new word 0.004656425
first word 0.004463355
such word 0.004343107
word graph 0.004221113
correct word 0.004189964
complete word 0.004178221
word german 0.004175353
word spanish 0.00415738
special word 0.0041534829999999995
word stroke 0.004152461
word graphs 0.0041240849999999996
translation model 0.00411188
incorrect word 0.004109318
word french 0.004088809
word basis 0.004087177
different translation 0.003488843
translation models 0.0034523110000000004
new translation 0.0033899050000000003
translation hypothesis 0.00317855
machine translation 0.0030729
interactive translation 0.0030487070000000003
translation process 0.002991042
final translation 0.0029414080000000004
partial translation 0.0029396080000000002
reference translation 0.002920769
complete translation 0.002911701
translation direction 0.0028988480000000003
equivalent translation 0.002837622
candidate translation 0.0028298330000000003
translation environment 0.0028237
monotonic translation 0.0028225340000000002
imt system 0.0027142
translation 0.00252926
language model 0.002523735
source phrase 0.00221033
source sentence 0.00220432
smt model 0.002184566
target phrase 0.002065785
target sentence 0.0020597750000000002
smt system 0.002037816
system evaluation 0.001975884
baseline system 0.001862308
target words 0.001860437
complete system 0.001818311
ing system 0.001810286
source text 0.0017476079999999999
imt framework 0.001731846
new suffix 0.001711548
imt procedure 0.001696217
real imt 0.0016400960000000002
typical imt 0.00161355
input sentence 0.001610008
imt sys 0.001608915
information source 0.001606412
target text 0.001603063
imt scenario 0.001595639
imt paradigm 0.001590022
imt session 0.0015851700000000001
model 0.00158262
traditional imt 0.001577915
imt scenarios 0.001576502
tional imt 0.001575133
imt proce 0.001571051
whole sentence 0.0015496680000000001
complete sentence 0.0015474310000000002
same models 0.00153889
spanish sentence 0.00152659
smt models 0.0015249970000000002
new hypothesis 0.001509935
new prefix 0.0015075520000000001
suffix hypothesis 0.001500193
search algorithm 0.0014869409999999999
statistical approach 0.0014586339999999999
previous suffix 0.0014543730000000001
system 0.00143587
search process 0.001382978
single words 0.001360429
different trans 0.0013555
text prefix 0.001355185
linear models 0.001353024
source sen 0.00134943
search procedure 0.001339083
test set 0.0013313210000000001
several suffix 0.001309391
different suffixes 0.0013070719999999998
argmax search 0.001306509
new input 0.001305663
language pairs 0.001295923
different approaches 0.001280982
possible translations 0.001279718
target phrases 0.001271174
full target 0.001270885
training set 0.00126831
other systems 0.0012567680000000001
english wsr 0.001256026
different authors 0.001254285
correct suffix 0.001245087
statistical machine 0.001232721
beam search 0.001221059
new action 0.0012200420000000002
