target word 0.00449065
translation model 0.00424451
source word 0.0042376
language word 0.00405262
word alignments 0.003847904
different word 0.0038190489999999997
word lexicon 0.003478313
word error 0.003442988
standard word 0.003428716
discriminative word 0.003377875
several word 0.003364291
word selection 0.003340018
word sense 0.003324269
word lists 0.00332422
content word 0.0033052809999999998
word choice 0.003303706
word triggers 0.003298854
word classes 0.0032987579999999997
tive word 0.0032840919999999997
tent word 0.0032813779999999997
word triplets 0.003276903
word lexicons 0.0032719389999999998
empty word 0.003271581
native word 0.003268837
criminative word 0.00326713
incorrect word 0.003266635
word lexi 0.003265867
word lex 0.003265418
word remarks 0.003262484
translation models 0.003161288
language model 0.00310739
phrase translation 0.003103584
phrase model 0.003025474
translation results 0.00292105
large translation 0.002708362
translation approach 0.002705106
source target 0.00267139
machine translation 0.002666845
additional model 0.0026483789999999998
translation system 0.002642303
target words 0.0025976899999999997
model score 0.002595399
ibm model 0.002559148
first model 0.002549338
lexicon model 0.0025330829999999998
text translation 0.002531098
reference translation 0.002516326
translation process 0.0025134190000000002
translation tasks 0.002499509
second model 0.0024921879999999998
translation probabilities 0.0024891080000000003
target language 0.0024864099999999997
phrasal translation 0.002478154
full translation 0.002474261
translation quality 0.002461874
target sentence 0.00246093
model parameters 0.0024572509999999997
translation example 0.002444838
translation hypothesis 0.002443994
model discriminative 0.0024326449999999998
discriminative model 0.0024326449999999998
linear model 0.002427075
final model 0.002413316
model probabilities 0.002410998
translation lexicons 0.002404819
model com 0.0024001079999999998
translation candi 0.002397378
chine translation 0.002396617
eral translation 0.00239585
selection model 0.002394788
model assumption 0.0023686059999999997
triplet model 0.002352939
separate model 0.002351074
model the 0.002344727
source words 0.0023446400000000003
only model 0.002331639
model variants 0.002327254
model discrimina 0.00232703
dwl model 0.002322431
trained model 0.0023198209999999997
source language 0.00223336
source sentence 0.00220788
translation 0.00216131
same target 0.002157608
source phrase 0.002151444
sentence words 0.0021341800000000003
model 0.0020832
language models 0.002024168
phrase models 0.0019422520000000002
target sentences 0.001825862
training data 0.0018040679999999998
target vocabulary 0.001798624
different models 0.0017905970000000001
target side 0.001765178
phrase alignments 0.001761748
entire target 0.001728452
ing words 0.001714677
target vocabu 0.001696806
repeated target 0.001696036
sentence context 0.0016722210000000002
