probability model 0.0033135
discriminative model 0.003216971
generative model 0.003066754
latter model 0.00279594
accurate model 0.002789284
tive model 0.002781744
ative model 0.002773023
markov model 0.0027552419999999998
ity model 0.002748238
bility model 0.00274527
ability model 0.002738481
criminative model 0.002733689
dssn model 0.002729401
chosen model 0.002726138
model 0.00247906
standard training 0.002173333
training method 0.001992798
discriminative training 0.001990581
training data 0.0019876959999999997
probability models 0.0018926
network training 0.00188537
generative training 0.0018403640000000001
work training 0.0018195730000000001
discriminative models 0.001796071
language parsing 0.0017838490000000001
ing training 0.001764577
standard probability 0.001755103
training methods 0.0017450360000000002
previous models 0.001742211
training corpus 0.00172583
training parses 0.001723801
additional training 0.001692341
parsing models 0.0016871490000000002
other models 0.001679182
generative models 0.001645854
training techniques 0.001598236
same probability 0.0015948149999999999
training criteria 0.001593535
training times 0.00157325
discriminative probability 0.001572351
other words 0.001554295
training procedure 0.00154173
approximate training 0.001539842
training param 0.001539673
tion training 0.001536079
training process 0.001535596
natural language 0.0015264229999999998
appropriate training 0.001523194
training algorithms 0.001511366
first word 0.001436513
next word 0.001424069
language pars 0.001422364
generative probability 0.001422134
standard methods 0.001413029
previous results 0.001411529
set sentences 0.001407214
same set 0.0014014470000000001
linear models 0.001400826
ural language 0.001400788
word vocabulary 0.00138681
learning method 0.0013704020000000002
word pairs 0.001355526
probability distribution 0.00134499
gssn models 0.001344607
standard way 0.001343238
standard maximum 0.0013328349999999999
ity models 0.001327338
erative models 0.001320991
dgssn models 0.001313871
first probability 0.0012758259999999999
probability distributions 0.0012631349999999999
standard criteria 0.001261528
training 0.00125267
previous work 0.0012509539999999999
word vocabularies 0.001249002
word predictions 0.001248064
unknown words 0.0012410099999999999
conditional probability 0.00123264
discriminative methods 0.001230277
standard testing 0.00123022
second probability 0.001223417
probability estimation 0.001201768
joint probability 0.0012016940000000001
small set 0.001193883
discriminative optimization 0.001193076
same way 0.0011829499999999999
tional probability 0.001182288
standard measures 0.001176878
standard datasets 0.001172441
language 0.00115486
ditional probability 0.0011446030000000001
probability mass 0.001140126
such network 0.001130862
morphological features 0.001130315
ative probability 0.001128403
probability distri 0.001122839
learning methods 0.00112264
probability estimates 0.001114453
same criteria 0.0011012399999999999
test set 0.001099273
