evaluation metric 0.00365124
human evaluation 0.00348952
human translation 0.0032935300000000002
evaluation metrics 0.00302979
metric bleu 0.0027423079999999997
translation sentence 0.00262204
evaluation model 0.0025779229999999998
human reference 0.002544963
level evaluation 0.002434942
translation quality 0.002380966
metric coefficient 0.002366787
based metric 0.002359617
automatic evaluation 0.002353432
our metric 0.0022856179999999997
svm evaluation 0.0022835909999999997
reference translation 0.0022757330000000003
english translation 0.002252205
human references 0.0022349049999999997
machine translation 0.002233868
evaluation dataset 0.00223292
evaluation task 0.00220287
evaluation results 0.002201413
human judgments 0.002188762
tion evaluation 0.002182959
human assessment 0.002174441
human judges 0.00213832
human judgment 0.00211728
human judg 0.002108412
human perception 0.002106419
good translation 0.002071321
evaluation process 0.0020519699999999997
chinese evaluation 0.002051113
generalized evaluation 0.002049128
evaluation progress 0.002045886
nese evaluation 0.002033218
same metrics 0.002005453
metric 0.0019431
reference sentence 0.001873473
sentence feature 0.001866843
several metrics 0.001864911
language model 0.0018613050000000002
test sentence 0.0018254680000000002
model score 0.001799607
tion metrics 0.001796469
human 0.00178138
reference sentences 0.001772263
general metrics 0.0017659199999999998
parser score 0.001733544
reference translations 0.0017152349999999999
various metrics 0.001712197
evaluation 0.00170814
motivated metrics 0.001705283
linguistic information 0.001694989
semantic information 0.00168372
source sentence 0.0016587750000000001
target sentence 0.001585304
source sentences 0.0015575650000000001
source language 0.0015404070000000001
translation 0.00151215
english words 0.001505607
system level 0.001503717
fluency score 0.001488716
main dataset 0.00148843
good correlation 0.001478832
target language 0.001466936
phrase level 0.001442105
feature set 0.001440916
probability score 0.001440312
learning model 0.00143798
language models 0.0014370840000000001
linear correlation 0.0014366700000000001
sentence shares 0.001435627
word precision 0.001423254
training data 0.001417078
same level 0.001410605
based language 0.001408039
correct translations 0.001404687
large feature 0.001381045
function words 0.001379543
online news 0.001363467
different year 0.001355283
regression model 0.001349714
other features 0.0013447889999999999
correlation coefficient 0.001343348
xinhua news 0.001340432
content word 0.0013359370000000001
different kernels 0.001335091
poor translations 0.001331448
semantic meaning 0.001328999
assessment score 0.001322885
linguistic aspects 0.001321693
main trend 0.001321673
metrics 0.00132165
training set 0.0013197769999999998
rank correlation 0.001317228
such cases 0.001312158
afc news 0.001312044
large set 0.001308055
parse score 0.001307036
important information 0.001305308
