word features 0.00400653
individual features 0.0026057520000000002
following features 0.002569048
sparse features 0.002536912
novel features 0.002484479
dense features 0.002450412
adequacy features 0.00244919
fluency features 0.002447776
same feature 0.002313635
feature score 0.002312752
word order 0.00229644
features 0.002213
linear model 0.00216387
word ngrams 0.002128731
feature values 0.002043555
feature vectors 0.001983842
feature func 0.001981253
model our 0.001978629
our model 0.001978629
trainable model 0.001957516
training data 0.001759954
feature 0.00174096
model 0.0017201
function words 0.001712066
evaluation metric 0.00161588
target language 0.001528766
machine translation 0.001494703
content words 0.001481874
language pair 0.001477201
translation rule 0.0014571240000000002
metric score 0.001425721
language pairs 0.0014188360000000001
matched words 0.001391325
words accord 0.001378669
words pcont 0.0013773
words pall 0.0013773
training approach 0.0013426850000000001
new metric 0.001289294
source sentence 0.001259911
training objective 0.001238253
level evaluation 0.001228928
same set 0.001214733
sentence level 0.001210883
level metrics 0.001160753
human ranking 0.001146101
translation 0.00114239
words 0.00114154
same alignment 0.001134553
system level 0.001119925
human judgment 0.0011143820000000001
reordering building 0.0011043799999999999
atomic reordering 0.001092079
metric beer 0.001089534
same source 0.00108868
system performance 0.001084437
human judgments 0.00108235
human judgement 0.001075439
system output 0.001074926
corpus level 0.001068641
modelling sentence 0.001065584
language 0.00101408
system translations 9.97237E-4
evaluation demands 9.96613E-4
different ways 9.706860000000001E-4
system tuning 9.655009999999999E-4
different pets 9.44886E-4
used metrics 9.29654E-4
ing approach 9.29154E-4
system outputs 9.27873E-4
possible derivations 9.17366E-4
same way 9.04527E-4
smt system 8.9202E-4
ited set 8.76681E-4
intricate models 8.685609999999999E-4
individual sentences 8.67802E-4
metric 8.53929E-4
training 8.52002E-4
trainable models 8.342829999999999E-4
level nodes 8.33143E-4
kendall score 8.33112E-4
reordering 8.29012E-4
complex order 8.270739999999999E-4
ing hypothesis 8.24813E-4
matched function 8.20311E-4
order difference 8.16485E-4
same tables 8.13247E-4
modelling approach 8.123609999999999E-4
fill score 8.0698E-4
linear classifier 8.031819999999999E-4
derivation tree 8.006210000000001E-4
correct order 7.86957E-4
tree organization 7.80933E-4
segment level 7.71009E-4
linear modelling 7.65448E-4
novel approach 7.62162E-4
evaluation 7.61951E-4
pet figure 7.598430000000001E-4
binary monotone 7.5665E-4
weight vector 7.56244E-4
sentence 7.43906E-4
