word pos 0.00505074
character word 0.00429692
segmentation word 0.00389988
word segmentation 0.00389988
japanese word 0.003830252
unknown word 0.0038249950000000003
word dictionary 0.003725752
chinese word 0.003652245
word boundaries 0.0036445880000000003
word processing 0.00357455
word forms 0.0035576830000000003
word seg 0.0035247760000000003
known word 0.003485952
word segmen 0.003482345
word rate 0.0034621120000000003
word segmenta 0.003432363
word dictionaries 0.003430175
scoring word 0.0034292230000000003
pos tag 0.002829851
such words 0.002751743
unknown words 0.002693455
pos tags 0.002626291
pos tagging 0.0026083160000000003
ing words 0.002416006
known words 0.0023544119999999997
english pos 0.002326842
hierarchical pos 0.002258619
words 0.00197813
corpus number 0.0016970700000000002
single character 0.001659042
training data 0.001638034
training test 0.001612187
character types 0.0015814359999999999
corpus version 0.001569091
test data 0.001559237
corpora corpus 0.001558523
treebank corpus 0.001525116
japanese data 0.001513124
university corpus 0.001508782
edr corpus 0.001474093
kuc corpus 0.0014580700000000001
corpus scoring 0.001456983
rwcp corpus 0.0014561280000000001
pfr corpus 0.0014561280000000001
chinese data 0.001335117
statistical information 0.001321888
combined method 0.001307395
processing method 0.001303275
hybrid method 0.001282154
other methods 0.001268711
this method 0.001234794
method this 0.001234794
overall accuracy 0.0012233679999999999
level information 0.001219082
other processing 0.001217036
markov models 0.00120095
character 0.00118725
syntactic parsing 0.00117581
other corpora 0.001173249
first step 0.001170623
brid method 0.001162092
entropy models 0.001159005
same time 0.001152758
japanese corpora 0.001141675
corpus 0.00113743
number number 0.00111928
probabilistic models 0.0011177890000000001
high accuracy 0.001117077
previous work 0.001109382
important task 0.001100649
test ctb 0.001088973
combined approach 0.001039287
input sentence 0.001036099
poc tags 0.001004428
common step 9.650500000000001E-4
chinese corpora 9.63668E-4
large problem 9.60056E-4
many methods 9.47481E-4
chinese treebank 9.302609999999999E-4
possible candidates 9.2409E-4
same tendency 9.189059999999999E-4
ducted experiments 8.9202E-4
language processing 8.889939999999999E-4
penn chinese 8.883820000000001E-4
performance 8.8189E-4
later experiments 8.76798E-4
information 8.56587E-4
training 8.45492E-4
method 8.38395E-4
splitting methods 8.38146E-4
base forms 8.20657E-4
following sections 8.18796E-4
random fields 8.13688E-4
experimental settings 8.11237E-4
labeling problem 8.03223E-4
features 7.9815E-4
processing ging 7.94819E-4
segmentation 7.9021E-4
statistical signifi 7.88097E-4
juman version 7.865050000000001E-4
