translation model 0.00484429
target translation 0.00398117
phrase translation 0.003856635
translation sentence 0.0038226700000000002
translation models 0.003697857
translation system 0.003491787
different translation 0.0034739230000000003
translation rules 0.0034704090000000003
english translation 0.003417867
other translation 0.0033972480000000003
lexical translation 0.00339497
translation method 0.0033854420000000002
many translation 0.003368269
translation probabilities 0.0033403910000000003
machine translation 0.003294808
hierarchical translation 0.0032416980000000003
translation quality 0.003216384
new translation 0.0032081090000000002
string translation 0.003206642
translation experiments 0.003206623
discourse translation 0.003177659
entire translation 0.003176412
translation integrity 0.003163757
final translation 0.003146915
translation candidates 0.003133028
translation examples 0.0031318640000000003
golden translation 0.003130338
actual translation 0.0031278
translation qualities 0.0031278
word alignment 0.0030774
model feature 0.0030260900000000004
target word 0.00300084
language model 0.002938426
translation 0.00287972
reordering model 0.002737531
word corpus 0.002667211
source sentence 0.00264869
model score 0.0026097480000000003
such word 0.002521626
tion model 0.002379178
transfer model 0.002330847
form word 0.0022564399999999997
source sentences 0.002242118
cohesive word 0.002234507
new word 0.002227779
source figure 0.002222549
test source 0.0022100040000000002
connection word 0.002188303
second source 0.002180749
connective word 0.002171539
word align 0.002165196
thi word 0.002163267
word selec 0.00214795
indefinite word 0.00214795
source css 0.0021107450000000002
source string 0.002032662
source side 0.002029106
target words 0.0020280149999999998
original source 0.002009419
source sub 0.001969334
model 0.00196457
source compound 0.001954032
neighboring source 0.001954032
disjoint source 0.001954032
training data 0.00176837
different features 0.001763373
source 0.00170574
alignment results 0.001703847
training corpus 0.001662615
additional features 0.00166226
baseline features 0.001574128
additional feature 0.00155461
euc alignment 0.001552111
evaluation data 0.0015189000000000001
sentence structure 0.00151569
new alignment 0.0015063989999999998
original alignment 0.001481689
language pairs 0.0014731570000000001
english words 0.001464712
bilingual training 0.0014456109999999999
sentence pairs 0.0014422509999999999
incorrect alignment 0.0014349789999999999
target string 0.0014283719999999998
alignment complies 0.001426616
target side 0.001424816
several words 0.001409023
target span 0.0014035249999999999
transitional phrase 0.001403476
chinese sentence 0.001400999
chinese words 0.001384614
test data 0.00137784
projected target 0.0013603419999999998
jected target 0.001357419
local sentence 0.001353526
transitional words 0.001353126
target chains 0.001351657
thn target 0.001350385
tional feature 0.0013338550000000001
ditional feature 0.001317009
training process 0.001314841
