translation rules 0.00422304
translation rule 0.00383832
phrase translation 0.00382905
translation models 0.00375495
reordering model 0.0034157500000000004
translation systems 0.0033980160000000002
translation performance 0.003347976
new translation 0.003202313
machine translation 0.003173118
translation toolkit 0.003143109
translation probabilities 0.003073553
translation quality 0.003048113
translation process 0.003046654
translation speed 0.0030421
efficient translation 0.003022332
scale translation 0.0030199980000000003
translation engines 0.002988211
language model 0.002901854
translation 0.00271418
model score 0.00246012
reordering models 0.00245756
phrase decoding 0.00244173
second model 0.002421243
different reordering 0.002209137
different decoding 0.002119207
reordering probability 0.002031608
model 0.00199896
grammar rule 0.001938118
binary rules 0.0019089830000000002
scfg rules 0.0019068380000000001
reordering problem 0.001886235
training data 0.001834278
different models 0.001833117
minimal rules 0.001820351
glue rules 0.0017948590000000002
nearby rules 0.001781645
larger rules 0.001781645
discriminative reordering 0.00176726
msd reordering 0.001733386
word alignment 0.001719291
hierarchical phrase 0.001715233
other features 0.001708874
reordering types 0.001693853
decoding methods 0.001689923
target language 0.001675731
decoding process 0.001659334
data structure 0.001621125
source sentence 0.001619639
decoding stages 0.001605171
decoding modes 0.00159999
smt models 0.001553085
training corpus 0.001527778
phrase table 0.001521347
rule extraction 0.001517808
new features 0.001511833
rules 0.00150886
phrase extraction 0.001508538
language corpus 0.001502405
phrase penalty 0.001500877
syntactic labels 0.001469699
additional features 0.001468299
feature weights 0.001466884
source machine 0.001418239
reordering 0.00141679
main features 0.00140613
language side 0.001405879
rule application 0.001399489
feature engineering 0.001398233
source toolkit 0.0013882299999999998
source span 0.001375632
syntactic information 0.001374908
feature values 0.001368763
complex features 0.001358195
decoding 0.00132686
word alignments 0.001323772
parsing algorithm 0.001305152
weight training 0.001290772
bleu scores 0.001270616
bilingual training 0.001264852
first system 0.001260579
new system 0.001259987
entire source 0.0012471869999999999
rate training 0.001243482
open source 0.001243002
bilingual data 0.001242596
basic data 0.001240062
bleu score 0.001238338
smt systems 0.001196151
fast language 0.0011880950000000001
gram language 0.00118333
bleu improvement 0.001131171
translations table 0.001129025
test set 0.001124614
parse tree 0.0011193330000000001
phrase 0.00111487
mert system 0.001111994
standard framework 0.0011114620000000001
reference translations 0.001106644
feature 0.00108161
various systems 0.001078356
