Examples of org.apache.clerezza.rdf.core.LiteralFactory

org.apache.clerezza.rdf.core.LiteralFactory
This class provides methods to convert java objects to typed literals and vice versa. While the default implementation will provide literal objects storing the data's lexical form in memory, other implementations may create literal optimized for processing within the store. Note: this class uses the notion of "Convertor" (in the Exception naming), but does not currently provide a mechanism to register such Convertors. An implementation is said to provide Convertors for the types it supports. @since 0.3 @author reto

      throw new EngineException("Error wile encoding/decoding the request/" + "response to the CELI lemmatizer service!", e);
    }
    // get a write lock before writing the enhancements
    ci.getLock().writeLock().lock();
    try {
      LiteralFactory literalFactory = LiteralFactory.getInstance();
      for (LexicalEntry le : terms) {


        List<CeliMorphoFeatures> mFeatures = this.convertLexicalEntryToMorphFeatures(le, language);
        for (CeliMorphoFeatures feat : mFeatures) {
          // Create a text annotation for each interpretation produced by the morphological analyzer
          UriRef textAnnotation = EnhancementEngineHelper.createTextEnhancement(ci, this);
          g.add(new TripleImpl(textAnnotation, ENHANCER_SELECTED_TEXT, new PlainLiteralImpl(le.getWordForm(), lang)));
          if (le.from >= 0 && le.to > 0) {
            g.add(new TripleImpl(textAnnotation, ENHANCER_START, literalFactory.createTypedLiteral(le.from)));
            g.add(new TripleImpl(textAnnotation, ENHANCER_END, literalFactory.createTypedLiteral(le.to)));
            g.add(new TripleImpl(textAnnotation, ENHANCER_SELECTION_CONTEXT, new PlainLiteralImpl(getSelectionContext(text, le.getWordForm(), le.from), lang)));
          }
          g.addAll(feat.featuresAsTriples(textAnnotation, lang));
        }
      }

View Full Code Here

            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!");
        }
        Language lang = new Language(language); //used for the palin literals in TextAnnotations
    try {
      List<SentimentExpression> lista = this.client.extractSentimentExpressions(text, language);
      LiteralFactory literalFactory = LiteralFactory.getInstance();


      MGraph g = ci.getMetadata();


      for (SentimentExpression se : lista) {
        try {
          UriRef 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);
        }
      }

View Full Code Here

        return ENHANCE_SYNCHRONOUS;
    }


    public void computeEnhancements(ContentItem ci) throws EngineException {
        FeatureStructureListHolder holder;
        LiteralFactory literalFactory = LiteralFactory.getInstance();




        try {
            UriRef uimaUriRef = new UriRef(uimaUri);
            logger.info(new StringBuilder("Trying to load holder for ref:").append(uimaUri).toString());
            holder = ci.getPart(uimaUriRef, FeatureStructureListHolder.class);
            for (String source : sourceNames) {
                logger.info(new StringBuilder("Processing UIMA source:").append(source).toString());
                List<FeatureStructure> sourceList = holder.getFeatureStructureList(source);
                if (sourceList != null) {
                    logger.info(new StringBuilder("UIMA source:").append(source)
                            .append(" contains ").append(sourceList.size()).append(" annotations.").toString());
                } else {
                    logger.info(new StringBuilder("Source list is null:").append(source).toString());
                    continue;
                }
                for (FeatureStructure fs : sourceList) {
                    String typeName = fs.getTypeName();
                    logger.debug(new StringBuilder("Checking ").append(typeName).toString());
                    if (tnfs.checkFeatureStructureAllowed(typeName, fs.getFeatures())) {
                        logger.debug(new StringBuilder("Adding ").append(typeName).toString());
                        UriRef textAnnotation = EnhancementEngineHelper.createTextEnhancement(
                                ci, this);
                        MGraph metadata = ci.getMetadata();
                        String uriRefStr = uimaUri + ":" + typeName;
                        if (mappings.containsKey(typeName)) {
                            uriRefStr = mappings.get(typeName);
                        }
                        metadata.add(new TripleImpl(textAnnotation, DC_TYPE, new UriRef(uriRefStr)));


                        if (fs.getFeature("begin") != null) {
                            metadata.add(new TripleImpl(textAnnotation, ENHANCER_START,
                                    literalFactory.createTypedLiteral(fs.getFeature("begin").getValueAsInteger())));
                        }
                        if (fs.getFeature("end") != null) {
                            metadata.add(new TripleImpl(textAnnotation, ENHANCER_END,
                                    literalFactory.createTypedLiteral(fs.getFeature("end").getValueAsInteger())));
                        }
                        if (fs.getCoveredText() != null && !fs.getCoveredText().isEmpty()) {
                            metadata.add(new TripleImpl(textAnnotation, ENHANCER_SELECTED_TEXT, new PlainLiteralImpl(fs.getCoveredText())));
                        }

View Full Code Here

        }
        UriRef precision = new UriRef(NamespaceEnum.fise + "classifier/precision");
        UriRef recall = new UriRef(NamespaceEnum.fise + "classifier/recall");
        UriRef f1 = new UriRef(NamespaceEnum.fise + "classifier/f1");


        LiteralFactory lf = LiteralFactory.getInstance();
        ci.getLock().writeLock().lock();
        try {
            // Global text annotation to attach all the topic annotation to it.
            UriRef textAnnotation = EnhancementEngineHelper.createTextEnhancement(ci, this);
            metadata.add(new TripleImpl(textAnnotation,
                    org.apache.stanbol.enhancer.servicesapi.rdf.Properties.DC_TYPE,
                    OntologicalClasses.SKOS_CONCEPT));
            for (TopicSuggestion topic : topics) {
                UriRef enhancement = EnhancementEngineHelper.createEntityEnhancement(ci, this);
                metadata.add(new TripleImpl(enhancement,
                        org.apache.stanbol.enhancer.servicesapi.rdf.Properties.RDF_TYPE,
                        TechnicalClasses.ENHANCER_TOPICANNOTATION));
                metadata.add(new TripleImpl(enhancement,
                        org.apache.stanbol.enhancer.servicesapi.rdf.Properties.DC_RELATION, textAnnotation));


                // add link to entity
                metadata.add(new TripleImpl(enhancement,
                        org.apache.stanbol.enhancer.servicesapi.rdf.Properties.ENHANCER_ENTITY_REFERENCE,
                        new UriRef(topic.conceptUri)));
                metadata.add(new TripleImpl(enhancement,
                        org.apache.stanbol.enhancer.servicesapi.rdf.Properties.ENHANCER_ENTITY_TYPE,
                        OntologicalClasses.SKOS_CONCEPT));


                // add confidence information
                metadata.add(new TripleImpl(enhancement,
                        org.apache.stanbol.enhancer.servicesapi.rdf.Properties.ENHANCER_CONFIDENCE, lf
                                .createTypedLiteral(Double.valueOf(topic.score))));


                // add performance estimates of the classifier if available
                ClassificationReport perf = getPerformanceEstimates(topic.conceptUri);
                if (perf.uptodate) {
                    metadata.add(new TripleImpl(enhancement, precision, lf.createTypedLiteral(Double
                            .valueOf(perf.precision))));
                    metadata.add(new TripleImpl(enhancement, recall, lf.createTypedLiteral(Double
                            .valueOf(perf.recall))));
                    metadata.add(new TripleImpl(enhancement, f1, lf.createTypedLiteral(Double
                            .valueOf(perf.f1))));
                }
                // fetch concept label from the entityhub or a referenced site if available
                Entity entity = entityhub.getEntity(topic.conceptUri);
                if (entity == null) {

View Full Code Here

                new Mapping.Converter(){//we need to convert from MByte/min to kByte/sec
                    @Override
                    public Resource convert(Resource value) {
                        if(value instanceof TypedLiteral &&
                                XSD.double_.equals(((TypedLiteral)value).getDataType())){
                            LiteralFactory lf = LiteralFactory.getInstance();
                            double mm = lf.createObject(Double.class, (TypedLiteral)value);
                            return lf.createTypedLiteral(Double.valueOf(
                                mm*1024/60));
                        } else {
                            return value; //do not convert
                        }
                    }

View Full Code Here

        confidence
    }


    private void initOccurrences() {
        MGraph graph = contentItem.getMetadata();
        LiteralFactory lf = LiteralFactory.getInstance();
        Map<UriRef,Collection<NonLiteral>> suggestionMap = new HashMap<UriRef,Collection<NonLiteral>>();
        // 1) get Entity Annotations
        Map<NonLiteral,Map<EAProps,Object>> entitySuggestionMap = new HashMap<NonLiteral,Map<EAProps,Object>>();
        Iterator<Triple> entityAnnotations = graph.filter(null, RDF.type, ENHANCER_ENTITYANNOTATION);
        while(entityAnnotations.hasNext()){

View Full Code Here

    /**
     * @return an RDF/JSON descriptions of places for the word map widget
     */
    public String getPlacesAsJSON() throws ParseException, UnsupportedEncodingException {
        MGraph g = new IndexedMGraph();
        LiteralFactory lf = LiteralFactory.getInstance();
        MGraph metadata = contentItem.getMetadata();
        for (EntityExtractionSummary p : getPlaceOccurrences()) {
            EntitySuggestion bestGuess = p.getBestGuess();
            if (bestGuess == null) {
                continue;
            }
            UriRef uri = new UriRef(bestGuess.getUri());
            Iterator<Triple> latitudes = metadata.filter(uri, GEO_LAT, null);
            if (latitudes.hasNext()) {
                g.add(latitudes.next());
            }
            Iterator<Triple> longitutes = metadata.filter(uri, GEO_LONG, null);
            if (longitutes.hasNext()) {
                g.add(longitutes.next());
                g.add(new TripleImpl(uri, Properties.RDFS_LABEL, lf.createTypedLiteral(bestGuess.getLabel())));
            }
        }
        ByteArrayOutputStream out = new ByteArrayOutputStream();
        serializer.serialize(out, g, SupportedFormat.RDF_JSON);

View Full Code Here

        } else {
            language = null;
        }
        log.debug("findNamedEntities typeUri={}, type={}, text=", 
                new Object[]{ typeUri, typeLabel, StringUtils.abbreviate(text, 100) });
        LiteralFactory literalFactory = LiteralFactory.getInstance();
        MGraph g = ci.getMetadata();
        Map<String,List<NameOccurrence>> entityNames = extractNameOccurrences(nameFinderModel, text);
        //lock the ContentItem while writing the RDF data for found Named Entities
        ci.getLock().writeLock().lock();
        try {
            Map<String,UriRef> previousAnnotations = new LinkedHashMap<String,UriRef>();
            for (Map.Entry<String,List<NameOccurrence>> nameInContext : entityNames.entrySet()) {
    
                String name = nameInContext.getKey();
                List<NameOccurrence> occurrences = nameInContext.getValue();
    
                UriRef firstOccurrenceAnnotation = null;
    
                for (NameOccurrence occurrence : occurrences) {
                    UriRef textAnnotation = EnhancementEngineHelper.createTextEnhancement(ci, this);
                    g.add(new TripleImpl(textAnnotation, ENHANCER_SELECTED_TEXT, 
                        new PlainLiteralImpl(name, language)));
                    g.add(new TripleImpl(textAnnotation, ENHANCER_SELECTION_CONTEXT, 
                        new PlainLiteralImpl(occurrence.context, language)));
                    g.add(new TripleImpl(textAnnotation, DC_TYPE, typeUri));
                    g.add(new TripleImpl(textAnnotation, ENHANCER_CONFIDENCE, literalFactory
                            .createTypedLiteral(occurrence.confidence)));
                    if (occurrence.start != null && occurrence.end != null) {
                        g.add(new TripleImpl(textAnnotation, ENHANCER_START, literalFactory
                                .createTypedLiteral(occurrence.start)));
                        g.add(new TripleImpl(textAnnotation, ENHANCER_END, literalFactory
                                .createTypedLiteral(occurrence.end)));
                    }
    
                    // add the subsumption relationship among occurrences of the same
                    // name

View Full Code Here

     *
     * @param occs a Collection of entity information
     * @param ci the content item
     */
    public void createEnhancements(Collection<CalaisEntityOccurrence> occs, ContentItem ci) {
        LiteralFactory literalFactory = LiteralFactory.getInstance();
        final Language language; // used for plain literals representing parts fo the content
        String langString = getMetadataLanguage(ci.getMetadata(), null);
        if(langString != null && !langString.isEmpty()){
            language = new Language(langString);
        } else {
            language = null;
        }
        //TODO create TextEnhancement (form, start, end, type?) and EntityAnnotation (id, name, type)
        HashMap<Resource, UriRef> entityAnnotationMap = new HashMap<Resource, UriRef>();
        for (CalaisEntityOccurrence occ : occs) {
            UriRef textAnnotation = EnhancementEngineHelper.createTextEnhancement(
                    ci, this);
            MGraph model = ci.getMetadata();
            model.add(new TripleImpl(textAnnotation, DC_TYPE, occ.type));
            // for autotagger use the name instead of the matched term (that might be a pronoun!)
            if (onlyNERMode) {
                model.add(new TripleImpl(textAnnotation, ENHANCER_SELECTED_TEXT,new PlainLiteralImpl(occ.name,language)));
            }
            else {
                model.add(new TripleImpl(textAnnotation, ENHANCER_SELECTED_TEXT, new PlainLiteralImpl(occ.exact,language)));
            }
            model.add(new TripleImpl(textAnnotation, ENHANCER_START, literalFactory.createTypedLiteral(occ.offset)));
            model.add(new TripleImpl(textAnnotation, ENHANCER_END, literalFactory.createTypedLiteral(occ.offset + occ.length)));
            model.add(new TripleImpl(textAnnotation, ENHANCER_SELECTION_CONTEXT, new PlainLiteralImpl(occ.context,language)));
            //create EntityAnnotation only once but add a reference to the textAnnotation
            if (entityAnnotationMap.containsKey(occ.id)) {
                model.add(new TripleImpl(entityAnnotationMap.get(occ.id), DC_RELATION, textAnnotation));
            } else {

View Full Code Here

     *
     * @return the URI of the new enhancement instance
     */
    protected static UriRef createEnhancement(MGraph metadata,
            EnhancementEngine engine, UriRef contentItemId){
        LiteralFactory literalFactory = LiteralFactory.getInstance();


        UriRef enhancement = new UriRef("urn:enhancement-"
                + EnhancementEngineHelper.randomUUID());
        //add the Enhancement Type
        metadata.add(new TripleImpl(enhancement, Properties.RDF_TYPE,
                TechnicalClasses.ENHANCER_ENHANCEMENT));
        //add the extracted from content item
        metadata.add(new TripleImpl(enhancement,
                Properties.ENHANCER_EXTRACTED_FROM, contentItemId));
        // creation date
        metadata.add(new TripleImpl(enhancement, Properties.DC_CREATED,
                literalFactory.createTypedLiteral(new Date())));


        // the engines that extracted the data
        // TODO: add some kind of versioning info for the extractor?
        // TODO: use a public dereferencing URI instead? that would allow for
        // explicit versioning too
        /* NOTE (Rupert Westenthaler 2010-05-26):
         * The Idea is to use the  ComponentContext in the activate() method of
         * an Enhancer to get the bundle name/version and use that as an
         * URI for the creator.
         * We would need to add getEnhancerID() method to the enhancer interface
         * to access this information
          */
        metadata.add(new TripleImpl(enhancement, Properties.DC_CREATOR,
                literalFactory.createTypedLiteral(engine.getClass().getName())));
        return enhancement;
    }

View Full Code Here

0 1 2 3 4 5 6

TOP

Related Classes of org.apache.clerezza.rdf.core.LiteralFactory

org.apache.clerezza.platform.content.WebDavUtils

org.apache.clerezza.rdf.jena.tdb.storage.TdbMGraphTest

org.apache.stanbol.cmsadapter.cmis.mapping.CMISRDFMapper

org.apache.stanbol.cmsadapter.core.mapping.DefaultRDFBridgeImpl

org.apache.stanbol.cmsadapter.core.mapping.RDFBridgeHelper

org.apache.stanbol.cmsadapter.jcr.mapping.JCRRDFMapper

org.apache.stanbol.commons.indexedgraph.IndexedGraphTest

org.apache.stanbol.contenthub.web.resources.ContentItemResource

org.apache.stanbol.enhancer.engine.topic.TopicClassificationEngine

org.apache.stanbol.enhancer.engines.celi.lemmatizer.impl.CeliLemmatizerEnhancementEngine

All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.