translation model 0.005284489999999999
language model 0.00396833
model score 0.003511082
network model 0.003424809
linear model 0.003363628
language translation 0.00334886
model parameter 0.00333435
model parameters 0.003317106
tion model 0.003301061
above model 0.003298483
lation model 0.003283392
generative model 0.0032231499999999997
adnn model 0.0032230469999999997
guage model 0.0032228789999999997
model och 0.003212328
ear model 0.003212328
translation models 0.0031826149999999997
model 0.00295198
translation performance 0.002906107
machine translation 0.0027980179999999998
translation probabilities 0.002793352
english translation 0.002793319
new translation 0.002771702
translation tasks 0.002764628
translation states 0.002762497
translation rules 0.002757998
translation state 0.002752082
translation candidate 0.0027518819999999998
linear translation 0.002744158
following translation 0.002732006
current translation 0.002725646
lexical translation 0.002683622
based translation 0.002657647
translation framework 0.002653673
partial translation 0.0026481169999999997
high translation 0.002645441
complex translation 0.002622343
chine translation 0.002597391
translation evalua 0.0025965579999999997
translation 0.00233251
feature vector 0.002184632
local features 0.0020956209999999998
additional features 0.002058158
embedding features 0.00205062
training corpus 0.002037257
potential features 0.002003515
arbitrary features 0.002000924
surface features 0.001993465
default features 0.001977613
word embedding 0.00196824
cal features 0.001961425
reasonable features 0.001956089
ding features 0.001954321
trary features 0.001954139
word penalty 0.001918721
corresponding word 0.001914481
new feature 0.001909952
adnn word 0.001879507
word embeddings 0.001877452
local feature 0.001875561
language models 0.001866455
additional feature 0.001838098
feature vectors 0.001828279
feature engineering 0.001799576
training method 0.001790257
other words 0.001693955
features 0.00169082
source language 0.001654585
training objective 0.001561471
joint training 0.001502166
feature 0.00147076
source sentence 0.001463374
rnn language 0.0013477200000000002
complex language 0.001306183
gram language 0.001287113
beam search 0.001264474
phrase pairs 0.001256771
same search 0.001230532
objective function 0.001214265
heuristic function 0.00120865
neural network 0.001207198
words 0.00119021
training 0.00118976
efficient decoding 0.0011701020000000001
weight vector 0.001163972
additive models 0.001157932
function sigmoid 0.001130678
sigmoid function 0.001130678
search state 0.001116635
target side 0.001099014
experiment results 0.001086825
decoding efficiency 0.001082684
exact decoding 0.001079109
bilingual tree 0.001078347
conjugate information 0.001072618
search heuristic 0.0010631590000000002
search strategy 0.001052949
neural networks 0.001047531
erarchical phrase 0.0010440850000000002
additive neural 0.001042196
