Package org.ictclas4j.bean

Examples of org.ictclas4j.bean.AdjoiningPos

317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
        sn = snList.get(i);
        posList = sn.getAllPos();
        for (int j = 0; posList != null && j < posList.size(); j++) {
          double minFee = 1000000;
          int minPrev = 100000;
          AdjoiningPos pos = posList.get(j);
          SegNode psn = snList.get(i - 1);
          ArrayList<AdjoiningPos> pposList = psn.getAllPos();
          for (int k = 0; pposList != null && k < pposList.size(); k++) {
            double temp = -Math.log(context
                .getPossibility(0, pposList.get(k).getPos().getTag(), pos.getPos().getTag()));
            temp += pposList.get(k).getValue();// Add the fees
            if (temp < minFee) {
              minFee = temp;
              minPrev = k;
            }
          }
          pos.setPrev(minPrev);
          pos.setValue(pos.getValue() + minFee);
        }
      }
    }
  }
View Full Code Here
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
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
                curWord = "��";
            }
          }

          if (sn.getPos() < 0) {
            AdjoiningPos pos = new AdjoiningPos( 0 , 0);
            sn.addPos(pos);
          } else {
            // ��unknownDict�ʵ���л�ȡ��ǰ�����д���
            SegAtom sa = unknownDict.getSegAtom(curWord, gbkID);
            for (int j = 0; sa != null && j < sa.getPosCount(); j++) {
              Pos pos = sa.getPos(j);
              double value = -Math.log((1 + pos.getFreq()));
              value += Math.log((context.getFreq(pos.getTag()) + sa.getPosCount() + 1));
              AdjoiningPos apos = new AdjoiningPos(pos , value);
              sn.addPos(apos);
            }

            if (Utility.SENTENCE_BEGIN.equals(curWord))
              sn.addPos(new AdjoiningPos( 100 , 0));

            else if (Utility.SENTENCE_END.equals(curWord))
              sn.addPos(new AdjoiningPos( 101 , 0));
            else {
              int freq = 0;
              sa = coreDict.getSegAtom(curWord, gbkID);
              if (sa != null) {
                double value = -Math.log((double) (1 + freq));
                value += Math.log((double) (context.getFreq(0) + sa.getPosCount()));
                sn.addPos(new AdjoiningPos( 0 , value));

              }
            }
          }
        } else {
          if (sn.getPos() > 0) {
            int tag = sn.getPos();
            double value = -Math.log(1 + sn.getFreq());
            value += Math.log(1 + context.getFreq(tag));
            if (value < 0)
              value = 0;
            sn.addPos(new AdjoiningPos( tag,  value));
          } else {
            if (sn.getPos() < 0) {
              sn.setPos(-sn.getPos());
              sn.addPos(new AdjoiningPos( -sn.getPos(),  sn.getFreq()));
            }
            SegAtom sa = coreDict.getSegAtom(curWord, gbkID);
            if (sa != null) {
              for (int j = 0; j < sa.getPosCount(); j++) {
                Pos pos = sa.getPos(j);
                double value = -Math.log(1 + pos.getFreq());
                value += Math.log(context.getFreq(pos.getTag()) + sa.getPosCount());
                sn.addPos(new AdjoiningPos(pos , value));
              }
            }
          }
        }

        if (sn.getAllPos() == null)
          guessPos(tagType, sn);

        // ���һ���ʽڵ��Ӧ��allPosΪnull����˵�����޷������ɴ�
        // ���Ĵ�������һ���ʵĴ���,���ǽ�����ʶ��ĩ##ĩ������
        if (i - 1 >= 0 && sns.get(i - 1).getPosSize() == -1) {
          if (sn.getPosSize() > 0) {
            Pos pos = sn.getAllPos().get(0).getPos();
            int ipos = pos.getTag() == POSTag.SEN_END ? POSTag.UNKNOWN : pos.getTag();
            AdjoiningPos apos = new AdjoiningPos( ipos , 0);
            sns.get(i - 1).addPos(apos);
          }
        }
      }

      // ���һ��������
      SegNode last = sns.get(i - 1);
      if (last != null) {
        SegNode sn = new SegNode();
        int tag = 0;
        if (tagType != Utility.TAG_TYPE.TT_NORMAL)
          tag = 101;
        else
          tag = 1;
        AdjoiningPos pos = new AdjoiningPos( tag, 0);
        sn.addPos(pos);
        sns.add(sn);
      }
    }
  }
View Full Code Here
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
    if (sns != null && context != null) {
      for (int i = 0; i < sns.size(); i++) {
        if (i == 0) {
          int pos = tagType != Utility.TAG_TYPE.TT_NORMAL ? 100 : 0;
          prevAllPos = new ArrayList<AdjoiningPos>();
          prevAllPos.add(new AdjoiningPos(pos, 0));
        } else {
          prevAllPos = sns.get(i - 1).getAllPos();
        }
        allPos = sns.get(i).getAllPos();
        if (allPos != null)
          for (AdjoiningPos pos : allPos) {
            // �ҳ�ǰһ�����Ժ͵�ǰ�������п����ڽӵĴ���
            int bestPrev = 0;
            double minValue = 10000000;
            for (int k = 0; prevAllPos != null && k < prevAllPos.size(); k++) {
              AdjoiningPos prevPos = prevAllPos.get(k);
              double temp = context.computePossibility(prevPos.getPos().getTag(), pos.getPos().getTag());
              temp = -Math.log(temp) + prevPos.getValue();
              if (temp < minValue) {
                minValue = temp;
                bestPrev = k;
              }
            }
View Full Code Here
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
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
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
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
449
      case TT_NORMAL:
        break;
      case TT_PERSON:
        if (word.indexOf("����") != -1) {
          freq = (double) 1 / (double) (context.getFreq(6) + 1);
          sn.addPos(new AdjoiningPos(6, freq));
        } else {
          freq = (double) 1 / (double) (context.getFreq(0) + 1);
          sn.addPos(new AdjoiningPos(0, freq));

          if (sn.getLen() >= 4) {
            freq = (double) 1 / (double) (context.getFreq(0) + 1);
            sn.addPos(new AdjoiningPos(0, freq));
            freq = (double) 1 / (double) (context.getFreq(11) * 8);
            sn.addPos(new AdjoiningPos(11, freq));
            freq = (double) 1 / (double) (context.getFreq(12) * 8);
            sn.addPos(new AdjoiningPos(12, freq));
            freq = (double) 1 / (double) (context.getFreq(13) * 8);
            sn.addPos(new AdjoiningPos(13, freq));
          } else if (sn.getLen() == 2) {
            freq = (double) 1 / (double) (context.getFreq(0) + 1);
            sn.addPos(new AdjoiningPos(0, freq));
            charType = Utility.charType(word);
            if (charType == Utility.CT_OTHER || charType == Utility.CT_CHINESE) {
              freq = (double) 1 / (double) (context.getFreq(1) + 1);
              sn.addPos(new AdjoiningPos(1, freq));
              freq = (double) 1 / (double) (context.getFreq(2) + 1);
              sn.addPos(new AdjoiningPos(2, freq));
              freq = (double) 1 / (double) (context.getFreq(3) + 1);
              sn.addPos(new AdjoiningPos(3, freq));
              freq = (double) 1 / (double) (context.getFreq(4) + 1);
              sn.addPos(new AdjoiningPos(4, freq));
            }
            freq = (double) 1 / (double) (context.getFreq(11) * 8);
            sn.addPos(new AdjoiningPos(11, freq));
            freq = (double) 1 / (double) (context.getFreq(12) * 8);
            sn.addPos(new AdjoiningPos(12, freq));
            freq = (double) 1 / (double) (context.getFreq(13) * 8);
            sn.addPos(new AdjoiningPos(13, freq));
          }
        }
        break;
      case TT_PLACE:
        freq = (double) 1 / (double) (context.getFreq(0) + 1);
        sn.addPos(new AdjoiningPos(0, freq));

        if (sn.getLen() >= 4) {
          freq = (double) 1 / (double) (context.getFreq(11) * 8);
          sn.addPos(new AdjoiningPos(11, freq));
          freq = (double) 1 / (double) (context.getFreq(12) * 8);
          sn.addPos(new AdjoiningPos(12, freq));
          freq = (double) 1 / (double) (context.getFreq(13) * 8);
          sn.addPos(new AdjoiningPos(13, freq));
        } else if (sn.getLen() == 2) {
          freq = (double) 1 / (double) (context.getFreq(0) + 1);
          sn.addPos(new AdjoiningPos(0, freq));
          charType = Utility.charType(word);
          if (charType == Utility.CT_OTHER || charType == Utility.CT_CHINESE) {

            freq = (double) 1 / (double) (context.getFreq(1) + 1);
            sn.addPos(new AdjoiningPos(1, freq));
            freq = (double) 1 / (double) (context.getFreq(2) + 1);
            sn.addPos(new AdjoiningPos(2, freq));
            freq = (double) 1 / (double) (context.getFreq(3) + 1);
            sn.addPos(new AdjoiningPos(3, freq));
            freq = (double) 1 / (double) (context.getFreq(4) + 1);
            sn.addPos(new AdjoiningPos(4, freq));
          }
          freq = (double) 1 / (double) (context.getFreq(11) * 8);
          sn.addPos(new AdjoiningPos(11, freq));
          freq = (double) 1 / (double) (context.getFreq(12) * 8);
          sn.addPos(new AdjoiningPos(12, freq));
          freq = (double) 1 / (double) (context.getFreq(13) * 8);
          sn.addPos(new AdjoiningPos(13, freq));
        }
        break;
      case TT_TRANS_PERSON:
        freq = (double) 1 / (double) (context.getFreq(0) + 1);
        sn.addPos(new AdjoiningPos(0, freq));
        if (!Utility.isAllChinese(word)) {
          if (Utility.isAllLetter(word)) {
            freq = (double) 1 / (double) (context.getFreq(1) + 1);
            sn.addPos(new AdjoiningPos(1, freq));
            freq = (double) 1 / (double) (context.getFreq(11) + 1);
            sn.addPos(new AdjoiningPos(11, freq));
            freq = (double) 1 / (double) (context.getFreq(2) * 2 + 1);
            sn.addPos(new AdjoiningPos(2, freq));
            freq = (double) 1 / (double) (context.getFreq(3) * 2 + 1);
            sn.addPos(new AdjoiningPos(3, freq));
            freq = (double) 1 / (double) (context.getFreq(12) * 2 + 1);
            sn.addPos(new AdjoiningPos(12, freq));
            freq = (double) 1 / (double) (context.getFreq(13) * 2 + 1);
            sn.addPos(new AdjoiningPos(13, freq));
          }
          freq = (double) 1 / (double) (context.getFreq(41) * 8);
          sn.addPos(new AdjoiningPos(41, freq));
          freq = (double) 1 / (double) (context.getFreq(42) * 8);
          sn.addPos(new AdjoiningPos(42, freq));
          freq = (double) 1 / (double) (context.getFreq(43) * 8);
          sn.addPos(new AdjoiningPos(43, freq));
        } else if (sn.getLen() >= 4) {
          freq = (double) 1 / (double) (context.getFreq(41) * 8);
          sn.addPos(new AdjoiningPos(41, freq));
          freq = (double) 1 / (double) (context.getFreq(42) * 8);
          sn.addPos(new AdjoiningPos(42, freq));
          freq = (double) 1 / (double) (context.getFreq(43) * 8);
          sn.addPos(new AdjoiningPos(43, freq));
        } else if (sn.getLen() == 2) {
          charType = Utility.charType(word);
          if (charType == Utility.CT_OTHER || charType == Utility.CT_CHINESE) {
            freq = (double) 1 / (double) (context.getFreq(1) * 2 + 1);
            sn.addPos(new AdjoiningPos(1, freq));
            freq = (double) 1 / (double) (context.getFreq(2) * 2 + 1);
            sn.addPos(new AdjoiningPos(2, freq));
            freq = (double) 1 / (double) (context.getFreq(3) * 2 + 1);
            sn.addPos(new AdjoiningPos(3, freq));
            freq = (double) 1 / (double) (context.getFreq(30) * 8 + 1);
            sn.addPos(new AdjoiningPos(30, freq));
            freq = (double) 1 / (double) (context.getFreq(11) * 4 + 1);
            sn.addPos(new AdjoiningPos(11, freq));
            freq = (double) 1 / (double) (context.getFreq(12) * 4 + 1);
            sn.addPos(new AdjoiningPos(12, freq));
            freq = (double) 1 / (double) (context.getFreq(13) * 4 + 1);
            sn.addPos(new AdjoiningPos(13, freq));
            freq = (double) 1 / (double) (context.getFreq(21) * 2 + 1);
            sn.addPos(new AdjoiningPos(21, freq));
            freq = (double) 1 / (double) (context.getFreq(22) * 2 + 1);
            sn.addPos(new AdjoiningPos(22, freq));
            freq = (double) 1 / (double) (context.getFreq(23) * 2 + 1);
            sn.addPos(new AdjoiningPos(23, freq));
          }
          freq = (double) 1 / (double) (context.getFreq(41) * 8);
          sn.addPos(new AdjoiningPos(41, freq));
          freq = (double) 1 / (double) (context.getFreq(42) * 8);
          sn.addPos(new AdjoiningPos(42, freq));
          freq = (double) 1 / (double) (context.getFreq(43) * 8);
          sn.addPos(new AdjoiningPos(43, freq));
        }
        break;
      default:
        break;
      }
View Full Code Here
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
      // �����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
TOP

Related Classes of org.ictclas4j.bean.AdjoiningPos

  • org.ictclas4j.segment.PosTagger
  • org.ictclas4j.segment.SegGraph
    • 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.