translation model 0.003927170000000001
phrase translation 0.00360369
source word 0.00343658
translation probability 0.003239456
machine translation 0.00301766
target sentence 0.002976908
target language 0.0029288559999999996
translation performance 0.002888353
word alignment 0.0028625839999999996
translation hypotheses 0.002795348
test target 0.002785499
training data 0.00276504
translation experiments 0.0027636880000000003
interactive translation 0.002751734
word alignments 0.002749527
translation sys 0.0027360870000000003
bilingual word 0.002728099
translation rules 0.002718367
translation initiative 0.002706477
many word 0.002649693
target sentences 0.0026265159999999997
automatic word 0.0024946169999999998
translation 0.00243097
source words 0.0024024800000000002
corresponding target 0.002399527
language model 0.0023971360000000002
target hypotheses 0.0023922979999999997
target perplexity 0.002364195
conditional target 0.0023616089999999998
target terms 0.002352317
individual target 0.002342776
arbitrary target 0.002335796
target sen 0.002326575
target documents 0.002309173
vant target 0.002306609
jth target 0.002303157
parallel data 0.00226058
source sentence 0.002187628
source language 0.002139576
training corpus 0.00207794
target 0.00202792
source vector 0.002022897
data configuration 0.0019273250000000001
source sentences 0.001837236
background model 0.001813282
model biasing 0.001775484
parallel training 0.00174566
source input 0.001743959
source space 0.0017297710000000002
training sentences 0.001723656
test corpus 0.0017104590000000001
maximum source 0.0017041320000000001
other words 0.001669859
source vocabulary 0.001659488
different vector 0.001622764
smt system 0.001619489
date source 0.0015783190000000001
parallel corpus 0.00157348
candidate source 0.0015711310000000001
source terms 0.001563037
source sen 0.0015372950000000002
times source 0.0015361910000000001
source span 0.001518313
source docu 0.001515197
test test 0.001515158
language sentences 0.001499532
model 0.0014962
baseline smt 0.001473766
rate training 0.001459092
language input 0.001406255
smt performance 0.001401415
language space 0.0013920669999999999
sentence level 0.001382422
parallel test 0.001378179
monolingual corpus 0.001375203
sentence length 0.001356859
test set 0.001338607
language pairs 0.001323782
smt decoder 0.001322627
weighted corpus 0.001309812
smt hypotheses 0.00130841
unit sentence 0.001285032
smt systems 0.001283965
smt experiments 0.00127675
vector space 0.0012753880000000001
interactive smt 0.001264796
test input 0.001262898
initial smt 0.00125292
source 0.00123864
smt applications 0.001238509
smt perfor 0.001234337
large size 0.001230725
smt architecture 0.001228847
high probability 0.001206589
vocabulary vector 0.001205105
language vocabularies 0.00118884
maximum bleu 0.001188364
pairs test 0.001180425
phrase 0.00117272
words 0.00116384
