probability model 0.003581219
corpus model 0.003496931
model classifier 0.003425581
hmm model 0.003301162
markov model 0.003244781
tagging model 0.003161781
model overall 0.003151789
chunking model 0.003105267
model prec 0.003095639
guage model 0.003089531
trigram model 0.003085234
hybrid model 0.00308037
complicated model 0.003078715
other word 0.00307237
word segmentation 0.003034075
chinese word 0.002997605
word segmenter 0.002895143
model 0.0028412
errors word 0.002800951
word sense 0.0027572029999999997
single word 0.002744056
word con 0.002689555
word boundaries 0.002679163
word seg 0.002636233
rect word 0.002625131
ieer word 0.002624774
word trigram 0.002623994
word granularity 0.002623947
word segmen 0.00262279
nese word 0.002619901
training data 0.002502909
same data 0.002270683
test data 0.002265354
data size 0.0021261830000000002
linguistic data 0.002003759
data sets 0.00199096
ment data 0.001980237
experiment data 0.001979109
development data 0.0019685930000000003
ieer data 0.001965424
opment data 0.00195904
velopment data 0.00195807
data consortium 0.00195807
lexical features 0.001830413
good features 0.001626003
such features 0.0015830430000000001
conditioning features 0.0014105
features capitalization 0.0014105
other words 0.001395955
language learning 0.0013897990000000002
training set 0.001389109
system performance 0.0012707
feature type 0.001240137
probability distribution 0.0012381789999999998
learning classifier 0.001234336
important feature 0.001231499
classification method 0.001208386
ing training 0.001195166
training size 0.001187872
features 0.00117183
evaluation set 0.0011668450000000001
class probability 0.001163772
test set 0.001151554
different ways 0.001148534
speech tags 0.0011374830000000001
conditional probability 0.00113583
speech recognition 0.001129825
entity boundary 0.001122616
natural language 0.001119911
training examples 0.001114264
training time 0.00110289
test test 0.001089488
corpus type 0.001083466
other types 0.00108228
recognition chinese 0.001081834
sequence clas 0.001078286
segmentation errors 0.0010751060000000002
classification distribution 0.001073315
recognition system 0.001071282
other characters 0.001070699
probability interpolation 0.001070044
hmm system 0.001067055
feature vector 0.001064941
chinese text 0.001057335
ble sequence 0.001055643
feature templates 0.001051859
useful feature 0.001051842
ful feature 0.001050664
hmm classifier 0.001044343
classification task 0.0010414690000000002
entity recognition 0.001038469
tag unigram 0.001035892
tag bigram 0.001032367
other examples 0.001024375
combination performance 0.001022656
mandarin training 0.0010198289999999999
source information 0.001013472
probability distributions 0.0010049149999999999
experimental results 0.001002862
correct classification 9.98737E-4
