word translation 0.00539882
phrase translation 0.0053532200000000005
training phrase 0.00437138
translation model 0.00425017
translation features 0.0037746900000000002
translation system 0.003601684
system translation 0.003601684
target word 0.003597732
translation feature 0.003577601
target phrase 0.003552132
word alignment 0.003497524
source word 0.0034563500000000004
translation models 0.003428908
source phrase 0.0034107499999999997
language translation 0.0033620060000000003
phrase models 0.003331608
baseline translation 0.003300476
translation results 0.003297013
phrase pair 0.003289465
translation table 0.003259703
lexicon translation 0.003259254
machine translation 0.003254536
translation task 0.003218882
translation probabilities 0.003203722
training data 0.0031703900000000004
phrase table 0.003162403
reference translation 0.003158628
significant translation 0.00314416
translation hypotheses 0.003131753
translation performance 0.003128914
train phrase 0.003125513
possible word 0.0031251440000000003
phrase length 0.003120596
word alignments 0.003120222
translation metric 0.0031089810000000002
translation quality 0.003100194
translation process 0.003096712
ted translation 0.0030852960000000004
phrase pairs 0.003080681
phrase alignments 0.003074622
maximum phrase 0.003065717
translation tasks 0.003057689
the translation 0.00305367
translation measure 0.00305309
common phrase 0.003028569
optimal translation 0.003021494
translation benchmark 0.003019356
phrase extraction 0.003018178
phrase counts 0.003013165
word count 0.003004562
phrase level 0.003002552
translation result 0.0029998110000000002
desirable translation 0.00299648
parameters phrase 0.00295917
phrase count 0.002958962
phrase segmentation 0.002951786
bleu training 0.0029497100000000004
forward phrase 0.002924954
secondary phrase 0.002908552
primary phrase 0.002900622
translation 0.00272526
training algorithm 0.0026312410000000003
phrase 0.00262796
training set 0.002570016
bleu bleu 0.00241258
training corpus 0.002375256
training method 0.002372402
reordering model 0.002336507
training lexicon 0.002277414
sentence bleu 0.00226628
parallel training 0.002207409
full training 0.00219288
ibm model 0.00218828
training iteration 0.0021766290000000002
discriminative training 0.002123921
training process 0.0021148720000000003
training procedure 0.002114563
training objective 0.002112942
model weights 0.002102138
training iterations 0.00207545
stage training 0.002060621
lexicon model 0.002058904
rate training 0.002058753
training corpora 0.0020546880000000003
training dataset 0.002041413
entire training 0.002023734
training schedule 0.002015739
target sentence 0.0019841620000000003
english data 0.0019796980000000002
linear model 0.001962346
smt model 0.001941339
bleu score 0.001932506
tuning data 0.001910067
prior model 0.001891905
parallel data 0.001890959
target phrases 0.0018690970000000001
markov model 0.001864618
model parameters 0.00185612
source sentence 0.00184278
corpus bleu 0.001838126
