same model 0.002804677
probability model 0.002755386
supervised model 0.002727883
crf model 0.002589166
model size 0.002568577
training data 0.002565798
unlabeled data 0.002565025
markov model 0.002542119
probabilistic model 0.002447819
labeled data 0.002440173
model parameters 0.002428352
graphical model 0.002395604
model representations 0.002392941
mrf model 0.002357185
such data 0.0023565269999999998
feature function 0.002124688
model 0.00211729
much data 0.002051755
classifieds data 0.002020418
fieds data 0.002013898
feature functions 0.001745938
label information 0.001709375
same label 0.001666509
labeled features 0.001631812
feature vector 0.001626634
word type 0.001543078
supervised learning 0.0015407160000000001
other words 0.001488652
single word 0.001445808
learning results 0.0014323970000000002
learning objective 0.0014219459999999999
label prototype 0.0013754280000000002
additional features 0.001369142
word collocation 0.001364482
accuracy performance 0.0013417260000000001
prototype words 0.001332384
training time 0.001317555
such information 0.0013103400000000001
learning methods 0.001289894
training objective 0.001281181
function form 0.001280141
machine learning 0.0012756920000000001
example words 0.001271625
feature 0.00127007
distance function 0.001266163
useful features 0.001266018
same set 0.001261788
state sequence 0.001258458
conditional models 0.001232377
features kitchen 0.0012277400000000002
label consistency 0.0012217
joint probability 0.001213018
good accuracy 0.001203743
same way 0.0012031490000000001
learning setting 0.001202623
labeling information 0.0011885300000000001
stop words 0.001181437
ample words 0.001176119
driven learning 0.001173056
pervised learning 0.0011702940000000001
partition function 0.001164244
same crf 0.001159263
unlabeled examples 0.001151919
unlabeled dataset 0.001134934
normalization function 0.001128756
different graph 0.0011270009999999999
prototype information 0.001126559
text vectors 0.001121469
conditional probability 0.001117984
decision function 0.001103475
test time 0.001096449
maximum accuracy 0.00109263
unlabeled instances 0.001089299
supervised crf 0.001082469
accuracy improvement 0.001066441
inference method 0.001066082
natural way 0.001063202
conditional entropy 0.001061504
training settings 0.001060339
input sequence 0.001042292
unlabeled instance 0.001042101
necessary training 0.001035937
graphical models 0.0010308029999999998
labeled examples 0.001027067
cient training 0.001026861
ical models 0.001011251
state crf 0.001008044
similar type 0.001003551
sparse information 0.001001906
sequential models 9.996809999999999E-4
probability distribution 9.95384E-4
same collocation 9.88729E-4
further accuracy 9.8161E-4
label 9.79122E-4
conditional random 9.78305E-4
such prototype 9.76393E-4
other values 9.7508E-4
significant accuracy 9.74405E-4
such weights 9.693569999999999E-4
hypothesis sequence 9.684030000000001E-4
