system combination 0.002923242
new system 0.002787415
same system 0.002768227
machine translation 0.00270153
task system 0.002697308
different model 0.00258405
test data 0.00257429
system comparisons 0.002565586
system outputs 0.002540438
translation implementation 0.0025170030000000003
group system 0.002458725
system correlation 0.002457219
system pairs 0.002453809
incremental system 0.002448634
system combi 0.002444981
system com 0.002435955
available system 0.002428525
workshop system 0.002407232
system changes 0.002401824
comparable system 0.002398934
system variation 0.0023973270000000003
translation our 0.0023868220000000002
test set 0.00231665
test bleu 0.002223689
test corpus 0.002155256
new test 0.002142815
system 0.00214254
translation 0.0021272
same test 0.002123627
large test 0.00207934
significance test 0.0020747219999999998
other test 0.002038318
same model 0.001997567
possible test 0.001989365
chinese test 0.00195277
bleu different 0.0019379190000000002
small test 0.001919034
sample test 0.00189694
test sets 0.0018885179999999998
test size 0.001877189
word alignment 0.0018598339999999999
news test 0.00185777
different models 0.001812344
test gain 0.0018117329999999998
hansard test 0.001803201
test corpora 0.0017875819999999998
test distribution 0.0017805619999999999
dependent test 0.0017699419999999998
different approach 0.0017638100000000002
various test 0.00176246
resampled test 0.0017572999999999998
simulated test 0.001752899
fixed test 0.0017515039999999999
test instances 0.0017491219999999999
base model 0.0017425539999999999
different research 0.00173499
model type 0.001727533
new data 0.001721225
model types 0.001704375
simple model 0.001697312
training set 0.001697226
single systems 0.001694453
different source 0.001658619
similar systems 0.00164466
different base 0.001582844
word aligners 0.0015535199999999999
previous systems 0.001549397
unsupervised word 0.0015434959999999999
different sys 0.00153873
new training 0.001523391
such data 0.001522374
english sentences 0.001521431
rouge different 0.00151195
nlp systems 0.001502419
different groups 0.001494709
different runs 0.00148607
systems par 0.001475723
different ranges 0.001472854
past systems 0.0014649010000000002
different trends 0.001464225
aer different 0.001464225
systems gener 0.001451361
same set 0.001444397
ferent systems 0.001442395
trained systems 0.00143796
dependency parsing 0.001403435
set performance 0.001390416
nal data 0.001373746
model 0.00137188
chinese training 0.001333346
original training 0.001282016
training sets 0.001269094
bleu score 0.0012600649999999999
parsing task 0.001243999
many corpus 0.001240822
training cor 0.001231981
english translations 0.0012150569999999999
different 0.00121217
combination comparisons 0.001203748
hansard training 0.001183777
