Package org.apache.uima.ruta.textruler.learner.lp2

Examples of org.apache.uima.ruta.textruler.learner.lp2.OptimizedLP2

147
148
149
150
151
152
153
154
155
156
157
      if (true) {
        rule = rule.copy();
        LP2RuleItem item = rule.getMarkingRuleItem();
        // TextRulerToolkit.log("CONTEXTUAL RULE CANDIDATE: "+rule.getRuleString()+"  ;  "+rule.getCoveringStatistics());
        item.setContextConstraint(new MLLP2ContextConstraint(slotMaximumTokenCountMap.get(rule
                .getTarget().getSingleSlotRawTypeName()), rule));
        rule.setIsContextualRule(true);

        rule.setNeedsCompile(true);
View Full Code Here
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
      currentBestRules.cutToMaxSize();
    } else if (!tooFewPositives && (ctxStartRules.size() > 0)) {
      // new: use precalculated ctx startrules:
      for (LP2Rule ctxStartRule : ctxStartRules) {

        MLLP2ContextConstraint ctxConstraint = ctxStartRule.getMarkingRuleItem()
                .getContextConstraint();
        LP2Rule newCTXRule = newRule.copy();
        newCTXRule.setIsContextualRule(true);
        newCTXRule.getMarkingRuleItem().setContextConstraint(ctxConstraint.copy());
        newCTXRule.setNeedsCompile(true);
        newCTXRule.compileRuleString();
        newCTXRule.setCoveringStatistics(getCoveringIntersection(newRule.getCoveringStatistics(),
                ctxStartRule.getCoveringStatistics()));
        // if
View Full Code Here
453
454
455
456
457
458
459
460
461
462
463
    // TODO make all other tags contextual tags here. for now we take only
    // the counterpart
    // tag of the current learning process: (opening/closing tags)

    LP2RuleItem ctxItem = new LP2RuleItem();
    MLLP2ContextConstraint ctxConstraint = new MLLP2ContextConstraint(
            slotMaximumTokenCountMap.get(aStartRule.getTarget().getSingleSlotRawTypeName()),
            aStartRule);
    ctxItem.setContextConstraint(ctxConstraint);
    LP2Rule ctxStartRule = new LP2Rule(this, aStartRule.getTarget());
    ctxStartRule.setIsContextualRule(true);
View Full Code Here
147
148
149
150
151
152
153
154
155
156
157
      if (true) {
        rule = rule.copy();
        LP2RuleItem item = rule.getMarkingRuleItem();
        // TextRulerToolkit.log("CONTEXTUAL RULE CANDIDATE: "+rule.getRuleString()+"  ;  "+rule.getCoveringStatistics());
        item.setContextConstraint(new MLLP2ContextConstraint(slotMaximumTokenCountMap.get(rule
                .getTarget().getSingleSlotRawTypeName()), rule));
        rule.setIsContextualRule(true);

        rule.setNeedsCompile(true);
View Full Code Here
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
      currentBestRules.cutToMaxSize();
    } else if (!tooFewPositives && (ctxStartRules.size() > 0)) {
      // new: use precalculated ctx startrules:
      for (LP2Rule ctxStartRule : ctxStartRules) {

        MLLP2ContextConstraint ctxConstraint = ctxStartRule.getMarkingRuleItem()
                .getContextConstraint();
        LP2Rule newCTXRule = newRule.copy();
        newCTXRule.setIsContextualRule(true);
        newCTXRule.getMarkingRuleItem().setContextConstraint(ctxConstraint.copy());
        newCTXRule.setNeedsCompile(true);
        newCTXRule.compileRuleString();
        newCTXRule.setCoveringStatistics(getCoveringIntersection(newRule.getCoveringStatistics(),
                ctxStartRule.getCoveringStatistics()));
        // if
View Full Code Here
453
454
455
456
457
458
459
460
461
462
463
    // TODO make all other tags contextual tags here. for now we take only
    // the counterpart
    // tag of the current learning process: (opening/closing tags)

    LP2RuleItem ctxItem = new LP2RuleItem();
    MLLP2ContextConstraint ctxConstraint = new MLLP2ContextConstraint(
            slotMaximumTokenCountMap.get(aStartRule.getTarget().getSingleSlotRawTypeName()),
            aStartRule);
    ctxItem.setContextConstraint(ctxConstraint);
    LP2Rule ctxStartRule = new LP2Rule(this, aStartRule.getTarget());
    ctxStartRule.setIsContextualRule(true);
View Full Code Here
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
    for (int i = leftContext.size() - 1; i >= 0; i--) {
      TextRulerAnnotation annot = new TextRulerAnnotation(leftContext.get(i), example.getDocument());
      LP2RuleItem item = new LP2RuleItem();
      item.setWordConstraint(annot);
      if (item.getWordConstraint().isRegExpConstraint())
        item.addOtherConstraint(new MLLP2OtherConstraint(annot, annot));
      rule.addPreFillerItem(item);
    }

    for (AnnotationFS afs : rightContext) {
      TextRulerAnnotation annot = new TextRulerAnnotation(afs, example.getDocument());
      LP2RuleItem item = new LP2RuleItem();
      item.setWordConstraint(annot);
      if (item.getWordConstraint().isRegExpConstraint())
        item.addOtherConstraint(new MLLP2OtherConstraint(annot, annot));

      rule.addPostFillerItem(item);
    }
    TextRulerToolkit.log("INITIAL RULE: " + rule.getRuleString());
    return rule;
View Full Code Here
280
281
282
283
284
285
286
287
288
289
290
      newItem = new LP2RuleItem();
      for (MLLP2OtherConstraint c : currentConstraintTuple)
        newItem.addOtherConstraint(c.copy());
      result.add(newItem);
    } else {
      MLLP2OtherConstraint currentConstraint = otherConstraints.get(currentConstraintIndex);
      // recurse WITH and WITHOUT this key:
      recursiveGeneralizeRuleItem(baseItem, otherConstraints, currentConstraintIndex + 1,
              currentConstraintTuple, result);
      currentConstraintTuple.add(currentConstraint);
      recursiveGeneralizeRuleItem(baseItem, otherConstraints, currentConstraintIndex + 1,
View Full Code Here
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
      if (wordItem.getWordConstraint().isRegExpConstraint()) {
        LP2RuleItem basicItem = new LP2RuleItem();
        // basicItem.setOtherConstraint("basicTM", new
        // MLLP2OtherConstraint(tokenAnnotation, tokenAnnotation));
        basicItem.addOtherConstraint(new MLLP2OtherConstraint(tokenAnnotation, tokenAnnotation));
        result.add(createStartRuleForConstraint(example.getTarget(), prePostIndex + 1, isPre,
                basicItem));
      }

      // // POS-Tags created by our test hmm tagger.
      // Type posTagsRootType = ts.getType("org.apache.uima.ml.ML.postag");
      // if (posTagsRootType != null)
      // {
      // List<AnnotationFS> posTagAnnotations =
      // TextRulerToolkit.getAnnotationsWithinBounds(example.getDocumentCAS(),
      // tokenAnnotation.getBegin(), tokenAnnotation.getEnd(), null,
      // posTagsRootType);
      // if (posTagAnnotations.size()>0)
      // {
      // if (TextRulerToolkit.DEBUG && posTagAnnotations.size()>1)
      // {
      // TextRulerToolkit.logIfDebug("HOW CAN ONE TOKEN HAVE MORE THAN ONE POS TAG ?? "+tokenAnnotation.getBegin()+":"+tokenAnnotation.getEnd()+"="+tokenAnnotation.getCoveredText());
      // for (AnnotationFS afs : posTagAnnotations)
      // {
      // System.out.print(afs.getType().getShortName()+":"+afs.getCoveredText()+" "+afs.getBegin()+":"+afs.getEnd()+"\n");
      // }
      // TextRulerToolkit.logIfDebug("");
      // }
      //
      // TextRulerAnnotation posTagAnnotation = new
      // TextRulerAnnotation(posTagAnnotations.get(0),
      // example.getDocument());
      // LP2RuleItem basicItem = new LP2RuleItem();
      // basicItem.setOtherConstraint("postag", new
      // MLLP2OtherConstraint(posTagAnnotation, posTagAnnotation));
      // result.add(createStartRuleForConstraint(example.getTarget(),
      // prePostIndex+1, isPre, basicItem));
      // }
      // }

      // new dynamic system: grab everything we get from the annotation
      // index that lies over this token:
      // (annotations WITHIN (with smaller bounds than the token itself)
      // are ignored for now! we could
      // add using them with the CONTAINS constraint. but our
      // MLLP2OtherConstraint is not yet capable of this!

      List<AnnotationFS> featureAnnotations = TextRulerToolkit.getOtherAnnotationsOverToken(docCas,
              tokenAFS, filterSetWithSlotNames);
      if (TextRulerToolkit.DEBUG && featureAnnotations.size() > 1) {
        TextRulerToolkit.log("FOUND MORE THAN ONE EXTRA TOKEN FEATURE ANNOTATION !");
        for (AnnotationFS featA : featureAnnotations)
          TextRulerToolkit.log(featA.toString());
        TextRulerToolkit.log("--------------------------------");
      }
      for (AnnotationFS featA : featureAnnotations) {
        TextRulerAnnotation featureAnnot = new TextRulerAnnotation(featA, example.getDocument());
        LP2RuleItem basicItem = new LP2RuleItem();
        basicItem.addOtherConstraint(new MLLP2OtherConstraint(tokenAnnotation, featureAnnot));
        result.add(createStartRuleForConstraint(example.getTarget(), prePostIndex + 1, isPre,
                basicItem));
      }

    }
View Full Code Here
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
    for (int i = leftContext.size() - 1; i >= 0; i--) {
      TextRulerAnnotation annot = new TextRulerAnnotation(leftContext.get(i), example.getDocument());
      LP2RuleItem item = new LP2RuleItem();
      item.setWordConstraint(annot);
      if (item.getWordConstraint().isRegExpConstraint())
        item.addOtherConstraint(new MLLP2OtherConstraint(annot, annot));
      rule.addPreFillerItem(item);
    }

    for (AnnotationFS afs : rightContext) {
      TextRulerAnnotation annot = new TextRulerAnnotation(afs, example.getDocument());
      LP2RuleItem item = new LP2RuleItem();
      item.setWordConstraint(annot);
      if (item.getWordConstraint().isRegExpConstraint())
        item.addOtherConstraint(new MLLP2OtherConstraint(annot, annot));

      rule.addPostFillerItem(item);
    }
    TextRulerToolkit.log("INITIAL RULE: " + rule.getRuleString());
    return rule;
View Full Code Here
TOP

Related Classes of org.apache.uima.ruta.textruler.learner.lp2.OptimizedLP2

  • org.apache.uima.cas.CAS
  • org.apache.uima.cas.text.AnnotationFS
  • org.apache.uima.cas.Type
  • org.apache.uima.cas.TypeSystem
  • org.apache.uima.ruta.textruler.core.TextRulerAnnotation
  • org.apache.uima.ruta.textruler.core.TextRulerExample
  • org.apache.uima.ruta.textruler.core.TextRulerRuleItem
  • org.apache.uima.ruta.textruler.core.TextRulerRuleList
  • org.apache.uima.ruta.textruler.core.TextRulerRulePattern
  • org.apache.uima.ruta.textruler.core.TextRulerShiftExample
    • 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.