Package edu.umd.hooka

Examples of edu.umd.hooka.Alignment$AIterator

48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
    System.out.println("OLD:\n" + at);
    System.out.println("NEW:\n" + ac);
    System.out.println("\nnew:\n"+tc);
    //if (true) return;
    PerplexityReporter cr = new PerplexityReporter();
    Alignment a = hmm.viterbiAlign(pp, cr);
    double ce1 = cr.getCrossEntropy();
    System.out.println(a.toStringVisual() + "\n"+cr);
    assertTrue(a.aligned(0, 0));
    System.out.println(hmm.backtrace);
    hmm = new HMM(tc, ac);
    hmm.buildHMMTables(pp);
    cr.reset();
    a = hmm.viterbiAlign(pp, cr);
    double ce2 = cr.getCrossEntropy();
    // perplexity should decrease!
    assertTrue(ce1 > ce2);
    assert(tc.get(1,1) > 0.0f);
    System.out.println(a.toStringVisual() + "\n" + cr + "\nPG::\n" + hmm.computeAlignmentPosteriors(pp));
    assertEquals(a.getELength(), e.size());
    assertEquals(a.getFLength(), f.size());
  }
View Full Code Here
90
91
92
93
94
95
96
97
98
99
    Phrase e = new Phrase(ew, 1);
    PhrasePair pp = new PhrasePair(f, e);
    hmm = new HMM(tt, at);
    hmm.buildHMMTables(pp);
    PerplexityReporter cr = new PerplexityReporter();
    Alignment a = hmm.viterbiAlign(pp, cr);
    assertEquals(a.getELength(), e.size());
    assertEquals(a.getFLength(), f.size());
    System.out.println(a.toStringVisual() + "\n"+cr);
  }
View Full Code Here
127
128
129
130
131
132
133
134
135
136
137
138
139
    PhrasePair pp = new PhrasePair(f, e);
    tt.normalize();
    hmm = new HMM(tt, at);
    hmm.buildHMMTables(pp);
    PerplexityReporter cr = new PerplexityReporter();
    Alignment a = hmm.viterbiAlign(pp, cr);
    assertEquals(a.getELength(), e.size());
    assertEquals(a.getFLength(), f.size());
    System.err.println(hmm.emission);
    System.err.println(hmm.transition);
    System.out.println(tt);
    System.err.println(hmm.viterbi);
    System.err.println(hmm.backtrace);
View Full Code Here
173
174
175
176
177
178
179
180
181
182
183
184
185
186
    System.out.println("COUNTS:\n" + tc);
    tc.normalize();
    ac.normalize();
   
    PerplexityReporter cr = new PerplexityReporter();
    Alignment a = hmm.viterbiAlign(pp, cr);
    assertEquals(a.getELength(), e.size());
    assertEquals(a.getFLength(), f.size());
    AlignmentPosteriorGrid pg = hmm.computeAlignmentPosteriors(pp);
    System.out.println(a.toStringVisual() + "\nPG:\n"+pg + "\n"+cr + "\n");
    System.out.println(hmm.emission);
    System.out.println(hmm.transition);
    System.out.println("Done NULL");
  }
View Full Code Here
68
69
70
71
72
73
74
75
76
77
78
        Path pathTTable = new Path(ttable);
        TTable tt = loadTTable(pathTTable);
        m1 = new Model1(tt, true);
      }
      cr.reset();
      Alignment a = m1.viterbiAlign(value, cr);
      out.set(a.toString());
      output.collect(key, out);
      reporter.incrCounter(CrossEntropyCounters.LOGPROB, (long)(cr.getTotalLogProb()));
      reporter.incrCounter(CrossEntropyCounters.WORDCOUNT, cr.getTotalWordCount());
    }
View Full Code Here
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
  @Override
  public Alignment viterbiAlign(PhrasePair sentence, PerplexityReporter viterbiPerp) {
    int[] es = sentence.getE().getWords();
    int[] fs = sentence.getF().getWords();
    float threshold = 0.27f;
    Alignment res = new Alignment(fs.length, es.length);
    AlignmentPosteriorGrid g = computeAlignmentPosteriors(sentence);
    for (int j=0; j<fs.length; j++) {
      for (int i=1; i<es.length; i++) {
        float post = g.getAlignmentPointPosterior(j, i);
        if (post > threshold) { res.align(j, i-1); }
      }
    }
    return res;
  }
View Full Code Here
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
  }

  public Alignment realViterbiAlign(PhrasePair sentence, PerplexityReporter viterbiPerp) {
    int[] es = sentence.getE().getWords();
    int[] fs = sentence.getF().getWords();
    Alignment res = new Alignment(fs.length, es.length);
    float viterbiScore = 0.0f;
    for (int j=0; j<fs.length; j++) {
      float bestProb = -1.0f;
      int besti = -1;
      int starti = 0; // TODO - should use NULL?
      if (_includeEnglishNullWord) starti = -1;
      for (int i=starti; i<es.length; i++) {
        float curProb = 0.0f;
        if (i == -1)
          curProb = tmodel.get(0, fs[j]);
        else
          curProb = tmodel.get(es[i], fs[j]);
        if (curProb > bestProb) { bestProb = curProb; besti = i; }
      }
      if (besti < 0) {
        ; //throw new RuntimeException("Implement or ignore!");
      } else {
        res.align(j, besti);
      }
      viterbiScore += Math.log(bestProb);
    }
    viterbiPerp.addFactor(viterbiScore - (fs.length * Math.log(es.length + 1.0)), fs.length);
    return res;
View Full Code Here
9
10
11
12
13
14
15
16
17
18
19
20
21
  public static float computeAER(Iterator<ReferenceAlignment> refi, Iterator<Alignment> testi) {
    float num = 0.0f;
    float den = 0.0f;
    while (refi.hasNext()) {
      ReferenceAlignment ref = refi.next();
      Alignment test = testi.next();
      num += (ref.countProbableHits(test) + ref.countSureHits(test));
      den += (test.countAlignmentPoints() + ref.countAlignmentPoints());
    }
    if (testi.hasNext())
      throw new RuntimeException("Mismatch in lengths");
    return num / den;
  }
View Full Code Here
342
343
344
345
346
347
348
349
350
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
393
  @Override
  public Alignment viterbiAlign(PhrasePair sentence,
      PerplexityReporter reporter) {
    this.buildHMMTables(sentence);
    Alignment res = new Alignment(sentence.getF().size(), sentence.getE().size());
    int J = sentence.getF().size() + 1;
    int numStates = getNumStates();
    viterbi.resize(J, getNumStates());
    backtrace.resize(J, getNumStates());
    viterbi.fill(Float.NEGATIVE_INFINITY);
    viterbi.set(0, 0, 0.0f);
    int lene = sentence.getE().getWords().length;
    Alignment m1a = null;
    if (m1_post != null)
      m1a = m1_post.alignPosteriorThreshold(THRESH);

    //System.out.println(emission);
    for (int j = 1; j < J; j++) {
      //System.out.println("J="+j);
      boolean valid = false;
      for (int s = 1; s < numStates; s++) {
        float best = Float.NEGATIVE_INFINITY;
        int best_s = -1;
        double emitLogProb = Math.log(emission.get(j, s));
        if (emitLogProb == Float.NEGATIVE_INFINITY) {
          //System.out.println("BAD STATE: " + j + " " + s);
          continue;
        }
        //System.out.println("j="+j + " s="+s+ "  ep"+emitLogProb);
        for (int s_prev = 0; s_prev < numStates; s_prev++) {
          float m1boost = 1.0f;
          float m1penalty = 0.0f;
          boolean use_m1 = false;
          if (m1a != null && m1a.isFAligned(j-1)) {
            float m1post = 0.0f;
            use_m1 = true;
            for (int i=0; i<lene; i++) {
              if (m1a.aligned(j-1, i))
                m1post = m1_post.getAlignmentPointPosterior(j-1, i+1);
            }
            m1boost = (float)Math.sqrt(m1post);
            m1penalty = 1.0f - m1boost;
          }
          float trans = getTransitionProb(s_prev, s);
          if (use_m1) {
            if (s <= l && s > 0 && m1a.aligned(j-1, s-1))
              trans = m1boost;
            else
              trans *= m1penalty;
          }
          float cur = (float)(viterbi.get(j - 1, s_prev) +
View Full Code Here
46
47
48
49
50
51
52
53
54
55
56
57
  static class GrowDiagFinalAndRefiner extends Refiner {
    static Koehn koehn = new Koehn();
    static KoehnFinal koehnFinal = new KoehnFinal();
    public Alignment refine(Alignment a2, Alignment a1) {
      Alignment au = Alignment.union(a1, a2);
      Alignment a = Alignment.intersect(a1, a2);
     
      grow(a, koehn, false, au);
      //System.out.println(a1.toString());
      grow(a, koehnFinal, true, a1);
      grow(a, koehnFinal, true, a2);
View Full Code Here
TOP

Related Classes of edu.umd.hooka.Alignment$AIterator

  • edu.umd.hooka.alignment.aer.AER
  • edu.umd.hooka.alignment.HadoopAlign$AlignMapper
  • edu.umd.hooka.alignment.HadoopAlign$EMapper
  • edu.umd.hooka.alignment.hmm.HMM
  • edu.umd.hooka.alignment.hmm.HMMTest
  • edu.umd.hooka.alignment.HSymAlign$Reduce
  • edu.umd.hooka.alignment.M1ViterbiExtract$M1ViterbiMapper
  • edu.umd.hooka.alignment.model1.Model1
  • edu.umd.hooka.alignment.RefinerFactory$GrowDiagFinalAndRefiner
  • java.util.regex.Matcher
    • 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.