character word 0.00392783
word segmentation 0.0038333300000000002
chinese word 0.003546336
joint word 0.00346095
word dictionary 0.003246052
word table 0.003223118
word seg 0.0031978650000000003
basic word 0.0031971630000000003
discriminative word 0.003183374
current word 0.0031732030000000003
word boundaries 0.0031430720000000002
word boundary 0.003142902
word segmen 0.003123789
nese word 0.0031131880000000002
word hypotheses 0.003104171
ous word 0.0030980450000000002
word segmenta 0.003097289
word basis 0.003095651
unknown words 0.0026868129999999997
hybrid model 0.002468834
baseline model 0.002468205
other words 0.002467032
markov model 0.002372552
model weight 0.002360272
entropy model 0.00235184
brid model 0.002269837
trained model 0.002265314
model yields 0.002264274
oov words 0.002238041
frequent words 0.002209715
known words 0.002207904
infrequent words 0.002170321
pos tag 0.00213297
pos tagging 0.0020361809999999998
training corpus 0.001993433
model 0.00199213
pos tags 0.0019791369999999997
words 0.00188938
training data 0.001792465
simple pos 0.001744013
training method 0.00172684
regular pos 0.001720961
pos bigrams 0.001714008
pos category 0.001684385
character sequence 0.001682429
assigned pos 0.001674909
own pos 0.001674909
character type 0.0016346020000000002
class features 0.0016110080000000001
corpus data 0.001608098
bigram features 0.001600804
training set 0.00159158
character hybrid 0.0015822840000000002
character types 0.001538008
general character 0.001498105
large training 0.0014929680000000001
generalized features 0.0014848580000000001
small training 0.001444687
unigram features 0.001435307
end character 0.001428795
training example 0.001420836
training cor 0.001420737
test data 0.001419519
beginning character 0.00140515
intermediate character 0.0014001260000000002
igram features 0.001395963
formative features 0.001395963
features appro 0.001395963
bitrary features 0.001395963
training iterations 0.0013909060000000001
whole training 0.0013793170000000001
character basis 0.001378981
learning algorithm 0.001376235
training times 0.001375574
training phase 0.00137267
training sam 0.0013628160000000002
moderate training 0.0013611080000000002
main corpus 0.001324046
chinese characters 0.001318268
feature templates 0.001299253
corpus experiments 0.00127164
new approach 0.001263111
learning problem 0.00126091
hybrid approach 0.001229349
corpus size 0.001226325
test set 0.0012186340000000001
annotated corpus 0.001196104
machine learning 0.001171149
ing data 0.0011259249999999998
different studies 0.001122834
output nodes 0.001122765
features 0.0011226
different policies 0.001112305
system output 0.001109781
online learning 0.001108848
character 0.00110558
training 0.0010889
optimal performance 0.001085864
chinese treebank 0.001075441
test sets 0.0010714
