Examples of SegNode

Examples of org.ictclas4j.bean.SegNode

607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
            else
              break;
          }
        }
        if (end > start) {
          SegNode newsn = new SegNode();
          newsn.setRow(sns.get(start).getRow());
          newsn.setCol(sns.get(end - 1).getCol());
          newsn.setPos(pos);
          newsn.setWord(unknownFlags);
          newsn.setSrcWord(srcWord);
          double value = computePossibility(start, end - start + 1, sns);
          newsn.setWeight(value);
          segGraph.insert(newsn, true);
        }
      }
    }
  }
View Full Code Here

Examples of org.ictclas4j.bean.SegNode

623
624
625
626
627
628
629
630
631
632
633
    }
  }

  private int getBestTag(ArrayList<SegNode> sns, int index) {
    if (sns != null && index >= 0 && index < sns.size()) {
      SegNode sn = sns.get(index);
      return getBestTag(sn);

    }

    return -1;
View Full Code Here

Examples of org.ictclas4j.bean.SegNode

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
    if (sns != null) {
      int size = sns.size();

      // �����ǿ�ʼ�ͽ������
      for (int i = size - 1, j = 0; i >= 0; i--) {
        SegNode sn = sns.get(i);
        ArrayList<AdjoiningPos> allPos = sn.getAllPos();
        if (allPos != null && allPos.size() > j) {
          AdjoiningPos pos = allPos.get(j);
          pos.setBest(true);
          j = pos.getPrev();
        } else if (i + 1 < size - 1) {
          int tag = getBestTag(sns.get(i + 1));
          AdjoiningPos pos = new AdjoiningPos(tag, 0);
          pos.setBest(true);
          sns.get(i).addPos(pos);
        }
        // ����ô�����ĸ���֣��������û��������ҵʵ�������������øôʵĴ���
        if (sn.getPos() == POSTag.NOUN_LETTER || sn.getPos() == POSTag.NUM) {
          for (AdjoiningPos pos : allPos) {
            if (pos.isBest() && pos.getPos().getTag() > 0) {
              sn.setPos(pos.getPos().getTag());
              break;
            }
          }
        }
      }
View Full Code Here

Examples of org.ictclas4j.bean.SegNode

754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
  private double computePossibility(int startPos, int length, ArrayList<SegNode> sns) {
    double retValue = 0, posPoss;

    if (sns != null && unknownDict != null && context != null) {
      for (int i = startPos; sns != null && i < startPos + length && i < sns.size(); i++) {
        SegNode sn = sns.get(i);
        int bestTag = getBestTag(sn);
        if (bestTag != -1) {
          int gbkID = sn.getGbkID();// dictLib.getGBKID(sn.getSrcWord());
          int freq = unknownDict.getFreq(sn.getSrcWord(), bestTag, gbkID);
          posPoss = Math.log((double) (context.getFreq(sn.getPos()) + 1));
          posPoss += -Math.log((double) (freq + 1));
          retValue += posPoss;
        }
      }
    }
View Full Code Here

Examples of org.ictclas4j.bean.SegNode

20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
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
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
   * @param atoms
   * @return
   */
  public static SegGraph generate(ArrayList<Atom> atoms, DictLib dictLib) {
    SegGraph segGraph = null;
    SegNode sn = null;
    Atom atom = null;

    if (atoms != null && atoms.size() > 0 && dictLib != null) {
      segGraph = new SegGraph();
      Dictionary dict = dictLib.getCoreDict();

      // �ȰѷǺ����ַ��Ĵ���ʶ�����
      for (int i = 0; i < atoms.size(); i++) {
        atom = atoms.get(i);
        String word = atom.getWord();
        if (atom.getPos() == Utility.CT_CHINESE)
          sn = new SegNode(i, i + 1, 0, 0, atom.getWord());
        else {
          int pos = 0;
          double value = Utility.MAX_FREQUENCE;

          switch (atom.getPos()) {
          case Utility.CT_INDEX:
          case Utility.CT_NUM:
            pos = -POSTag.NUM;// 'm'*256
            word = Utility.UNKNOWN_NUM;
            value = 0;
            break;
          case Utility.CT_DELIMITER:
            pos = POSTag.PUNC;// 'w'*256;
            break;
          case Utility.CT_LETTER:
            pos = -POSTag.NOUN_LETTER;//
            value = 0;
            word = Utility.UNKNOWN_LETTER;
            break;
          case Utility.CT_SINGLE:// 12021-2129-3121
            if (Utility.getCharCount("+-1234567890", atom.getWord()) == atom.getLen()) {
              pos = -POSTag.NUM;// 'm'*256
              word = Utility.UNKNOWN_NUM;
            } else {
              pos = -POSTag.NOUN_LETTER;//
              word = Utility.UNKNOWN_LETTER;
            }
            value = 0;
            break;
          default:
            pos = atom.getPos();// '?'*256;
            break;
          }

          int gbkID = dictLib.getGBKID(word);
          sn = new SegNode(i, i + 1, pos, value, word);
          sn.setGbkID(gbkID);
        }

        sn.setSrcWord(atom.getWord());
        segGraph.insert(sn, true);
      }

      StringBuffer words = new StringBuffer();
      for (int i = 0; i < atoms.size(); i++) {
        int j = i + 1;
        words.delete(0, words.length());
        words.append(atoms.get(i).getWord());

        // ����ǡ��·ݡ�����Ҫ�ָ�
        boolean flag = false;
        if (j < atoms.size()) {
          Atom a2 = atoms.get(j);
          if ("��".equals(words.toString()) && "��".equals(a2.getWord())) {
            segGraph.delete(i, j);
            segGraph.delete(i + 1, j + 1);
            words.append(a2.getWord());
            flag = true;
            j++;
          }
        }

        SegAtom sa = null;
        String word = words.toString();
        int gbkID = dictLib.getGBKID(word);
        int wordMaxLen = dict.getWordMaxLen(word, gbkID);
        for (; j <= atoms.size() && word.length() < wordMaxLen; j++) {
          word = words.toString();
          sa = dict.getSegAtom(word, gbkID);
          if (sa != null) {
            // 1���ڣ�1999��ĩ
            // if (word.length() == 2 && segGraph.getSize() > 0) {
            // SegNode g2 = segGraph.getLast();
            // if (Utility.isAllNum(g2.getWord()) ||
            // Utility.isAllChinese(g2.getWord())
            // && (g2.getWord().indexOf("��") == 0 ||
            // g2.getWord().indexOf("��") == 0)) {
            // if ("ĩ���е�ǰ���".indexOf(words.substring(1)) != -1)
            // break;
            // }
            // }
            // ֻ��һ���Դʣ�������
            SegNode sg = null;
            if (sa.getPosCount() == 1) {
              Pos pos = sa.getPos(0);
              sg = new SegNode(i, j, pos.getTag(), sa.getTotalFreq(), word);
            } else
              sg = new SegNode(i, j, 0, sa.getTotalFreq(), word);
            sg.setSrcWord(word);
            sg.setGbkID(gbkID);
            segGraph.insert(sg, true);
          }

          if (j < atoms.size()) {
            String word2 = atoms.get(j).getWord();
View Full Code Here

Examples of org.ictclas4j.bean.SegNode

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
194
195
196
197
198
199
200
201
202
203
204
205
206
207
    if (seg != null && dict != null && biDict != null) {
      segGraph = new SegGraph();
      ArrayList<SegNode> sgs = seg.getSnList();

      for (int i = 0; sgs != null && i < sgs.size(); i++) {
        SegNode sg = sgs.get(i);
        if (sg.getPos() >= 0)
          curFreq = sg.getWeight();
        else {
          int gbkID = sg.getGbkID();// dictLib.getGBKID(sg.getWord());
          curFreq = dict.getFreq(sg.getWord(), 2, gbkID);
        }

        // �õ�������ֵ�͸���ֵ��ȵ�����Ԫ��
        ArrayList<SegNode> nextSgs = seg.getNextElements(i);
        for (SegNode graph : nextSgs) {
          String twoWords = sg.getWord();
          twoWords += Utility.WORD_SEGMENTER;
          twoWords += graph.getWord();
          int gbkID = sg.getGbkID();// dictLib.getGBKID(twoWords);

          // ��������������֮���ƽ��ֵ
          // -log{a*P(Ci-1)+(1-a)P(Ci|Ci-1)} Note 0<a<1
          int twoFreq = biDict.getFreq(twoWords, 3, gbkID);
          double temp = (double) 1 / Utility.MAX_FREQUENCE;
          double value = smoothParam * (1 + curFreq) / (Utility.MAX_FREQUENCE + 80000);
          value += (1 - smoothParam) * ((1 - temp) * twoFreq / (1 + curFreq) + temp);
          value = -Math.log(value);

          if (value < 0) {
            value += sg.getFreq();
          }

          SegNode sg2 = new SegNode();
          // �ָ���@ǰ�Ĵ��������е�λ��
          int wordIndex = getWordIndex(sgs, sg);
          sg2.setRow(wordIndex);

          // �ָ���@��Ĵ��������е�λ��
          wordIndex = getWordIndex(sgs, graph);
          sg2.setCol(wordIndex);
          sg2.setWord(twoWords);
          sg2.setPos(sg.getPos());
          sg2.setWeight(value);
          sg2.setGbkID(gbkID);
          segGraph.insert(sg2, false);
        }
      }
    }
    return segGraph;
View Full Code Here

Examples of org.ictclas4j.bean.SegNode

29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
  public SegGraph(ArrayList<SegNode> snList) {
    this.snList = snList;
  }

  public SegNode getElement(int row, int col) {
    SegNode result = new SegNode();
    result.setWeight(  Utility.INFINITE_VALUE );
    // if (row > m_nRow || col > m_nCol)
    // return null;

    int index = 0;
    if (snList != null) {
      if (isRowFirst) {
        for (int i = 0; i < snList.size(); i++, index++) {
          SegNode sg = snList.get(i);
          if (row != -1 && sg.getRow() < row || col != -1 && sg.getRow() == row && sg.getCol() < col)
            continue;
          else
            break;
        }
      } else {
        for (int i = 0; i < snList.size(); i++, index++) {
          SegNode sg = snList.get(i);
          if (col != -1 && sg.getCol() < col || row != -1 && sg.getCol() == col && sg.getRow() < row)
            continue;
          else
            break;
        }
      }

      // Find it and return the value
      if (index < snList.size()) {
        SegNode sg = snList.get(index);
        if ((sg.getRow() == row || row == -1) && (sg.getCol() == col || col == -1))
          result = sg;
      }
    }
    return result;
View Full Code Here

Examples of org.ictclas4j.bean.SegNode

77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
    if (sg != null) {
      if (snList == null)
        snList = new ArrayList<SegNode>();

      int i = 0;
      SegNode sgTemp = null;
      if (isRowFirst) {
        for (i = 0; i < snList.size(); i++) {
          sgTemp = snList.get(i);
          if (sgTemp.getRow() < sg.getRow() || sgTemp.getRow() == sg.getRow()
              && sgTemp.getCol() < sg.getCol())
            continue;
          else
            break;
        }
      } else {
        for (i = 0; i < snList.size(); i++) {
          sgTemp = snList.get(i);
          if (sgTemp.getCol() < sg.getCol() || sgTemp.getCol() == sg.getCol()
              && sgTemp.getRow() < sg.getRow())
            continue;
          else
            break;
        }
      }

      if (sgTemp != null && sgTemp.getRow() == sg.getRow() && sgTemp.getCol() == sg.getCol())
        sgTemp = sg;
      else if (i > 0)
        snList.add(i - 1, sg);

    }
View Full Code Here

Examples of org.ictclas4j.bean.SegNode

122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
    ArrayList<SegNode> result = null;

    if (snList != null && snList.size() > 0 && curIndex >= 0) {
      result = new ArrayList<SegNode>();
      for (int i = 0; i < snList.size(); i++) {
        SegNode sg = snList.get(i);
        if (isColFirst) {
          if (sg.getCol() == curIndex)
            result.add(sg);
        } else {
          if (sg.getRow() == curIndex)
            result.add(sg);
        }

      }
    }
View Full Code Here

Examples of org.ictclas4j.bean.SegNode

503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
            nPos = j;
            personName = "";
            // Get the possible person name
            while (nPos < j + patterns[k].length()) {
              SegNode sn = sns.get(nPos);
              if (sn.getPos() < 4
                  && unknownDict.getFreq(sn.getWord(), sn.getPos()) < Utility.LITTLE_FREQUENCY)
                personName += sn.getWord();
              nPos += 1;
            }
            if ("CDCD".equals(patterns[k])) {
              if (GetForeignCharCount(personName) > 0)
                j += patterns[k].length() - 1;
              continue;
            }

            SegNode usn = new SegNode();
            usn.setRow(sns.get(j).getRow());
            usn.setCol(sns.get(j + patterns[k].length() - 1).getCol());
            usn.setWord(unknownFlags);
            usn.setSrcWord(personName);
            double value = -Math.log(factor[k]) + computePossibility(j, patterns[k].length(), sns);
            usn.setPos(pos);
            usn.setValue(value);
            segGraph.insert(usn, true);

            j += patterns[k].length();
            bMatched = true;
          }
View Full Code Here
TOP
Copyright © 2018 www.massapi.com. 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.