/
exp_l2s_predict.py
470 lines (317 loc) · 15.4 KB
/
exp_l2s_predict.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
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
133
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
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
246
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
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
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
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
# encoding=utf-8
__author__ = 'jma'
'''
Input:
1. the lexicon (the union of training lexcion + lexcicon extracted from *predicted* test corpus)
2. the raw corpus (a bunch of unsegmentd sentences)
3. scoring model that can evaluate
a implicit consistency required for the training code (exp_l2s_ins_to_train_score_model.py) is that both use the same
notabtion for labels, which are "T" and "F" to represent positive and negative labels.
'''
import math
import codecs
from lattice_build import gen_lattice # b_lattic_display
from viterbi_search import viterbi_search
from feature_gen import feature_gen
# from exp_l2s_gen_ins_to_train_score_model import gen_valid_state, gen_instance_by_traversal_lattice
from maxent import MaxentModel # need python interface of Zhang Le's maxent package
class ScoreModel(MaxentModel):
def __init__(self, path_me_model):
#print "initialization..."
super(ScoreModel, self).__init__()
self.load(path_me_model)
def score_it(self, bigram, incoming_char):
feature = feature_gen(bigram, incoming_char)
raw_score = self.eval([u_str.encode('utf-8') for u_str in feature],
"T") # maxent_model only takes utf-8 string as input
#print '\t\tRaw score=',raw_score
return math.log(raw_score, 10)
def test_single_sent():
#parameters
max_word_len = 15
dummy_start, dummy_end = u'$START#', u'$END#'
print '\nRunning test for exp_l2s_predict....'
path_me_model = "../working_data/train.set1.i80.model"
path_to_lexicon = "../working_data/train_testPredict.dict"
sent = u"材 料 利 用 率 高".split()
sent = u"下 雨 天 留 客 天 天 留 我 不 留".split()
print "sample sentence is:", " - ".join(sent)
#
# loading maximum entropy model as the score function
#
print '\nInitializing maximum entropy model as the scoring model'
model = ScoreModel(path_me_model)
print 'done'
#
# loading lexicion
#
print "\nLoading lexicion file..."
with codecs.open(path_to_lexicon, 'rU', 'utf-8') as f:
lexicon = [word for line in f for word in line.split()]
print "lexicion size=", len(lexicon), "example word in lexicion:", " ".join(lexicon[:5])
lexicon = set(lexicon)
print '\n====1 Bui latttice for the sample sentence====='
forward_unigram_lattice, backward_bigram_lattice = gen_lattice(lexicon, sent, max_word_len, dummy_start)
print '\n====2 Runing Viterbi search to decode===='
best_index_seq = viterbi_search(model.score_it, backward_bigram_lattice, sent, dummy_end)
#b_lattic_display(backward_bigram_lattice)
#f_lattice_display(forward_unigram_lattice)
x = best_index_seq[:-1]
y = best_index_seq[1:]
z = zip(x, y)
segmented = []
for index1, index2 in z:
word = u"".join(sent[index1:index2])
print word
segmented.append(word)
print '\nSegmented sent=', u" ".join(segmented)
def main(path_corpus, path_me_model, path_to_lexicon, path_to_output):
max_word_len = 12
dummy_start, dummy_end = u'$START#', u'$END#'
#
# loading maximum entropy model as the score function
#
print '\nInitializing maximum entropy model as the scoring model'
model = ScoreModel(path_me_model)
print 'done'
#
# loading lexicion
#
print "\nLoading lexicion file..."
with codecs.open(path_to_lexicon, 'rU', 'utf-8') as f:
lexicon = [word for line in f for word in line.split()]
print "lexicion size=", len(lexicon), "example word in lexicion:", " ".join(lexicon[:5])
lexicon = set(lexicon)
segmented_corpus = []
print "\nLoading corpus to be segmented..."
with codecs.open(path_corpus, 'rU', 'utf-8') as f:
raw_corpus = [u"".join(line.split()) for line in f]
print 'line count of raw_corpus=', len(raw_corpus)
print 'the first line is ', raw_corpus[0]
print "\n\n====Segmenting the corpus======"
for sent in raw_corpus:
#print '\n====1 Bui latttice for the sample sentence====='
forward_unigram_lattice, backward_bigram_lattice = gen_lattice(lexicon, sent, max_word_len, dummy_start)
#print '\n====2 Runing Viterbi search to decode===='
best_index_seq = viterbi_search(model.score_it, backward_bigram_lattice, sent, dummy_end)
#b_lattic_display(backward_bigram_lattice)
#f_lattice_display(forward_unigram_lattice)
x = best_index_seq[:-1]
y = best_index_seq[1:]
z = zip(x, y)
segmented = []
for index1, index2 in z:
word = u"".join(sent[index1:index2])
#print word
segmented.append(word)
print '\nSegmented sent=', u" ".join(segmented)
segmented_corpus.append(u" ".join(segmented))
print "\nSegmentation done, writing it to file", path_to_output, '...'
with codecs.open(path_to_output, 'w', 'utf-8') as f:
for sent in segmented_corpus:
f.write(sent + u'\n')
print 'done'
print "Program exit."
def test1():
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../working_data/train_testPredict.dict"
path_to_output = "../working_data/test.ctb5.seg.l2s"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test1_2():
print "Running test1_2, i.e. experiment Setting 1.2"
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../working_data/testPredict.dict"
path_to_output = "../working_data/test.ctb5.seg.l2s.1.2"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test2_1():
print "Running test2_2, i.e. experiment Setting 2.1"
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../working_data/test_crf_wordhood.dict"
path_to_output = "../working_data/test.ctb5.seg.l2s.2.1"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test2_2():
print "Running test2_2, i.e. experiment Setting 2.2"
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../working_data/train_test_crf_wordhood.dict"
path_to_output = "../working_data/test.ctb5.seg.l2s.22"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_fine():
print "Running fine, i.e. experiment with .fine (30i) merge-action prediciton model"
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../working_data/train_crf_wordhood.fine.dict"
path_to_output = "../working_data/test.ctb5.seg.fine"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_fine50():
print "Running fine, i.e. experiment with .fine.50i merge-action prediction model"
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../working_data/train_crf_wordhood.fine.50i.dict"
path_to_output = "../working_data/test.ctb5.seg.fine50"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_fine100():
print "Running fine, i.e. experiment with /fine.100i merge-action prediciton model"
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../working_data/train_crf_wordhood.fine.100i.dict"
path_to_output = "../working_data/test.ctb5.seg.fine100"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_fine15():
print "Running fine, i.e. experiment with /fine.15i merge-action prediciton model"
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../working_data/train_crf_wordhood.fine.15i.dict"
path_to_output = "../working_data/test.ctb5.seg.fine15"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_fine8():
print "Running fine, i.e. experiment with /fine.8i merge-action prediciton model"
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../working_data/train_crf_wordhood.fine.8i.dict"
path_to_output = "../working_data/test.ctb5.seg.fine8"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_fine2():
print "Running fine 2i..."
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../working_data/train_crf_wordhood.fine.2i.dict"
path_to_output = "../working_data/test.ctb5.seg.fine2"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_merge():
print "Running fine, i.e. experiment with /exp.merge/model.merge.o16"
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../exp.merge/o16.train.crf.dict"
path_to_output = "../exp.merge/test.ctb5.seg.o16"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_merge15():
print "Running fine, i.e. experiment with /exp.merge/model.merge.o16.i15"
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../exp.merge/o16.15i.train.crf.dict"
path_to_output = "../exp.merge/test.ctb5.seg.o16.15i"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_merge100():
print "Running fine, i.e. experiment with /exp.merge/model.merge.o16.100i"
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../exp.merge/o16.100i.train.crf.dict"
path_to_output = "../exp.merge/test.ctb5.seg.o16.100i"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_group():
print "Running fine, i.e. experiment with /exp.merge.group/model.merge.group.o16"
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../exp.merge.group/all.o16.train.crf.dict"
path_to_output = "../exp.merge.group/test.ctb5.seg.group"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_group_i15():
print "Running fine, i.e. experiment with /exp.merge.group/model.merge.group.o16.i15"
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../exp.merge.group/all.biye.group.i15.dict"
path_to_output = "../exp.merge.group/test.ctb5.seg.group.i15"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_group_no_train():
print "Running fine, i.e. experiment with /exp.merge.group/model.merge.group.o16"
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../exp.merge.group/group.crf.dict"
path_to_output = "../exp.merge.group/test.ctb5.seg.group.no_train"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_base():
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../exp.base/test.ctb5.segwp.model.dict"
path_to_output = "../exp.base/test.ctb5.seg.base"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_base50():
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../exp.base/test.ctb5.segwp.model.i50.dict"
path_to_output = "../exp.base/test.ctb5.seg.base.i50"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_base100():
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../exp.base/test.ctb5.segmodel.wp.i100.dict"
path_to_output = "../exp.base/test.ctb5.seg.base.i00"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
#test()
#test1_2()
#test2_1()
def test_base100_no_train():
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../exp.base/i100.crf.dict"
path_to_output = "../exp.base/test.ctb5.seg.base.i00"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_a2_100():
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../working_data/a2.i100.crf.dict"
path_to_output = "../working_data/a2.i100.seg"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_a2_150():
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../working_data/a2.i150.crf.dict"
path_to_output = "../working_data/a2.i150.seg"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_a2_200():
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../working_data/a2.i200.crf.dict"
path_to_output = "../working_data/a2.i200.seg"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_a2_100_nobase():
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../working_data/a2.nobase.100i.crf.dict"
path_to_output = "../working_data/a2.nobase.i100.seg"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_bichar_extend_150():
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../working_data/bichar.crf.e.i150.dict"
path_to_output = "../working_data/bichar.e.i150.seg"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_bichar_150():
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../working_data/bichar.crf.i150.dict"
path_to_output = "../working_data/bichar.i150.seg"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_bichar_e2_150():
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../working_data/bichar.e2.crf.i150.dict"
path_to_output = "../working_data/bichar.e2.i150.seg"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
def test_crf():
path_corpus = "../working_data/test.ctb5.seg"
path_me_model = "../working_data/train.set1.i100.model"
path_to_lexicon = "../working_data/test.crf.dict"
path_to_output = "../working_data/test.crf.seg"
main(path_corpus, path_me_model, path_to_lexicon, path_to_output)
# #TODO code to handle "fail to segment" error (make each char a possible wor
if __name__ == '__main__':
#test_merge()
#test_merge100()
#test_merge15()
# test_group_no_train()
#test_base()
#test_base50()
#test_base100_no_train()
# test_a2_100()
#test_a2_150()
#test_a2_200()
#test_crf()
#test_a2_100_nobase()
# test_bichar_extend_150()
#test_bichar_150()
test_bichar_e2_150()