unlabeled data 0.0024790430000000002
domain data 0.002433587
test data 0.002414354
training data 0.002354779
labeled data 0.002262515
same data 0.002258094
target data 0.00206887
data sets 0.002061122
input data 0.002008101
source data 0.001996667
question data 0.001994772
biomedical data 0.001960229
data sparsity 0.001891449
new features 0.001807952
function word 0.001739214
context features 0.001677149
ing word 0.0016753850000000002
supervised model 0.0016745800000000002
local features 0.001653002
feature function 0.001607924
sequence model 0.001607749
learning algorithm 0.001560925
other words 0.001553423
suffix features 0.001509472
word position 0.0014973669999999999
word order 0.00148604
known features 0.001482443
pos tag 0.001479822
discriminative model 0.001475203
many words 0.00147266
unknown word 0.001466627
domain training 0.001466466
right word 0.001446731
center word 0.001434161
pos tags 0.001423746
word rate 0.001396801
word identity 0.001392018
word occurrence 0.00139025
test set 0.001389082
such domain 0.001389028
standard graph 0.001384103
underlying model 0.00138157
graphical model 0.001373078
graph label 0.001361791
crf training 0.001356679
model ter 0.001349669
feature weights 0.001348981
unlabeled sentence 0.001340403
similarity graph 0.001339366
learning problems 0.001338325
feature trigram 0.00133366
training set 0.001329507
set performance 0.001319548
similar pos 0.0013181360000000001
pos tagging 0.001312186
optimal feature 0.001304298
description feature 0.001303294
graph vertices 0.001302487
labeled training 0.001295394
feature templates 0.001263049
different sentences 0.001258038
supervised training 0.001257709
graph con 0.001254323
features 0.00125208
main graph 0.0012359390000000001
same set 0.001232822
graph propagation 0.001226887
supervised crf 0.00122673
unlabeled target 0.001226013
new tags 0.001221163
ssl algorithm 0.0012193080000000001
new crf 0.001218722
tag sequence 0.001218416
sequence tag 0.001218416
graph regularizer 0.001216132
training sentences 0.001211483
ing training 0.001208154
different label 0.0012050099999999998
graph node 0.001199971
graph regular 0.001198922
learning process 0.001191841
new set 0.00119155
graph nodes 0.001189441
graph regularization 0.001185168
training objective 0.0011837269999999999
target domain 0.001180557
tag distribution 0.0011801469999999999
graph size 0.001170642
domain accuracy 0.001166206
other ssl 0.0011625820000000001
target test 0.001161324
graph smoothness 0.0011529140000000001
convex graph 0.0011520480000000001
standard crf 0.0011497880000000001
sparse graph 0.0011484120000000001
different target 0.001148304
different vertices 0.001145706
construction graph 0.001139178
graph construction 0.001139178
domain adaptation 0.001139015
