target word 0.00424572
translation model 0.00417679
source word 0.00417602
word probability 0.0034809190000000003
translation models 0.00340515
ing word 0.003300028
word alignments 0.0032391520000000004
word order 0.003231143
word level 0.003213941
last word 0.0032061420000000004
word input 0.003203324
history word 0.00317579
word class 0.003148992
word classes 0.003082133
unique word 0.003077064
word cluster 0.003061115
word embeddings 0.003055614
gram word 0.003049533
word probabil 0.003045222
reverse word 0.003042978
neural translation 0.003002689
translation probability 0.002959049
translation table 0.002958101
different translation 0.0029456639999999997
machine translation 0.002833489
english translation 0.002811707
target phrase 0.00272135
translation task 0.00270728
translation probabilities 0.002690531
recurrent translation 0.002681915
source phrase 0.00265165
language model 0.002634414
novel translation 0.002620864
rnn translation 0.002618312
bidirectional translation 0.002616437
target words 0.00261489
unidirectional translation 0.002603105
posterior translation 0.00259701
translation units 0.0025962109999999997
joint translation 0.0025944089999999998
translation tasks 0.002589827
phrasal translation 0.002582283
translation experiments 0.002581769
translation modeling 0.002577486
translation mod 0.002560973
different model 0.002555734
tional translation 0.002548949
source words 0.00254519
all translation 0.002540731
minimum translation 0.0025338
translation direc 0.002523001
reverse translation 0.0025211079999999998
translation paradigm 0.002519982
pure translation 0.002519434
model baseline 0.00250483
phrase words 0.00245526
ing model 0.002388228
translation 0.00228336
rnn model 0.002228382
bidirectional model 0.002226507
unidirectional model 0.002213175
joint model 0.002204479
phrase alignment 0.002180351
orientation model 0.0021689409999999998
target sentence 0.00215736
model combination 0.002149612
guage model 0.002148055
model archi 0.002140641
training data 0.00212972
source sentence 0.00208766
previous target 0.001998713
target phrases 0.0019088199999999999
model 0.00189343
target vocabulary 0.001871483
target order 0.001866403
language models 0.0018627739999999998
last target 0.001841402
neural models 0.001841119
previous phrase 0.001839083
target input 0.0018385839999999999
current target 0.00182406
target distribution 0.0018120739999999999
target history 0.00181105
english phrase 0.001809207
output phrase 0.001793268
last source 0.001771702
target sentences 0.001769207
source input 0.001768884
target side 0.001758369
target sequence 0.001746438
source information 0.001742268
source history 0.00174135
recent target 0.001732697
full source 0.001707054
phrase length 0.001704133
target identity 0.001702302
source window 0.0017015210000000001
network models 0.001698742
source projection 0.001695324
network training 0.001694432
