word reordering 0.00538635
reordering model 0.00515415
translation model 0.0045054
word system 0.0040763299999999995
translation system 0.0038252499999999997
distance reordering 0.0035090909999999998
learning reordering 0.00346466
reordering method 0.0034451300000000002
reordering rules 0.00342691
based reordering 0.003390731
different word 0.003343454
syntactic reordering 0.003323691
model training 0.00328274
reordering lattice 0.003273601
reordering alternatives 0.003265927
reordering patterns 0.003232306
reordering window 0.003231854
tance reordering 0.003221347
reordering pattern 0.003215612
reordering lattices 0.003213717
language model 0.0032077670000000003
word order 0.003079066
word alignment 0.0030429149999999998
based word 0.002993061
chinese word 0.0029339089999999997
machine translation 0.002918248
automatic word 0.002897664
reordering 0.00289201
translation score 0.002883798
word lattice 0.002875931
nese word 0.002870723
model scores 0.002827645
word segmen 0.0028160389999999998
translation quality 0.002749523
test data 0.0027297
translation process 0.002656678
ibm model 0.002652512
jump model 0.0026458930000000003
lattice translation 0.002624851
english system 0.00261374
lation model 0.0026104270000000002
overall translation 0.002610405
model history 0.0026037400000000002
intermediate model 0.002599198
model due 0.002594559
tortion model 0.0025942860000000003
fertility model 0.002585195
parallelizing model 0.002583245
model ect 0.002583245
training data 0.00258157
translation functionality 0.002564816
large data 0.002324413
model 0.00226214
smt system 0.0022582469999999997
bilingual data 0.002254339
translation 0.00224326
system performance 0.0022064569999999998
parallel data 0.002085453
source sentence 0.0020337330000000002
chinese system 0.002021559
full data 0.001991794
training corpus 0.001953451
overall system 0.0019491349999999998
consistent data 0.001940836
data processing 0.001940155
enough data 0.001888275
newswire data 0.0018824850000000001
test set 0.001845736
language models 0.00182637
chinese words 0.0017804090000000002
phrase table 0.001738216
bilingual training 0.001713969
unknown words 0.001672001
bilingual sentence 0.001634882
bilingual corpus 0.0016262199999999998
bleu score 0.001625905
same training 0.001590827
test sen 0.001589945
system 0.00158199
test sets 0.001575611
unseen test 0.001557344
several language 0.0015516269999999999
phrase pairs 0.001550388
statistical models 0.001531802
sentence alignment 0.001490088
phrase tables 0.001477117
english side 0.001470859
alignment models 0.0014293180000000002
gual training 0.00142462
english part 0.001416215
training time 0.001396108
training steps 0.001395678
english gigaword 0.0013775889999999998
gual sentence 0.001345533
sentence pairs 0.001344161
words 0.00134084
vocabulary language 0.001335146
language modeling 0.001333811
statistical machine 0.001326047
main improvements 0.001296266
