error correction 0.00260045
model score 0.00253263
correction score 0.00230842
language model 0.002261427
correction system 0.00221133
test data 0.002198481
input word 0.002144923
other error 0.0021408599999999996
training data 0.002069278
spelling correction 0.002067021
data set 0.002016767
task data 0.001989715
decoder model 0.0019802789999999997
word alignments 0.001955342
error corrections 0.0019402069999999998
linear model 0.001924552
single word 0.001883891
plural word 0.001864105
model weight 0.001860607
new error 0.00185794
baseline error 0.0018516829999999998
grammatical error 0.001846286
word surface 0.001844838
noun phrase 0.001832322
model expert 0.001825318
expert model 0.001825318
possible correction 0.001787963
same score 0.001754242
development data 0.001752248
data sets 0.001750816
particular error 0.0017501469999999999
hoo data 0.001741468
error correc 0.001723414
correction decoding 0.0017053939999999998
model the 0.001695166
error cat 0.001692465
guage model 0.001681515
specific error 0.001680692
coder model 0.001671778
nel model 0.001669861
hypothesis correction 0.0016644149999999998
error categories 0.0016590539999999999
tuning data 0.001657231
number correction 0.001648917
matical error 0.0016475889999999999
article correction 0.0016471289999999998
error category 0.0016450089999999998
data drops 0.001642339
skewed data 0.001641013
other words 0.001632881
search decoder 0.001627949
grammar correction 0.001624944
correction step 0.0016215919999999998
preposition correction 0.001612873
bleu score 0.001608694
correction systems 0.001607601
maximum phrase 0.001594535
hypothesis score 0.0015703549999999998
phrase length 0.001555754
phrase extraction 0.0015417719999999998
system output 0.0015284
smt system 0.001526463
correction count 0.0015150019999999999
valid correction 0.0015119069999999998
expert score 0.001507048
punctuation correction 0.0015041149999999999
search tree 0.0014990510000000002
noun errors 0.0014879519999999999
correction steps 0.001479912
next search 0.0014777660000000001
prepositional phrase 0.001467898
baseline system 0.001462563
likely correction 0.001453279
ror correction 0.0014525999999999999
search beam 0.0014525170000000002
beam search 0.0014525170000000002
correction choices 0.001450771
ber correction 0.0014462479999999998
aggregated correction 0.0014456019999999998
search procedure 0.001441378
confidence score 0.001439998
model 0.00142545
test set 0.001422988
metric score 0.001402174
overall score 0.0013999029999999999
error 0.00139921
test sentences 0.0013915260000000001
right score 0.001387127
input sentence 0.001383637
system edit 0.0013749489999999999
search iteration 0.001370768
search space 0.0013697870000000001
different decoder 0.001367119
average score 0.001366527
delta score 0.001358185
pothesis score 0.001355952
score definition 0.001353356
system edits 0.001343253
ing errors 0.001328232
number errors 0.001323777
