Package org.uimafit.factory

Examples of org.uimafit.factory.AggregateBuilder

290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
  }

  @Override
  protected AnnotationStatistics<String> test(CollectionReader collectionReader, File directory)
      throws Exception {
    AggregateBuilder aggregateBuilder = this.getPreprocessorAggregateBuilder();
    aggregateBuilder.add(CopyFromGold.getDescription(EventMention.class, TimeMention.class));
   
    aggregateBuilder.add(AnalysisEngineFactory.createPrimitiveDescription(
        RemoveCrossSentenceRelations.class,
        RemoveCrossSentenceRelations.PARAM_SENTENCE_VIEW,
        CAS.NAME_DEFAULT_SOFA,
        RemoveCrossSentenceRelations.PARAM_RELATION_VIEW,
        GOLD_VIEW_NAME));
    if (this.useClosure) {
      aggregateBuilder.add(
          AnalysisEngineFactory.createPrimitiveDescription(AddClosure.class),//AnalysisEngineFactory.createPrimitiveDescription(AddTransitiveContainsRelations.class),
          CAS.NAME_DEFAULT_SOFA,
          GOLD_VIEW_NAME);
         
//      aggregateBuilder.add(
//          AnalysisEngineFactory.createPrimitiveDescription(AddContain2Overlap.class),
//          CAS.NAME_DEFAULT_SOFA,
//          GOLD_VIEW_NAME);
      //      aggregateBuilder.add(
      //          AnalysisEngineFactory.createPrimitiveDescription(AddTransitiveBeforeAndOnRelations.class),
      //          CAS.NAME_DEFAULT_SOFA,
      //          GOLD_VIEW_NAME);
    }
   
//    aggregateBuilder.add(
//        AnalysisEngineFactory.createPrimitiveDescription(RemoveNonContainsRelations.class,
//            RemoveNonContainsRelations.PARAM_RELATION_VIEW,
//            GOLD_VIEW_NAME));
    aggregateBuilder.add(
        AnalysisEngineFactory.createPrimitiveDescription(RemoveEventEventRelations.class),
        CAS.NAME_DEFAULT_SOFA,
        GOLD_VIEW_NAME);

    aggregateBuilder.add(AnalysisEngineFactory.createPrimitiveDescription(RemoveRelations.class));
    aggregateBuilder.add(this.baseline ? RecallBaselineEventTimeRelationAnnotator.createAnnotatorDescription(directory) :
      EventTimeRelationAnnotator.createAnnotatorDescription(directory));
    if(this.i2b2Output != null){
      aggregateBuilder.add(AnalysisEngineFactory.createPrimitiveDescription(WriteI2B2XML.class, WriteI2B2XML.PARAM_OUTPUT_DIR, this.i2b2Output), "TimexView", CAS.NAME_DEFAULT_SOFA);
    }
    if (this.useClosure) {//add closure for system output
      aggregateBuilder.add(
          AnalysisEngineFactory.createPrimitiveDescription(AddClosure.class),//AnalysisEngineFactory.createPrimitiveDescription(AddTransitiveContainsRelations.class),
          GOLD_VIEW_NAME,
          CAS.NAME_DEFAULT_SOFA
          );
    }
    Function<BinaryTextRelation, ?> getSpan = new Function<BinaryTextRelation, HashableArguments>() {
      public HashableArguments apply(BinaryTextRelation relation) {
        return new HashableArguments(relation);
      }
    };
    Function<BinaryTextRelation, String> getOutcome = AnnotationStatistics.annotationToFeatureValue("category");

    AnnotationStatistics<String> stats = new AnnotationStatistics<String>();
    JCasIterable jcasIter =new JCasIterable(collectionReader, aggregateBuilder.createAggregate());
    JCas jCas = null;
    while(jcasIter.hasNext()) {
      jCas = jcasIter.next();
      JCas goldView = jCas.getView(GOLD_VIEW_NAME);
      JCas systemView = jCas.getView(CAS.NAME_DEFAULT_SOFA);
View Full Code Here
119
120
121
122
123
124
125
126
127
128
    }
  }

  @Override
  protected void train(CollectionReader collectionReader, File directory) throws Exception {
    AggregateBuilder aggregateBuilder = this.getPreprocessorAggregateBuilder();
    aggregateBuilder.add(CopyFromGold.getDescription(EventMention.class));
    aggregateBuilder.add(CopyFromGold.getDescription(EventProperties.class));
    SimplePipeline.runPipeline(collectionReader, aggregateBuilder.createAggregate());
  }
View Full Code Here
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
  @Override
  protected Map<String, AnnotationStatistics<String>> test(
      CollectionReader collectionReader,
      File directory) throws Exception {
    AggregateBuilder aggregateBuilder = this.getPreprocessorAggregateBuilder();
    aggregateBuilder.add(CopyFromGold.getDescription(EventMention.class));
    aggregateBuilder.add(CopyFromGold.getDescription(TimeMention.class));
    aggregateBuilder.add(SentenceAnnotator.getDescription());
    aggregateBuilder.add(TokenAnnotator.getDescription());
    aggregateBuilder.add(PosTaggerAnnotator.getDescription());
    aggregateBuilder.add(DefaultSnowballStemmer.getDescription("English"));
    aggregateBuilder.add(ParserAnnotator.getDescription());
    aggregateBuilder.add(EventToClearTKEventAnnotator.getAnnotatorDescription());//for every cTakes eventMention, create a cleartk event
    aggregateBuilder.add(ClearTKDocumentCreationTimeAnnotator.getAnnotatorDescription());//for every jCAS create an empty DCT, and add it to index
    aggregateBuilder.add(EventTenseAnnotator.FACTORY.getAnnotatorDescription("/org/cleartk/timeml/event/eventtenseannotator/model.jar"));
    aggregateBuilder.add(EventAspectAnnotator.FACTORY.getAnnotatorDescription("/org/cleartk/timeml/event/eventaspectannotator/model.jar"));
    aggregateBuilder.add(EventClassAnnotator.FACTORY.getAnnotatorDescription("/org/cleartk/timeml/event/eventclassannotator/model.jar"));
    aggregateBuilder.add(EventPolarityAnnotator.FACTORY.getAnnotatorDescription("/org/cleartk/timeml/event/eventpolarityannotator/model.jar"));
    aggregateBuilder.add(EventModalityAnnotator.FACTORY.getAnnotatorDescription("/org/cleartk/timeml/event/eventmodalityannotator/model.jar"));
    aggregateBuilder.add(TemporalLinkEventToDocumentCreationTimeAnnotator.FACTORY.getAnnotatorDescription("/org/cleartk/timeml/tlink/temporallinkeventtodocumentcreationtimeannotator/model.jar"));
    aggregateBuilder.add(ClearTKDocTimeRelAnnotator.getAnnotatorDescription());// for every tlink, check if it cover and event, add the tlink type to the event's docTimeRel attribute

    Function<EventMention, ?> eventMentionToSpan = AnnotationStatistics.annotationToSpan();
    Map<String, Function<EventMention, String>> propertyGetters;
    propertyGetters = new HashMap<String, Function<EventMention, String>>();
    for (String name : PROPERTY_NAMES) {
      propertyGetters.put(name, getPropertyGetter(name));
    }

    Map<String, AnnotationStatistics<String>> statsMap = new HashMap<String, AnnotationStatistics<String>>();
    statsMap.put(DOC_TIME_REL, new AnnotationStatistics<String>());
    for (JCas jCas : new JCasIterable(collectionReader, aggregateBuilder.createAggregate())) {
      JCas goldView = jCas.getView(GOLD_VIEW_NAME);
      JCas systemView = jCas.getView(CAS.NAME_DEFAULT_SOFA);
      String text = goldView.getDocumentText();
      for (Segment segment : JCasUtil.select(jCas, Segment.class)) {
        if (!THYMEData.SEGMENTS_TO_SKIP.contains(segment.getId())) {
View Full Code Here
98
99
100
101
102
103
104
105
106
107
108
  protected abstract void trainAndPackage(File directory) throws Exception;

  @Override
  protected void train(CollectionReader collectionReader, File directory) throws Exception {
    AggregateBuilder aggregateBuilder = this.getPreprocessorAggregateBuilder();
    aggregateBuilder.add(CopyFromGold.getDescription(this.annotationClass));
    aggregateBuilder.add(this.getDataWriterDescription(directory), "TimexView", CAS.NAME_DEFAULT_SOFA);
    SimplePipeline.runPipeline(collectionReader, aggregateBuilder.createAggregate());
    this.trainAndPackage(directory);
  }
View Full Code Here
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
  protected abstract Collection<? extends Annotation> getSystemAnnotations(JCas jCas, Segment segment);

  @Override
  protected AnnotationStatistics<String> test(CollectionReader collectionReader, File directory)
      throws Exception {
    AggregateBuilder aggregateBuilder = this.getPreprocessorAggregateBuilder();
    aggregateBuilder.add(this.getAnnotatorDescription(directory), "TimexView", CAS.NAME_DEFAULT_SOFA);
    if(this.i2b2Output != null){
      aggregateBuilder.add(AnalysisEngineFactory.createPrimitiveDescription(WriteI2B2XML.class, WriteI2B2XML.PARAM_OUTPUT_DIR, this.i2b2Output), "TimexView", CAS.NAME_DEFAULT_SOFA);
    }
    AnnotationStatistics<String> stats = new AnnotationStatistics<String>();
    Ordering<Annotation> bySpans = Ordering.<Integer> natural().lexicographical().onResultOf(
        new Function<Annotation, List<Integer>>() {
          @Override
          public List<Integer> apply(Annotation annotation) {
            return Arrays.asList(annotation.getBegin(), annotation.getEnd());
          }
        });
    for (JCas jCas : new JCasIterable(collectionReader, aggregateBuilder.createAggregate())) {
      JCas goldView = jCas.getView(GOLD_VIEW_NAME);
      JCas systemView = jCas.getView(CAS.NAME_DEFAULT_SOFA);
      for (Segment segment : JCasUtil.select(jCas, Segment.class)) {
        if (!THYMEData.SEGMENTS_TO_SKIP.contains(segment.getId())) {
          Collection<? extends Annotation> goldAnnotations = this.getGoldAnnotations(goldView, segment);
View Full Code Here
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
   
    JCasIterable[] casIters = new JCasIterable[nFolds];
    for (int fold = 0; fold < nFolds; ++fold) {
      List<Integer> xfoldTrain = selectTrainItems(allTrain, nFolds, fold);
      List<Integer> xfoldTest = selectTestItems(allTrain, nFolds, fold);
      AggregateBuilder aggregateBuilder = this.getPreprocessorAggregateBuilder();
      File modelDirectory = getModelDirectory(new File("target/eval/time-spans/fold_"+fold));
      for (Class<? extends JCasAnnotator_ImplBase> annotatorClass : annotatorClasses) {
        EvaluationOfTimeSpans evaluation = new EvaluationOfTimeSpans(
            new File("target/eval/time-spans/" ),
            this.rawTextDirectory,
            this.xmlDirectory,
            this.xmlFormat,
            this.xmiDirectory,
            this.treebankDirectory,
            1,
            0,
            annotatorClass,
            false,
            annotatorTrainingArguments.get(annotatorClass));
        evaluation.prepareXMIsFor(allTrain);
        String name = String.format("%s.errors", annotatorClass.getSimpleName());
        evaluation.setLogging(Level.FINE, new File("target/eval", name));

        // train on 4 of the folds of the training data:
        evaluation.train(evaluation.getCollectionReader(xfoldTrain), modelDirectory);
        if(fold == 0){
          // train the main model as well:
          evaluation.train(evaluation.getCollectionReader(allTrain), directory);
        }
       
      }
      casIters[fold] = new JCasIterable(getCollectionReader(xfoldTest), aggregateBuilder.createAggregate());
    }
    // run meta data-writer for this fold:
    AggregateBuilder writerBuilder = new AggregateBuilder();
    writerBuilder.add(CopyFromGold.getDescription(TimeMention.class));
    writerBuilder.add(this.getDataWriterDescription(directory));
    AnalysisEngine writer = writerBuilder.createAggregate();
    for(JCasIterable casIter : casIters){
      for(JCas jcas : casIter){
        SimplePipeline.runPipeline(jcas, writer);
      }
    }
View Full Code Here
61
62
63
64
65
66
67
68
69
70
71
    assertTrue(tokens.get(29).getCanonicalForm() == null);
   
  }
 
  public static AnalysisEngineDescription getPrerequisitePipeline() throws ResourceInitializationException{
    AggregateBuilder builder = new AggregateBuilder();
    builder.add(SimpleSegmentAnnotator.createAnnotatorDescription());
    builder.add(SentenceDetector.createAnnotatorDescription());
    builder.add(TokenizerAnnotatorPTB.createAnnotatorDescription());
    return builder.createAggregateDescription();
  }
View Full Code Here
69
70
71
72
73
74
75
76
77
78
    builder.add(TokenizerAnnotatorPTB.createAnnotatorDescription());
    return builder.createAggregateDescription();
  }
 
  public static AnalysisEngineDescription getDefaultPipeline() throws ResourceInitializationException, URISyntaxException{
    AggregateBuilder builder = new AggregateBuilder();
    builder.add(getPrerequisitePipeline());
    builder.add(LvgAnnotator.createAnnotatorDescription());
    return builder.createAggregateDescription();
  }
View Full Code Here
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
        + "The patient was well till 6 months ago, when he started having a little blood with stool.";
    JCas jcas = JCasFactory.createJCas();
    jcas.setDocumentText(note);

    // Get the default pipeline
    AggregateBuilder builder = new AggregateBuilder();
    builder.add(ClinicalPipelineFactory.getTokenProcessingPipeline());
    builder.add(AnalysisEngineFactory
        .createPrimitiveDescription(CopyNPChunksToLookupWindowAnnotations.class));
    builder.add(AnalysisEngineFactory
        .createPrimitiveDescription(RemoveEnclosedLookupWindows.class));
    // Commented out the Dictionary lookup for the test
    // Uncomment and set -Dctakes.umlsuser and -Dctakes.umlspw env params if
    // needed
    // builder.add(UmlsDictionaryLookupAnnotator.createAnnotatorDescription());
    builder.add(ClearNLPDependencyParserAE.createAnnotatorDescription());

    // Add EventAnnotator
    builder.add(EventAnnotator
        .createAnnotatorDescription("/org/apache/ctakes/temporal/ae/eventannotator/model.jar"));

    SimplePipeline.runPipeline(jcas, builder.createAggregateDescription());

    Collection<EventMention> mentions = JCasUtil.select(jcas,
        EventMention.class);

    ArrayList<String> temp = new ArrayList<>();
View Full Code Here
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
        + "The patient was well till 6 months ago, when he started having a little blood with stool.";
    JCas jcas = JCasFactory.createJCas();
    jcas.setDocumentText(note);

    // Get the default pipeline with umls dictionary lookup
    AggregateBuilder builder = new AggregateBuilder();
    builder.add(ClinicalPipelineFactory.getTokenProcessingPipeline());
    builder.add(AnalysisEngineFactory
        .createPrimitiveDescription(CopyNPChunksToLookupWindowAnnotations.class));
    builder.add(AnalysisEngineFactory
        .createPrimitiveDescription(RemoveEnclosedLookupWindows.class));
    // Commented out the Dictionary lookup for the test
    // Uncomment and set -Dctakes.umlsuser and -Dctakes.umlspw env params if
    // needed
    // builder.add(UmlsDictionaryLookupAnnotator.createAnnotatorDescription());
    builder.add(ClearNLPDependencyParserAE.createAnnotatorDescription());

    // Add BackwardsTimeAnnotator
    builder.add(BackwardsTimeAnnotator
        .createAnnotatorDescription("/org/apache/ctakes/temporal/ae/timeannotator/model.jar"));

    SimplePipeline.runPipeline(jcas, builder.createAggregateDescription());

    Collection<TimeMention> mentions = JCasUtil.select(jcas,
        TimeMention.class);

    ArrayList<String> temp = new ArrayList<>();
View Full Code Here
TOP

Related Classes of org.uimafit.factory.AggregateBuilder

  • de.tudarmstadt.ukp.similarity.experiments.coling2012.FeatureGeneration
  • de.tudarmstadt.ukp.similarity.experiments.coling2012.util.CharacterNGramIdfValuesGenerator
  • de.tudarmstadt.ukp.similarity.experiments.coling2012.util.WordIdfValuesGenerator
  • org.apache.ctakes.assertion.cr.NegExCorpusReader
  • org.apache.ctakes.assertion.eval.AssertionEvalBasedOnModifier
  • org.apache.ctakes.assertion.eval.AssertionEvaluation
  • org.apache.ctakes.assertion.eval.MergeGoldViewFromOneCasIntoInitialViewOfAnotherCas
  • org.apache.ctakes.assertion.medfacts.cleartk.CreateAssertionDescriptor
  • org.apache.ctakes.assertion.medfacts.cleartk.extractors.CueWordTest
  • org.apache.ctakes.assertion.medfacts.cleartk.extractors.CueWordTest2
    • Copyright © 2018 www.massapicom. All rights reserved.
    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.