString();
assertFalse("Morpho Feature MUST NOT be empty", feature.isEmpty());
if (feature.startsWith(OLIA_NAMESPACE)) {
String key = feature.substring(OLIA_NAMESPACE.length());
NumberFeature cat = NumberFeature.valueOf(key);
assertTrue("Number of " + TERM + " should be " + Gender.Feminine, (cat == NumberFeature.Singular));
}
}
morphoFeatureIterator = enhancements.filter(textAnnotation, CeliLemmatizerEnhancementEngine.hasLemmaForm, null);
assertTrue("No Number Morpho Feature value found for TextAnnotation " + textAnnotation + "!", morphoFeatureIterator.hasNext());
if (morphoFeatureIterator.hasNext()) {
RDFTerm morphoFeature = morphoFeatureIterator.next().getObject();
assertTrue("Lemma Forms value are expected of type Literal", morphoFeature instanceof Literal);
assertFalse("Lemma forms MUST NOT be empty", ((Literal) morphoFeature).getLexicalForm().isEmpty());
String feature = ((Literal) morphoFeature).getLexicalForm();
assertTrue("Lemma of " + TERM + " should be " + TERM, (feature.equals(TERM)));
}
}
use of org.apache.clerezza.commons.rdf.IRI in project stanbol by apache.
the class CeliLemmatizerEnhancementEngineTest method testCompleteMorphoAnalysis.
@Test
public void testCompleteMorphoAnalysis() throws Exception {
ContentItem ci = wrapAsContentItem(TERM);
// 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")));
CeliLemmatizerEnhancementEngine morphoAnalysisEngine = initEngine(true);
try {
morphoAnalysisEngine.computeEnhancements(ci);
} catch (EngineException e) {
RemoteServiceHelper.checkServiceUnavailable(e);
return;
}
TestUtils.logEnhancements(ci);
// validate enhancements
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> textAnnotationIterator = ci.getMetadata().filter(null, RDF_TYPE, ENHANCER_TEXTANNOTATION);
// test if a textAnnotation is present
// assertTrue(textAnnotationIterator.hasNext());
// -> this might be used to test that there are no TextAnnotations
int textAnnotationCount = 0;
while (textAnnotationIterator.hasNext()) {
IRI textAnnotation = (IRI) textAnnotationIterator.next().getSubject();
// test if selected Text is added
validateTextAnnotation(ci.getMetadata(), textAnnotation, TERM, expectedValues);
textAnnotationCount++;
// perform additional tests for "hasMorphologicalFeature" and "hasLemmaForm"
validateMorphoFeatureProperty(ci.getMetadata(), textAnnotation);
}
log.info("{} TextAnnotations found and validated ...", textAnnotationCount);
int entityAnnoNum = validateAllEntityAnnotations(ci.getMetadata(), expectedValues);
// no EntityAnnotations expected
Assert.assertEquals("No EntityAnnotations expected by this test", 0, entityAnnoNum);
shutdownEngine(morphoAnalysisEngine);
}
Aggregations