word alignment 0.004494
translation model 0.00449138
source word 0.00424055
target word 0.00405976
model features 0.00389268
model feature 0.0038583999999999997
word alignments 0.00380552
different word 0.003775254
language word 0.0036964140000000003
word reordering 0.003679373
word position 0.003599096
word sequence 0.003585249
distance word 0.0035267140000000002
word pairs 0.003475788
distortion model 0.003470057
word penalty 0.003464426
relative word 0.003402248
word order 0.003393531
language model 0.0033908439999999996
reordering model 0.003373803
japanese word 0.003364071
original word 0.00334747
word positions 0.0032995520000000003
word orders 0.003291571
appropriate word 0.003290218
word katta 0.003286357
nearby word 0.003286357
word spec 0.003286357
sequence model 0.0032796789999999998
model parameters 0.003277623
model task 0.003236981
pair model 0.003190572
tion model 0.003130652
model probabilities 0.003081219
cost model 0.003070093
probabilistic model 0.003068379
tic model 0.003049795
model cal 0.003041799
outbound model 0.0030224609999999997
model definition 0.003021149
inbound model 0.003020821
lation model 0.003012277
model the 0.003011168
model dis 0.0029986509999999997
tortion model 0.002992042
criminative model 0.002989967
quence model 0.002982795
model 0.0027367
english translation 0.002371652
source sentence 0.00233503
source words 0.00233259
translation task 0.0022549609999999998
machine translation 0.0021800969999999998
target words 0.0021517999999999997
translation rules 0.002141066
translation experiments 0.002116696
translation order 0.002105941
translation quality 0.002070725
translation subtasks 0.001999002
source phrase 0.0019654679999999997
function words 0.0018629799999999998
source language 0.0018524239999999999
feature function 0.00185037
target phrase 0.001784678
translation 0.00175468
sentence position 0.001693576
smt features 0.0016908420000000001
target language 0.001671634
source sentences 0.001667818
sentence figure 0.0016634649999999998
smt feature 0.0016565619999999999
class feature 0.001651984
additional feature 0.0016463580000000001
source side 0.00164057
feature set 0.001597628
sentence pairs 0.0015702679999999999
relative source 0.0015582579999999999
corresponding source 0.001554439
sentence level 0.0015468819999999999
training data 0.001539302
yesterday source 0.001512175
distortion models 0.0015035489999999999
same training 0.0014913499999999998
corresponding words 0.001490469
eos source 0.001474771
source sen 0.0014670809999999999
different distortion 0.001466341
target side 0.00145978
distortion probability 0.0014591410000000002
alignment 0.00145173
language models 0.001424336
binary feature 0.001413679
bilingual training 0.001412649
reordering models 0.0014072949999999998
feature functions 0.00139961
upper sentence 0.001389172
feature templates 0.001384215
same distortion 0.001380553
feature template 0.001372022
feature tem 0.0013668320000000001
