language model 0.00308599
training data 0.002607616
linear model 0.002459621
standard model 0.002407415
english data 0.002377844
model adaptation 0.002336403
data feature 0.002327375
model comparison 0.0023201629999999997
separate model 0.00231666
entropy model 0.0023144439999999997
model priors 0.002294433
data performance 0.0022846539999999997
model compari 0.002242885
perior model 0.002242885
same data 0.002225521
data set 0.002206455
large data 0.002166988
test data 0.0021263989999999997
evaluation data 0.0020969689999999997
esl data 0.002058723
model 0.00202611
data sets 0.0020163299999999998
much data 0.001996814
native data 0.001972474
linguistic data 0.0019719769999999998
annotated data 0.00197051
ferent data 0.001912823
tive data 0.001907242
data consortium 0.001867278
ble data 0.001867204
data asap 0.001866298
language models 0.0018012079999999999
source language 0.0017372630000000002
other training 0.001645564
first language 0.001619357
different training 0.0016002999999999998
same error 0.001545132
training set 0.001514631
other models 0.001428996
word sense 0.001400236
second language 0.001398692
esl error 0.0013783340000000002
language modeling 0.001370437
english text 0.001364692
confusable word 0.001340612
error correction 0.001334581
natural language 0.001333905
training sets 0.001324506
new training 0.001317653
language total 0.0013173450000000001
common error 0.001315835
learning models 0.001314883
feature performance 0.0013125889999999999
training example 0.001306127
language mod 0.0012879060000000001
articles language 0.001286313
native training 0.00128065
language applica 0.0012763660000000001
error types 0.001268583
training corpora 0.001265879
same feature 0.001253456
english sentences 0.001245068
specific error 0.001241135
particular error 0.001239144
legitimate words 0.001238821
error regularities 0.001235548
feature set 0.0012343900000000001
feature preposition 0.001234082
different features 0.001224846
training sizes 0.001222409
confusable words 0.0012082739999999998
other results 0.001206447
error patterns 0.001200111
error statistics 0.001193934
error pat 0.001186316
preposition errors 0.0011812699999999999
training exam 0.001174547
same features 0.001158243
performance results 0.001153713
first corpus 0.001147688
algorithm performance 0.001106153
other esl 0.001096671
candidate preposition 0.001094087
english prepositions 0.001081198
language 0.00105988
different distribution 0.001051602
different esl 0.0010514069999999999
native english 0.001050878
learning algorithm 0.001044774
feature sets 0.001044265
correct preposition 0.0010392069999999999
source prepositions 0.001030457
preposition context 0.0010121359999999999
different systems 0.0010103439999999998
linear learning 0.001007066
other candidates 0.001000905
feature weights 0.001000673
english speakers 9.976870000000001E-4
esl corpus 9.972140000000002E-4
learning methods 9.94342E-4
