Package com.clearnlp.constituent

Examples of com.clearnlp.constituent.CTNode

423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
   */
  private boolean findHeadsCoordination(HeadRule rule, CTNode curr)
  {
    // skip pre-conjunctions and punctuation
    int i, sId, size = curr.getChildrenSize();
    CTNode node;
   
    for (sId=0; sId<size; sId++)
    {
      node = curr.getChild(sId);
     
      if (!CTLibEn.isPunctuation(node) && !CTLibEn.isConjunction(node))
        break;
    }
   
    if (!CTLibEn.containsCoordination(curr, curr.getChildren(sId)))
      return false;
   
    // find conjuncts
    Pattern rTags = getConjunctPattern(curr, sId, size);
    CTNode prevHead = null, mainHead = null;
    boolean isFound = false;
    int bId = 0, eId = sId;
   
    for (; eId<size; eId++)
    {
View Full Code Here
536
537
538
539
540
541
542
543
544
545
546
  }
 
  /** Called by {@link EnglishC2DConverter#findHeadsCoordination(HeadRule, CTNode)}. */
  private CTNode findHeadsCoordinationAux(HeadRule rule, CTNode curr, int bId, int eId, CTNode lastHead)
  {
    CTNode currHead = (eId - bId == 1) ? curr.getChild(bId) : getHead(rule, curr.getChildren(bId, eId), SIZE_HEAD_FLAGS);
   
    if (lastHead != null)
    {
      String label = DEPLibEn.DEP_CONJ;
     
View Full Code Here
562
563
564
565
566
567
568
569
570
571
572
  }
 
  private boolean findHyphens(CTNode node)
  {
    int i, size = node.getChildrenSize();
    CTNode prev, hyph, next;
    boolean isFound = false;
    boolean isVP = node.isPTag(CTLibEn.PTAG_VP);
   
    for (i=0; i<size-2; i++)
    {
View Full Code Here
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
  private boolean findHeadsApposition(CTNode curr)
  {
    if (!curr.isPTagAny(CTLibEn.PTAG_NP, CTLibEn.PTAG_NML) || curr.containsTags("+NN.*"))
      return false;
   
    CTNode fst = curr.getFirstChild("+NP|NML"), snd;
    while (fst != null && fst.containsTags(CTLibEn.POS_POS))
      fst = fst.getNextSibling("+NP|NML");
   
    if (fst == null || fst.c2d.hasHead())  return false;
   
    int i, size = curr.getChildrenSize();
    boolean hasAppo = false;
   
    for (i=fst.getSiblingId()+1; i<size; i++)
    {
      snd = curr.getChild(i);
      if (snd.c2d.hasHead())  continue;
     
      if ((snd.isPTagAny(CTLibEn.PTAG_NP, CTLibEn.PTAG_NML) && !hasAdverbialTag(snd)) ||
View Full Code Here
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
    return hasAppo;
  }

  private boolean findHeadsSmallClause(CTNode node)
  {
    CTNode parent = node.getParent();
   
    if (node.isPTag(CTLibEn.PTAG_S) && !node.containsTags(CTLibEn.PTAG_VP))
    {
      CTNode sbj = node.getFirstChild("-"+CTLibEn.FTAG_SBJ);
      CTNode prd = node.getFirstChild("-"+CTLibEn.FTAG_PRD);
     
      if (sbj != null && prd != null)
      {
        if (parent.isPTag(CTLibEn.PTAG_SQ))
        {
          CTNode vb = parent.getFirstChild("+VB.*");
         
          if (vb != null)
          {
            sbj.c2d.setHead(vb, getDEPLabel(sbj, parent, vb));
            node.pTag = prd.pTag;
View Full Code Here
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
  // ============================= Get Stanford labels =============================
 
  @Override
  protected String getDEPLabel(CTNode C, CTNode P, CTNode p)
  {
    CTNode c = C.c2d.getPhraseHead();
    CTNode d = C.c2d.getDependencyHead();
    String label;
   
    // function tags
    if (hasAdverbialTag(C))
    {
      if (C.isPTagAny(CTLibEn.PTAG_S, CTLibEn.PTAG_SBAR, CTLibEn.PTAG_SINV))
        return DEPLibEn.DEP_ADVCL;
     
      if (C.isPTagAny(CTLibEn.PTAG_NML, CTLibEn.PTAG_NP, CTLibEn.PTAG_QP))
        return DEPLibEn.DEP_NPADVMOD;
    }
   
    if ((label = getSubjectLabel(C, d)) != null)
      return label;
   
    // coordination
    if (C.isPTag(CTLibEn.PTAG_UCP))
    {
      c.addFTags(C.getFTags());
      return getDEPLabel(c, P, p);
    }
   
    // complements
    if (P.isPTagAny(CTLibEn.PTAG_VP, CTLibEn.PTAG_SINV, CTLibEn.PTAG_SQ))
    {
      if (isAcomp(C))  return DEPLibEn.DEP_ACOMP;
      if ((label = getObjectLabel(C)) != null)  return label;
      if (isOprd(C))  return DEPLibEn.DEP_OPRD;
      if (isXcomp(C))  return DEPLibEn.DEP_XCOMP;
      if (isCcomp(C))  return DEPLibEn.DEP_CCOMP;
      if ((label = getAuxLabel(C)) != null)    return label;
    }
   
    if (P.isPTagAny(CTLibEn.PTAG_ADJP, CTLibEn.PTAG_ADVP))
    {
      if (isXcomp(C))  return DEPLibEn.DEP_XCOMP;
      if (isCcomp(C))  return DEPLibEn.DEP_CCOMP;
    }
   
    if (P.isPTagAny(CTLibEn.PTAG_NML, CTLibEn.PTAG_NP, CTLibEn.PTAG_WHNP))
    {
      if (isNfmod(C))  return isInfMod(C) ? DEPLibEn.DEP_INFMOD : DEPLibEn.DEP_PARTMOD;
      if (isRcmod(C))  return DEPLibEn.DEP_RCMOD;
      if (isCcomp(C))  return DEPLibEn.DEP_CCOMP;
    }
   
    if (isPoss(C, P))
      return DEPLibEn.DEP_POSS;
   
    // simple labels
    if ((label = getSimpleLabel(C)) != null)
      return label;
     
    // default
    if (P.isPTagAny(CTLibEn.PTAG_PP, CTLibEn.PTAG_WHPP))
    {
      if (p.getParent() == C.getParent())  // p and C are siblings
      {
        if (p.getSiblingId() < C.getSiblingId())
          return getPmodLabel(C, d);
      }
      else                // UCP
      {
        if (p.getFirstTerminal().getTerminalId() < C.getFirstTerminal().getTerminalId())
          return getPmodLabel(C, d);
      }
    }
   
    if (C.isPTag(CTLibEn.PTAG_SBAR) || isXcomp(C) || (P.isPTag(CTLibEn.PTAG_PP) && CTLibEn.isClause(C)))
      return DEPLibEn.DEP_ADVCL;
   
    if (C.isPTagAny(CTLibEn.PTAG_S, CTLibEn.PTAG_SINV, CTLibEn.PTAG_SQ, CTLibEn.PTAG_SBARQ))
      return DEPLibEn.DEP_CCOMP;
   
    if (P.isPTag(CTLibEn.PTAG_QP))
    {
      if (C.isPTagAny(CTLibEn.POS_CD))
        return DEPLibEn.DEP_NUMBER;
      else
        return DEPLibEn.DEP_QUANTMOD;
    }
   
    if (P.isPTagAny(CTLibEn.PTAG_NML, CTLibEn.PTAG_NP, CTLibEn.PTAG_NX, CTLibEn.PTAG_WHNP) || CTLibEn.isNoun(p))
      return getNmodLabel(C);
   
    if (c != null)
    {
      if ((label = getSimpleLabel(c)) != null)
        return label;
     
      if (d.isPTag(CTLibEn.POS_IN))
        return DEPLibEn.DEP_PREP;
     
      if (CTLibEn.isAdverb(d))
        return DEPLibEn.DEP_ADVMOD;
    }
View Full Code Here
862
863
864
865
866
867
868
869
870
871
872
    return null;
  }
 
  private String getSpecialLabel(CTNode C)
  {
    CTNode d = C.c2d.getDependencyHead();
   
    if (CTLibEn.isPunctuation(C) || CTLibEn.isPunctuation(d))
      return DEPLibEn.DEP_PUNCT;
   
    if (isIntj(C) || isIntj(d))
View Full Code Here
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
  private String getAuxLabel(CTNode node)
  {
    if (node.isPTagAny(CTLibEn.POS_MD, CTLibEn.POS_TO))
      return DEPLibEn.DEP_AUX;

    CTNode vp;
   
    if (CTLibEn.isVerb(node) && (vp = node.getNextSibling(CTLibEn.PTAG_VP)) != null)
    {
      if ((MPLibEn.isBe(node.form) || MPLibEn.isBecome(node.form) || MPLibEn.isGet(node.form)))
      {
        if (vp.containsTags("+VBN|VBD"))
          return DEPLibEn.DEP_AUXPASS;
       
        if (!vp.containsTags("+VB.*") && (vp = vp.getFirstChild(CTLibEn.PTAG_VP)) != null && vp.containsTags("+VBN|VBD"))
          return DEPLibEn.DEP_AUXPASS;
      }
     
      return DEPLibEn.DEP_AUX;
    }
View Full Code Here
951
952
953
954
955
956
957
958
959
960
961
962
963
964
 
  private boolean isAdv(CTNode C)
  {
    if (C.isPTag(CTLibEn.PTAG_ADVP) || CTLibEn.isAdverb(C))
    {
      CTNode P = C.getParent();
      int id = C.getSiblingId();
     
      if (P.isPTagAny(CTLibEn.PTAG_PP, CTLibEn.PTAG_WHPP) && id+1 == P.getChildrenSize() && P.getChild(id-1).isPTagAny(CTLibEn.POS_IN, CTLibEn.POS_TO))
        return false;

      return true;
    }
   
View Full Code Here
995
996
997
998
999
1000
1001
1002
1003
1004
1005
    if (curr.hasFTag(DEPLibEn.DEP_OPRD))
      return true;
   
    if (curr.isPTag(CTLibEn.PTAG_S) && !curr.containsTags(CTLibEn.PTAG_VP) && curr.containsTags("-"+CTLibEn.FTAG_PRD))
    {
      CTNode sbj = curr.getFirstChild("-"+CTLibEn.FTAG_SBJ);
      return sbj != null && sbj.isEmptyCategoryRec();
    }
   
    return false;
  }
View Full Code Here
TOP

Related Classes of com.clearnlp.constituent.CTNode

  • com.carrotsearch.hppc.IntArrayList
  • com.carrotsearch.hppc.IntOpenHashSet
  • com.clearnlp.conversion.AbstractC2DConverter
  • com.clearnlp.conversion.EnglishC2DConverter
  • com.clearnlp.conversion.KaistC2DConverter
  • com.clearnlp.experiment.GenerateOntoNotesCoref
  • com.clearnlp.experiment.Tmp
  • com.clearnlp.run.C2DConvertMulti
  • com.clearnlp.run.PBPostProcess
  • java.util.StringTokenizer
    • 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.