system combination 0.003674839
combination system 0.003674839
correction system 0.003543261
smt system 0.0034870689999999998
translation model 0.0033778799999999998
system performance 0.003354051
system hypothesis 0.003305563
multiple system 0.003213072
evaluation system 0.003213013
single system 0.003203911
final system 0.00314657
system outputs 0.003133949
system output 0.003123922
component system 0.003110931
additional system 0.003059281
system combina 0.003052425
combined system 0.003047491
rection system 0.003045813
system example 0.003039398
gec system 0.003035669
corrected system 0.003020223
pipeline system 0.0030180299999999997
system com 0.0030150479999999998
based system 0.003002555
camb system 0.002997164
nent system 0.00299663
nus system 0.002993966
cuui system 0.002993966
superior system 0.002993966
phrase translation 0.002943507
machine translation 0.0028663070000000002
system 0.00274495
translation table 0.002699206
translation approaches 0.002640347
translation tables 0.0025086880000000002
chine translation 0.0025010370000000002
training data 0.002355602
test data 0.002348708
word alignment 0.002331956
translation 0.00225113
word form 0.002195565
data set 0.0021873080000000002
different error 0.002146459
correction systems 0.002140481
possible word 0.00211622
word order 0.002104701
smt systems 0.002084289
language model 0.0020788589999999997
word level 0.002034074
plural word 0.0020141630000000002
error correction 0.001986021
other systems 0.001979847
model feature 0.001968892
development data 0.001966806
data sets 0.001961081
word choice 0.001960449
word sur 0.001952227
data structure 0.0018969490000000002
learner data 0.001893297
ment data 0.001871129
multiple systems 0.001810292
model score 0.001755507
number error 0.001748673
component systems 0.001708151
multiple error 0.001655832
grammatical error 0.001648416
top systems 0.001646853
single error 0.0016466710000000002
whole systems 0.001645267
combined systems 0.0016447109999999999
individual systems 0.0016277029999999999
pipeline systems 0.00161525
reordering model 0.00160656
systems increases 0.001594901
nent systems 0.00159385
ple systems 0.001593513
common error 0.001564441
error analysis 0.001561826
error cor 0.001555411
model log 0.0015437299999999999
specific error 0.001514134
error type 0.001492899
error types 0.001482498
combination method 0.001477585
individual error 0.001473243
various error 0.001470446
noun errors 0.0014649350000000001
number errors 0.0014616569999999999
sva error 0.001458244
language models 0.001439511
combination approach 0.001438866
guage model 0.0014312499999999998
lation model 0.0014303839999999998
correction task 0.0014159350000000001
test set 0.001398816
different classifiers 0.001396112
form errors 0.001394339
model formulation 0.001376308
english corpus 0.001374653
large english 0.001365121
