first word 0.002449444
features table 0.002371841
function features 0.002370564
last word 0.002307727
previous word 0.002263168
word index 0.002243227
current word 0.00222931
independent features 0.002218996
structural features 0.002180378
next word 0.002174588
expressive features 0.002163323
joint model 0.0021256770000000003
word bigrams 0.00211858
end word 0.0020865709999999997
training data 0.002005574
labeling model 0.001987953
feature vector 0.001977981
same feature 0.001967804
pos tag 0.001943643
generative model 0.001915114
features 0.00189787
training algorithm 0.0018711539999999999
different chunk 0.001867695
tag sequence 0.001798155
feature templates 0.001778744
same chunk 0.0017786540000000002
training corpus 0.0017474359999999998
chunk tags 0.001692644
feature representation 0.00168842
effective feature 0.0016558839999999998
level feature 0.001640134
particular feature 0.001639102
feature template 0.001633151
last chunk 0.001628647
local feature 0.001624421
chinese sentence 0.0016228599999999998
learning algorithm 0.0016164109999999999
feature representations 0.001609848
chinese corpus 0.0016086999999999998
key feature 0.001608193
structural feature 0.001602988
chunk size 0.001598001
global feature 0.001595752
chunk type 0.001587089
feature maps 0.001585417
previous chunk 0.001584088
training set 0.001562612
possible chunk 0.001559876
chunk length 0.001554434
current chunk 0.00155023
pos tags 0.0015289230000000002
other words 0.0015259359999999999
maximum chunk 0.001522205
training time 0.001506559
chunk boundaries 0.001505715
margin training 0.001502466
chunk candidate 0.001492251
learning models 0.001487913
corresponding chunk 0.001480498
chunk types 0.001475884
tag set 0.001472836
model 0.00145563
single chunk 0.001445819
chunk candidates 0.001439352
few chunk 0.001427155
valid chunk 0.001426919
preceding chunk 0.001426671
chunk identification 0.001411457
invalid chunk 0.001402647
online training 0.0013919919999999999
training examples 0.0013644759999999999
chinese phrase 0.001358452
training iterations 0.001355112
sequence labeling 0.001354444
joint segmentation 0.001346276
comparable training 0.001343951
chinese chunking 0.001337883
output sequence 0.0013350559999999998
feature 0.00132048
data representation 0.001307704
testing data 0.001300557
tag dictionary 0.001284662
previous learning 0.0012638249999999999
learning task 0.0012545619999999999
end pos 0.00124377
bio tag 0.0012406869999999999
corresponding label 0.001231926
chinese treebank 0.001229674
data sets 0.001217223
labeling models 0.001209169
conll data 0.001205286
significant performance 0.001204563
joint problem 0.0012024660000000001
different systems 0.001200541
single label 0.001197247
input sequence 0.00119047
transition information 0.001186609
joint system 0.001178447
test corpus 0.001177497
learning methods 0.001177309
