translation system 0.00579462
imt translation 0.00524994
bleu translation 0.00476969
first translation 0.004644314
machine translation 0.004641768
translation quality 0.004554441
translation error 0.004518126000000001
translation scores 0.004396154
second translation 0.004355993000000001
reference translation 0.004293746
automatic translation 0.004275358
final translation 0.0042625810000000005
translation predictions 0.004259775
initial translation 0.0042294970000000005
incorrect translation 0.004214804
perfect translation 0.004201486000000001
monotonic translation 0.004201109000000001
tial translation 0.004200063
third translation 0.0041940020000000005
translation 0.00391846
imt system 0.00320764
source word 0.00311875
word confidence 0.002818598
other word 0.00269811
word classification 0.0026831629999999997
word graph 0.002440394
smt system 0.00243744
word cms 0.002418751
word stroke 0.002406677
complete word 0.002382159
word graphs 0.0023682679999999998
system evaluation 0.002351654
different sentence 0.0021844300000000002
imt bleu 0.00218271
complete system 0.002174749
source sentence 0.00214399
imt approach 0.002138559
different language 0.001886142
system 0.00187616
conventional imt 0.0017234659999999999
imt sys 0.001716469
output translations 0.001711124
sentence classification 0.001708403
imt framework 0.001700255
wsr imt 0.001695776
input sentence 0.0016794000000000002
real imt 0.001669227
imt scenario 0.001637773
final translations 0.001636301
source sentences 0.001633434
imt procedure 0.0016233559999999998
tional imt 0.0016156999999999999
ventional imt 0.00160853
imt scenar 0.001607288
imt famework 0.001607288
initial translations 0.001603217
perfect translations 0.001575206
ence translations 0.001571199
bleu score 0.0014764560000000001
sentence cms 0.001443991
source sen 0.0014142439999999998
bleu smt 0.00141251
user effort 0.0014100179999999999
source string 0.0014071379999999998
sentence clasiffication 0.001387698
different values 0.001378857
different val 0.001361594
different statisti 0.001353321
ibm model 0.001349306
bleu scores 0.001328924
empty source 0.00132639
first error 0.00132552
statistical machine 0.001317064
translations 0.00129218
building phrase 0.001286375
training set 0.001252378
confidence scores 0.001212722
language pairs 0.001211302
smt models 0.001208099
test set 0.001180349
user types 0.001154642
bleu points 0.001129775
confidence measures 0.001117942
confidence estimation 0.001115737
quality scores 0.001113675
confidence value 0.0011093470000000001
sentence 0.00110881
following way 0.001107107
human effort 0.001097439
different 0.00107562
smt figure 0.001049071
first step 0.001044819
confidence values 0.001038265
source 0.00103518
classification threshold 0.0010129610000000002
grate confidence 0.001011616
phrase 0.00101042
human trans 0.001009844
test sets 0.001005041
predictive machine 9.99064E-4
