joint word 0.00419042
word segmentation 0.004089711
chinese word 0.003912193
pos features 0.00380214
pos tag 0.00378796
word information 0.003767699
current word 0.003695122
number word 0.003693783
partial word 0.003662486
word length 0.003662056
baseline word 0.003645822
maximum word 0.0036261619999999996
word frequency 0.003621195
word bigram 0.003618546
word seg 0.003551238
combined word 0.003531323
word size 0.003523892
complete word 0.003504617
joint pos 0.00349873
separate word 0.0034975229999999998
whole word 0.0034969349999999996
word boundaries 0.0034828719999999997
last word 0.0034824829999999998
word segmentor 0.0034806539999999997
measure word 0.0034771629999999997
word bigrams 0.003460257
imum word 0.003458892
word lengths 0.003454788
word segmen 0.003453679
pure word 0.003451674
word frequencies 0.003449819
plete word 0.003449819
quent word 0.003449819
word fre 0.0034491509999999997
word initialization 0.0034491509999999997
marked word 0.0034491509999999997
pos tags 0.0034213539999999997
pos tagging 0.0033262709999999996
chinese pos 0.003220503
pos tagger 0.003174821
perceptron pos 0.0030810739999999996
pos information 0.0030760089999999998
current pos 0.003003432
baseline pos 0.002954132
english pos 0.002929209
important pos 0.002867588
overall pos 0.0028531769999999997
line pos 0.0028143269999999997
pos pattern 0.0028137419999999997
pure pos 0.002759984
rare pos 0.002759951
tween pos 0.0027580549999999997
tag features 0.00265984
joint model 0.0024731700000000002
segmentation model 0.002372461
tagging model 0.002300711
segmentation features 0.002269901
tagging features 0.0021981509999999998
linear model 0.002175074
perceptron model 0.002055514
joint training 0.00204701
joint segmentation 0.001966491
baseline model 0.001928572
chinese character 0.0019026429999999999
tag set 0.001894448
chinese words 0.0018900129999999998
character sequence 0.00188769
segmentation accuracy 0.001877872
joint models 0.001849832
joint approach 0.0018485910000000001
length features 0.001842246
tag dictionary 0.001836235
baseline tag 0.0018118320000000002
training method 0.001809252
category features 0.00180693
tagging accuracy 0.0018061219999999998
tag bigram 0.001784556
feature vector 0.001771248
ging model 0.0017703390000000001
joint linear 0.001769104
character information 0.001758149
tag trigram 0.0017519800000000002
morphological features 0.001751444
tron model 0.001749361
additional features 0.0017448820000000001
brid model 0.001736703
model outper 0.001732222
tistical model 0.001732222
tag seg 0.0017172480000000002
open features 0.001703019
baseline feature 0.0016973919999999998
joint system 0.0016913380000000001
chinese segmentation 0.001688264
current character 0.001685572
previous character 0.001677249
training data 0.001675611
ging features 0.001667779
joint problem 0.001662311
partial words 0.0016403059999999998
sible tag 0.001636852
