translation system 0.00483291
translation systems 0.0035743
system bleu 0.0035685660000000004
system training 0.0034917670000000002
system performance 0.003368134
system combination 0.003358251
smt system 0.0032953410000000002
system baseline 0.003236989
baseline system 0.003236989
system selection 0.003181727
system results 0.0031075350000000002
oracle system 0.003035639
single system 0.003030811
english translation 0.002987915
strong system 0.002928399
identification system 0.002924874
individual system 0.002899665
tion system 0.002892903
machine translation 0.0028924190000000002
system parameters 0.002881606
system label 0.0028685200000000003
system all 0.0028666670000000003
system selec 0.002865795
acle system 0.0028437080000000003
translation experiments 0.002630236
system 0.00258068
different data 0.00254231
literal translation 0.002519708
test data 0.002327433
training data 0.002284977
translation 0.00225223
english data 0.002109575
data set 0.002084297
different training 0.002079507
english systems 0.002057755
language model 0.002038702
smt systems 0.002036731
word corpus 0.0019829640000000003
baseline systems 0.001978379
selection systems 0.001923117
same data 0.001892341
data class 0.001812111
different machine 0.001808609
average word 0.001780728
word length 0.001759565
language features 0.001749523
parallel data 0.001743154
weka data 0.001688877
test set 0.00166395
different weights 0.001656413
data mining 0.001649089
bolt data 0.0016488660000000001
data scarcity 0.001637545
ferent systems 0.0016345560000000001
different classifiers 0.0016277960000000001
test sentence 0.0016191439999999999
feature weights 0.001610343
bleu score 0.001600911
msa training 0.0015706700000000001
different dialects 0.001518795
different languages 0.001508821
different techniques 0.001489007
different types 0.0014787910000000001
different ways 0.0014667550000000001
msa performance 0.001447037
classifier training 0.0014358840000000001
feature sources 0.0014253550000000001
source language 0.001420908
language models 0.0014166690000000002
feature vectors 0.00139357
level features 0.001389981
test sets 0.001369973
bleu improvement 0.001355026
combination work 0.001343816
msa sentences 0.001337931
bleu scores 0.0013337000000000002
classification training 0.001327711
msa sentence 0.001325184
systems 0.00132207
overall bleu 0.0013120190000000002
bleu point 0.0013057420000000001
language fea 0.00130106
target language 0.001295275
absolute bleu 0.001265641
bleu name 0.0012647420000000001
selection approach 0.001262025
msa selection 0.00126063
bleu points 0.0012580960000000002
selection baseline 0.0012573559999999998
bleu diff 0.001254756
bleu bias 0.0012511830000000001
rate training 0.001249175
diglossic language 0.001240611
set results 0.001237262
oracle combination 0.00123253
same set 0.001228858
blind test 0.0012284519999999999
basic features 0.001228426
sentence level 0.001214531
main approach 0.001190613
