translation systems 0.004055021
translation performance 0.003912898
translation rule 0.0038431949999999998
target translation 0.003818865
machine translation 0.003816472
multiple translation 0.003764308
translation hypotheses 0.003744789
translation results 0.0036106759999999997
translation quality 0.003590186
translation confidence 0.00358931
system decoding 0.00356348
final translation 0.00352447
translation derivations 0.0035231109999999998
translation forests 0.003493169
translation spaces 0.003491862
translation tasks 0.003485168
reference translation 0.003476845
entire translation 0.00345977
translation pair 0.0034581409999999996
translation reconstruction 0.003448155
decoding model 0.00344736
translation 0.00315959
model feature 0.00315382
system combination 0.00307136
model bleu 0.00305242
model combination 0.00295524
smt system 0.002947458
smt model 0.002831338
reordering model 0.0027950830000000003
language model 0.002746095
model training 0.00265939
model score 0.002575075
component system 0.0025721060000000002
single system 0.002554236
different models 0.0024798700000000003
further system 0.002470775
component model 0.0024559860000000003
smt decoding 0.002455878
mixture model 0.002452151
single model 0.002438116
tree system 0.002437681
decoding method 0.00242165
model scores 0.002414012
additional system 0.002407221
own model 0.002359725
linear model 0.0023530720000000003
different smt 0.002306588
different systems 0.002282091
syntactic model 0.002281776
maximum model 0.002279146
new decoding 0.0022194989999999998
different rules 0.002203509
decoding consensus 0.002170726
consensus decoding 0.002170726
decoding techniques 0.002122247
mixture decoding 0.002076691
mbr decoding 0.002075241
system 0.00202753
smt models 0.0020131380000000003
reordering features 0.001981573
reordering models 0.001976883
decoding approaches 0.001970833
source sentence 0.001953435
feature set 0.001951136
decoding procedure 0.0019468819999999998
network decoding 0.0019325359999999999
combination method 0.00192953
level decoding 0.00191673
model 0.00191141
collaborative decoding 0.001885851
related decoding 0.001880498
joint decoding 0.00187602
different methods 0.001867532
decoding scheme 0.001864871
decoding algorithms 0.0018580889999999998
word alignments 0.0018492080000000002
hybrid decoding 0.001828295
decoding hybrid 0.001828295
different nist 0.001825353
decoding stages 0.0018248659999999999
based decoding 0.001824802
borative decoding 0.001823136
smt systems 0.001815359
bleu score 0.001804675
word stem 0.0017942260000000001
different sets 0.001776742
word count 0.0017701750000000001
consensus features 0.0017326759999999998
consensus models 0.001727986
bleu improvements 0.001724069
posterior feature 0.001718511
reordering rules 0.0017005219999999999
different paradigms 0.001695399
different levels 0.0016837190000000002
different ranges 0.0016822550000000001
test data 0.001677195
feature vector 0.001671904
new search 0.001662341
feature weight 0.001660266
training data 0.001655337
