translation model 0.00423148
feature model 0.0035006000000000004
language model 0.002982586
model score 0.002927428
smt model 0.0028630390000000004
distortion model 0.002650901
level model 0.002641853
cost model 0.002599862
model class 0.002572304
model implementation 0.002542934
guage model 0.0025229170000000004
parity model 0.002518011
model lookups 0.002508705
feature models 0.00241906
many translation 0.002400446
machine translation 0.002370902
reference translation 0.002298365
model 0.00226798
example translation 0.002257815
translation quality 0.002254343
document translation 0.002249083
translation process 0.0022464299999999998
actual translation 0.002245123
swedish translation 0.002215458
translation options 0.00220569
translation snippets 0.002204441
target word 0.0021147180000000002
source phrase 0.0021135299999999997
translation 0.0019635
source sentence 0.0019244969999999998
same models 0.0019114029999999999
search state 0.001911177
language models 0.001901046
new feature 0.001872168
feature bleu 0.001857593
smt feature 0.001827679
smt decoding 0.001799819
source language 0.001792346
baseline features 0.0017825900000000001
smt models 0.001781499
phrase translations 0.001761553
feature scores 0.001730322
standard features 0.00167675
standard models 0.00166278
feature weights 0.001658425
decoding algorithm 0.001653487
target words 0.0016487490000000001
word penalty 0.001621983
new state 0.00161957
word order 0.0016110820000000002
target language 0.0016057340000000002
word level 0.0015974630000000001
important word 0.001592309
simple features 0.001579408
multiple decoding 0.001579115
new search 0.0015707030000000001
limit feature 0.001567658
feature weight 0.001564467
common models 0.001558724
available feature 0.001548148
word sequences 0.001541503
corresponding features 0.001510601
readability features 0.0015040350000000001
consistency feature 0.001498781
indicator feature 0.001493639
different translations 0.001491808
readability models 0.001490065
decoding process 0.00148769
document features 0.0014859930000000001
contiguous word 0.0014841330000000001
search decoder 0.001478394
expensive feature 0.001474055
comments feature 0.001474055
word vector 0.0014715370000000002
consistency features 0.001466571
ovix features 0.001463466
subsequent decoding 0.0014578310000000001
decoding pass 0.001453325
traditional models 0.001450974
search results 0.001448802
dividual features 0.001448623
cient decoding 0.001447304
ency features 0.001442161
cheap models 0.001427286
phrase level 0.001409663
phrase table 0.001383557
baseline system 0.0013803119999999999
search algorithm 0.001379882
previous state 0.0013423200000000001
same training 0.0013268770000000002
local search 0.001324932
phrase dictionary 0.0013248399999999999
test data 0.001321759
source code 0.001320395
ambiguous source 0.001319038
current search 0.001315283
beam search 0.001309594
search space 0.001306178
new score 0.001298996
bleu score 0.001284421
