translation evaluation 0.00349176
reference translation 0.0026417289999999998
evaluation metrics 0.002637533
translation system 0.002559827
chinese word 0.002536001
word similarity 0.002468923
word segmentation 0.002462549
candidate translation 0.002380042
machine translation 0.0023203440000000002
english translation 0.002320266
general word 0.002287408
natural word 0.002243472
word boundaries 0.002230729
word segmenter 0.002230459
translation quality 0.002226915
word segmentations 0.00222378
ambiguous word 0.00222378
translation methods 0.0021949680000000003
prior translation 0.002126069
correlation bleu 0.0021122600000000004
correlation correlation 0.0021109
translation eval 0.002110893
translation outputs 0.002102098
evaluation tasks 0.002095288
tomatic translation 0.002089664
evaluation framework 0.002073495
high evaluation 0.002064441
metric score 0.002057676
manual evaluation 0.0020560129999999998
evaluation costs 0.0020068349999999998
human judgments 0.0019945510000000002
human judgment 0.00191814
same matching 0.0018049289999999998
translation 0.00178914
level score 0.001773557
level bleu 0.001705717
evaluation 0.00170262
matching problem 0.00166944
chinese words 0.0016665909999999998
rank correlation 0.001637874
phrase matching 0.001616171
level metric 0.001581933
different segmentation 0.001530554
matching algorithm 0.0015285429999999998
different machine 0.001523039
level scores 0.001519167
rank metric 0.00151545
reference translations 0.001509232
automatic scores 0.001465759
chinese reference 0.0014647599999999998
final score 0.001462111
matching process 0.001430577
different values 0.001428641
chinese task 0.001427523
system translations 0.00142733
bleu baseline 0.001421196
system level 0.0014195940000000002
pearson correlation 0.0014161220000000001
consistency correlation 0.0014097520000000002
different purposes 0.001408627
bleu award 0.001387106
matching characters这 0.0013867559999999998
gram matching 0.0013672699999999999
bleu base 0.0013632380000000001
language text 0.0013499929999999999
covered matching 0.001346889
matching rule 0.001344008
linguistic data 0.001337361
fluency scores 0.001314613
metric type 0.0013087910000000001
different studies 0.00130197
different characteristics 0.001297863
such phrases 0.001282906
metrics award 0.001265209
source sentence 0.00125676
tesla metrics 0.001250889
metric tesla 0.001249002
generation metrics 0.001247014
celab metric 0.001243344
processing task 0.001240976
uation metrics 0.001237542
chinese sentences 0.001229989
challenge task 0.001206247
recent research 0.0011934329999999998
such as下星 0.001191629
such as雨伞 0.001191629
data set 0.001190638
training data 0.001163401
nist data 0.001162484
programming problem 0.0011517699999999999
test set 0.001141028
task the 0.001139488
automatic machine 0.001126703
score 0.00112465
linear programming 0.001124402
language processing 0.001119109
task chance 0.00111766
target language 0.001113288
same expression 0.001079551
system parameters 0.001071422
