Package org.apache.uima.ruta.textruler.core

Examples of org.apache.uima.ruta.textruler.core.TextRulerRuleList

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
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
    exampleDocuments.createExamplesForTarget(target);
    examples = exampleDocuments.getAllPositiveExamples();

    if (shouldAbort())
      return null;
    TextRulerRuleList bestRulesPool = new TextRulerRuleList();
    TextRulerRuleList contextRulesPool = new TextRulerRuleList();
    String slotName = target.getSingleSlotRawTypeName();
    bestRulesPoolMap.put(slotName, bestRulesPool);
    contextRulesPoolMap.put(slotName, contextRulesPool);

    coveredExamples = new HashSet<TextRulerExample>();
    int roundNumber = 0;
    for (TextRulerExample e : examples)
      if (!coveredExamples.contains(e)) {
        if (shouldAbort())
          break;
        roundNumber++;
        currentBestRules = new LP2CurrentBestRulesQueue(maxCurrentBestRulesCount);
        currentContextualRules = new LP2CurrentBestRulesQueue(maxCurrentContextualRulesCount);
        // TextRulerToolkit.log("Example: "+e.getAnnotation().getBegin()+" : "+e.getAnnotation().getEnd());

        induceRulesFromExample(e, roundNumber);

        // TextRulerToolkit.log("Best Rules from this Seed: "+currentBestRules.size());
        // if (TextRulerToolkit.DEBUG && currentBestRules.size()>1)
        // {
        // for (TextRulerRule r : currentBestRules)
        // {
        // TextRulerToolkit.log("\tp="+r.getCoveringStatistics().getCoveredPositivesCount()+"; n="+r.getCoveringStatistics().getCoveredNegativesCount()+";  "+r.getRuleString());
        // for (TextRulerExample ex :
        // r.getCoveringStatistics().getCoveredPositiveExamples())
        // {
        // TextRulerToolkit.log("\t\te="+ex.getAnnotation().getBegin());
        //
        // }
        // }
        // }
        for (LP2Rule bestRule : currentBestRules) {
          addToFinalBestRulesPool(bestRule);
        }
        for (LP2Rule ctxRule : currentContextualRules) {
          addToFinalContextRulesPool(ctxRule);
        }
        sendStatusUpdateToDelegate("New Rules added.", TextRulerLearnerState.ML_RUNNING, true);
      }
    if (TextRulerToolkit.DEBUG) {
      bestRulesPool.saveToRulesFile(getIntermediateRulesFileName(), getFileHeaderString(true));
      // for (TextRulerRule r : bestRulesPool)
      // {
      // TextRulerToolkit.log("p="+r.getCoveringStatistics().getCoveredPositivesCount()+"; n="+r.getCoveringStatistics().getCoveredNegativesCount()+";  "+r.getRuleString());
      // }
    }

    TextRulerRuleList result = bestRulesPool;
    if (contextualRules != null)
      for (TextRulerRule r : contextRulesPool)
        contextualRules.add(r);
    return result;
  }
View Full Code Here
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
      return;
    slotMaximumTokenCountMap.put(slotName, histogram.size() - 1); // -1 since the
    // zero-histogram point
    // also needs a place!

    TextRulerRuleList ctxRules = new TextRulerRuleList();
    TextRulerRuleList bestRules = learnTaggingRules(new TextRulerTarget(slotName,
            MLTargetType.SINGLE_LEFT_BOUNDARY, this), ctxRules); // learn
    // left
    // boundary
    // best
    // rules
    if (bestRules != null) {
      leftBoundaryBestRulesMap.put(slotName, bestRules.getRulesString(""));
      leftBoundaryContextualRulesMap.put(slotName, ctxRules.getRulesString("\t"));
      bestRules.clear(); // free som memory/references
    }
    if (shouldAbort())
      return;
    ctxRules.clear();
    bestRules = learnTaggingRules(new TextRulerTarget(slotName, MLTargetType.SINGLE_RIGHT_BOUNDARY,
            this), ctxRules); // learn
    // right
    // boundary best
    // rules
    if (bestRules != null) {
      rightBoundaryBestRulesMap.put(slotName, bestRules.getRulesString(""));
      rightBoundaryContextualRulesMap.put(slotName, ctxRules.getRulesString("\t"));
    }

    // TODO add correction rule learn stuff
    // testTaggingRulesAndCreateCorrectionRulesExamples(null, STANDARD_MAX_CONTEXTUAL_RULES_COUNT)

    // correct left start
    TextRulerTarget lsTarget = new TextRulerTarget(slotName, MLTargetType.SINGLE_LEFT_CORRECTION,
            this);
    lsTarget.setMaxShiftDistance(shiftSize);
    TextRulerRuleList correctLeftRules = learnTaggingRules(lsTarget, null);

    // resultString = "CAP{REGEXP(\"PM\")} ALL{->MARKONCE(stimeEND)};";
    // try {
    // FileUtils.saveString2File(resultString, file);
    // } catch (IOException e) {
View Full Code Here
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
      String leftBoundaryBestRulesString = leftBoundaryBestRulesMap.get(eachSlot);
      String rightBoundaryBestRulesString = rightBoundaryBestRulesMap.get(eachSlot);
      String leftBoundaryContextualRulesString = leftBoundaryContextualRulesMap.get(eachSlot);
      String rightBoundaryContextualRulesString = rightBoundaryContextualRulesMap.get(eachSlot);
      TextRulerRuleList bestRulesPool = bestRulesPoolMap.get(eachSlot);
      TextRulerRuleList contextRulesPool = contextRulesPoolMap.get(eachSlot);

      sb.append("\n// Slot: " + TextRulerToolkit.getTypeShortName(eachSlot) + "\n");
      sb.append("// LEFT BOUNDARY RULES:\n");
      if (leftBoundaryBestRulesString != null) {
        sb.append(leftBoundaryBestRulesString);
        sb.append("\n// RIGHT BOUNDARY RULES:\n");
        if (rightBoundaryBestRulesString != null)
          sb.append(rightBoundaryBestRulesString);
        else if (bestRulesPool != null)
          sb.append(bestRulesPool.getRulesString(""));

        sb.append("\nBLOCK(contextualRules_" + TextRulerToolkit.getTypeShortName(eachSlot)
                + ") Document{} {\n"
                + "\tDocument{->ASSIGN(redoContextualRules, false)}; // reset flag\n");
        sb.append("\n\t// LEFT BOUNDARY CONTEXTUAL RULES:\n");
        sb.append(leftBoundaryContextualRulesString);

        sb.append("\n\t// RIGHT BOUNDARY CONTEXTUAL RULES:\n");
        if (rightBoundaryBestRulesString != null)
          sb.append(rightBoundaryContextualRulesString);
        else if (contextRulesPool != null)
          sb.append(contextRulesPool.getRulesString("\t"));

        sb.append("\n\t//Document{IF(redoContextualRules)->CALL(thisFile.contextualRules_"
                + TextRulerToolkit.getTypeShortName(eachSlot) + ")};\n}\n");
      } else if (bestRulesPool != null) {
        sb.append(bestRulesPool.getRulesString(""));
        sb.append("\n\t// LEFT BOUNDARY CONTEXTUAL RULES:\n");
        if (contextRulesPool != null)
          sb.append(contextRulesPool.getRulesString(""));
      }
    }

    for (String eachSlot : slotNames) {
      String leftBoundary = TextRulerToolkit.getTypeShortName((new TextRulerTarget(eachSlot,
View Full Code Here
124
125
126
127
128
129
130
131
132
133
134
      examples = exampleDocuments.getAllPositiveExamples();

      if (shouldAbort())
        return;

      slotRules = new TextRulerRuleList();
      ruleList = new RapierRulePriorityQueue(ruleListSize);

      TextRulerToolkit.log("--- RAPIER START for Slot " + currentSlotName);

      sendStatusUpdateToDelegate("Creating initial rule base...",
View Full Code Here
110
111
112
113
114
115
116
117
118
119
120
    // this is the inner loop of the WHISK pseudo-code:
    // For each inst in Training
    // for each tag

    cachedTestedRuleStatistics.clear();
    ruleList = new TextRulerRuleList();
    coveredExamples = new HashSet<TextRulerExample>();

    sendStatusUpdateToDelegate("Creating examples...", TextRulerLearnerState.ML_RUNNING, false);
    TextRulerTarget target = new TextRulerTarget(slotNames[0], this); // only
    // single-slot-target
View Full Code Here
123
124
125
126
127
128
129
130
131
132
133
    // this is the inner loop of the WHISK pseudo-code:
    // For each inst in Training
    // for each tag

    cachedTestedRuleStatistics.clear();
    ruleList = new TextRulerRuleList();
    coveredExamples = new HashSet<TextRulerExample>();

    sendStatusUpdateToDelegate("Creating examples...", TextRulerLearnerState.ML_RUNNING, false);
    for (int i = 0; i < slotNames.length; i++) {
      TextRulerTarget target = new TextRulerTarget(slotNames[i], this);
View Full Code Here
63
64
65
66
67
68
69
70
71
72
73
  @Override
  protected TextRulerRuleList learnTaggingRules(TextRulerTarget target,
          TextRulerRuleList contextualRules) {
    cachedTestedStartRuleStatisticsHitCounter = 0;
    cachedTestedStartRuleStatistics.clear();
    TextRulerRuleList result = super.learnTaggingRules(target, contextualRules);
    TextRulerToolkit
            .log("[OptimizedLP2.learnTaggingRules] saved rule testings due to start rule results cache: "
                    + cachedTestedStartRuleStatisticsHitCounter);
    TextRulerToolkit.log("[OptimizedLP2.learnTaggingRules] cacheSize at end of induction: "
            + cachedTestedStartRuleStatistics.size());
View Full Code Here
124
125
126
127
128
129
130
131
132
133
134
      examples = exampleDocuments.getAllPositiveExamples();

      if (shouldAbort())
        return;

      slotRules = new TextRulerRuleList();
      ruleList = new RapierRulePriorityQueue(ruleListSize);

      TextRulerToolkit.log("--- RAPIER START for Slot " + currentSlotName);

      sendStatusUpdateToDelegate("Creating initial rule base...",
View Full Code Here
114
115
116
117
118
119
120
121
122
123
124
    // this is the inner loop of the WHISK pseudo-code:
    // For each inst in Training
    // for each tag

    cachedTestedRuleStatistics.clear();
    ruleList = new TextRulerRuleList();
    coveredExamples = new HashSet<TextRulerExample>();

    sendStatusUpdateToDelegate("Creating examples...", TextRulerLearnerState.ML_RUNNING, false);
    for (int i = 0; i < slotNames.length; i++) {
      TextRulerTarget target = new TextRulerTarget(slotNames[i], this);
View Full Code Here
110
111
112
113
114
115
116
117
118
119
120
    // this is the inner loop of the WHISK pseudo-code:
    // For each inst in Training
    // for each tag

    cachedTestedRuleStatistics.clear();
    ruleList = new TextRulerRuleList();
    coveredExamples = new HashSet<TextRulerExample>();

    sendStatusUpdateToDelegate("Creating examples...", TextRulerLearnerState.ML_RUNNING, false);
    TextRulerTarget target = new TextRulerTarget(slotNames[0], this); // only
    // single-slot-target
View Full Code Here
TOP

Related Classes of org.apache.uima.ruta.textruler.core.TextRulerRuleList

  • org.apache.uima.ruta.textruler.learner.lp2.BasicLP2
  • org.apache.uima.ruta.textruler.learner.lp2.OptimizedLP2
  • org.apache.uima.ruta.textruler.learner.rapier.Rapier
  • org.apache.uima.ruta.textruler.learner.whisk.generic.Whisk
  • org.apache.uima.ruta.textruler.learner.whisk.token.Whisk
  • java.io.File
    • 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.