pos tag 0.00320419
pos errors 0.00306196
word features 0.003029378
training pos 0.00278715
pos information 0.002728396
other pos 0.002716914
pos tagger 0.002689985
pos tagging 0.0026537310000000003
pos tags 0.002621528
same pos 0.002591209
category pos 0.00247056
pos category 0.00247056
word boundary 0.002429773
similar pos 0.002374996
supervised pos 0.0023714490000000003
description word 0.002336363
pos taggers 0.002309004
pos experiment 0.002307591
pos substantive 0.002305644
minor pos 0.002304043
bank pos 0.002292429
similarity pos 0.0022827200000000002
ous pos 0.0022817460000000003
pos informa 0.002278517
different error 0.0021510540000000003
tag features 0.002129568
training data 0.0020379
data feature 0.002010651
different tag 0.0019640639999999997
error type 0.0019244730000000001
different errors 0.001821834
data set 0.001796508
category error 0.0017953
test data 0.001766438
error analysis 0.0017527460000000001
error rate 0.0017490070000000001
overall error 0.001740467
error types 0.001714505
tag set 0.0016835079999999998
lexical features 0.001678272
error rates 0.001667307
tagging errors 0.001649251
minor error 0.001628783
experimental data 0.001627043
serious error 0.00161802
development data 0.0016177259999999999
previous words 0.001610367
tag category 0.00160831
data sets 0.001602965
error differentiation 0.0016016980000000001
intra error 0.0016016980000000001
inter error 0.0016016980000000001
dard data 0.00152894
correct tag 0.001478817
learning method 0.0014747229999999998
category errors 0.00146608
tag categories 0.0014558269999999998
tag fea 0.0014473709999999998
serious tag 0.0014310299999999998
bank tag 0.001430179
entropy model 0.001384428
error 0.00135796
machine learning 0.001343705
chunking model 0.00133004
total errors 0.001325779
errors reduction 0.001324024
learning methods 0.0013214989999999999
model complexity 0.001310659
minor errors 0.001299563
ging errors 0.001290731
serious errors 0.0012888
tree loss 0.001286082
same feature 0.00128467
iob model 0.001276439
rious errors 0.001271834
tagger accuracy 0.00126618
combination features 0.0012492509999999998
loss function 0.001248652
learning svms 0.001245485
learning machines 0.001243572
learning techniques 0.001235886
linguistic features 0.0012355489999999998
sequential features 0.001230281
tive features 0.001216893
tagging problem 0.001203111
fix features 0.001202467
informative features 0.001202467
learning community 0.0011998690000000001
same performance 0.001199411
sensitive learning 0.001198961
sitive learning 0.001192898
different types 0.001149639
category loss 0.001136929
words 0.00113543
new loss 0.001126695
additional loss 0.001125347
different cost 0.001111018
high loss 0.001107825
loss class 0.001085763
combined feature 0.00108002
