feature vectors 0.00203371
training data 0.0017701610000000001
feature 0.00176228
standard features 0.001728342
different sequence 0.00170999
features space 0.001595476
training set 0.0015910339999999999
parameter learning 0.001565718
training corpus 0.001557119
sequence probability 0.001538144
viterbi algorithm 0.0015001910000000001
annotated sequence 0.0014878909999999999
possible sequence 0.001473414
different data 0.0014720800000000002
sequence labeling 0.001449198
labeled sequence 0.001433326
token annotation 0.001420944
label information 0.001415857
maximum sequence 0.001394265
algorithm the 0.0013859780000000002
training examples 0.001365433
few training 0.001353771
probable sequence 0.001326266
entire sequence 0.00132019
new label 0.001304575
sequence assignment 0.001304355
sequence assignments 0.001298023
whole sequence 0.001295668
sequence probabili 0.001286559
dard sequence 0.001286436
sequence probabil 0.001283194
sociated sequence 0.001282243
features 0.00126488
human annotation 0.001259118
label sequences 0.0012566489999999999
other methods 0.001256517
annotated data 0.001249981
initial training 0.001231781
positive annotation 0.001229281
other tokens 0.0012258149999999999
conditional random 0.001225351
entire training 0.001216661
positive label 0.0012137229999999999
annotation effort 0.001207062
label estimation 0.001185672
partial label 0.0011854909999999999
learning strategies 0.0011841970000000001
man annotation 0.001182052
active learning 0.001181218
entire annotation 0.00116856
annotation unit 0.0011653190000000002
negative label 0.001165204
annotation cost 0.001150006
special label 0.001148615
different labels 0.001139372
label assignment 0.0011371670000000001
algorithm 0.00113607
mum annotation 0.001133308
basic annotation 0.001132261
kens annotation 0.001132261
random fields 0.001131661
other strategies 0.001129585
crf model 0.001128211
learning algorithms 0.001126147
signed label 0.001115557
learning framework 0.001115506
good performance 0.001106519
standard set 0.001102225
likelihood function 0.001075483
sequence 0.0010558
data statistics 0.001051574
very function 0.0010507910000000001
data heterogeneity 0.0010449860000000001
data characteristics 0.0010449860000000001
other datasets 0.001020827
same dataset 0.001007747
tional random 0.001007218
other ones 0.001005427
test set 0.001000111
token probability 9.99118E-4
unlabeled set 9.97731E-4
entity recognition 9.89366E-4
random drops 9.856399999999999E-4
recognition task 9.74403E-4
labeling task 9.73373E-4
token entropy 9.627220000000001E-4
training 9.52271E-4
same sequences 9.49818E-4
conditional probability 9.4962E-4
new information 9.43208E-4
different datasets 9.41737E-4
annotated tokens 9.24626E-4
large amount 9.22841E-4
standard results 9.12842E-4
optimization method 9.05795E-4
annotation 9.0417E-4
possible labels 9.02796E-4
entropy approach 9.004740000000001E-4
labeling tasks 8.98147E-4
few tokens 8.940350000000001E-4
