language model 0.0036249999999999998
model sentence 0.00345161
model sentences 0.003139107
model distribution 0.002937196
baseline model 0.002820319
current model 0.0027681010000000002
probabilistic model 0.002762622
real model 0.0027152210000000003
final model 0.002660575
wsme model 0.002653616
model perplexity 0.0026293090000000002
guage model 0.002625118
true model 0.002624902
rent model 0.0026174090000000002
model 0.00238783
previous features 0.0020404
language models 0.0019182309999999998
training language 0.001869272
corresponding features 0.0018534990000000002
current word 0.0018533409999999999
language corpus 0.0018283409999999998
trained features 0.0018258780000000002
word types 0.001801572
word position 0.0017983139999999999
next word 0.001781481
word sequences 0.001714705
other sentence 0.0016711030000000002
sentence length 0.001631725
generative language 0.0016302259999999998
discriminative language 0.0016079699999999998
language modeling 0.0015941809999999999
features 0.001592
natural language 0.0015884329999999998
statistical language 0.0015842869999999998
new feature 0.001581423
first sentence 0.001542268
single feature 0.00153798
next feature 0.001515721
language processing 0.0015029789999999998
optimal feature 0.0015023530000000001
large sentence 0.001496285
language mod 0.0014801959999999998
initial language 0.001479273
average sentence 0.0014738630000000002
ral language 0.0014671999999999999
sulting language 0.0014665729999999999
sample sentence 0.00143831
training data 0.001415735
real sentence 0.001391171
sentence rejection 0.0013695550000000002
other words 0.001363675
sparse sentence 0.0013294460000000002
sentence representation 0.001316752
probability distribution 0.0013104050000000002
different sampling 0.001296776
language 0.00123717
training corpus 0.001223273
feature 0.00120731
probability value 0.0012059570000000001
global information 0.001152641
similar sentences 0.001139557
training set 0.0011319809999999998
real data 0.001111024
different approach 0.001102235
certain probability 0.001090005
real sentences 0.001078668
probability values 0.001077822
rejection probability 0.0010668140000000001
sentence 0.00106378
adjacent words 0.0010389829999999998
training methods 0.001018072
machine translation 0.0010160030000000001
probability mass 0.001015138
synthetic data 0.001013582
learning algorithm 0.001012703
other methods 9.932930000000001E-4
erated sentences 9.88838E-4
instance sentences 9.82686E-4
possible way 9.68345E-4
entropy models 9.66326E-4
correct translation 9.52337E-4
online algorithm 9.385769999999999E-4
same sort 9.31266E-4
conditional probabilities 9.11015E-4
abilistic models 9.109589999999999E-4
test set 9.09675E-4
experimental results 9.00487E-4
other forms 8.85301E-4
experimental work 8.79592E-4
good results 8.75345E-4
new set 8.739919999999999E-4
empirical distribution 8.57283E-4
first sample 8.53018E-4
decision function 8.46561E-4
future work 8.44445E-4
sampling process 8.34507E-4
possible values 8.314749999999999E-4
rejection probabilities 8.29051E-4
guage corpus 8.28459E-4
new class 8.24609E-4
