different corpus 0.002835763
learner corpus 0.002724203
english corpus 0.002334558
corpus approach 0.0023037739999999998
native corpus 0.002210147
jle corpus 0.002161964
news corpus 0.002154335
clc corpus 0.002106155
corpus 0.00185272
other classifiers 0.001742863
different corpora 0.0017194150000000002
different test 0.001717569
different classifiers 0.001711636
different methods 0.0016697790000000001
base classifier 0.001632293
other base 0.001620773
learner corpora 0.001607855
different base 0.001589546
feature set 0.001504966
different classification 0.00150061
test data 0.001493041
feature extraction 0.001449288
several corpora 0.0014194099999999999
training set 0.001402234
learning approach 0.001361479
language learning 0.001354866
individual classifier 0.001351606
base classifiers 0.001335096
different characteristics 0.001291712
several base 0.001289541
different parts 0.001275497
different sources 0.001269444
classification method 0.001256134
training instances 0.0012509889999999999
test set 0.001240721
different datasets 0.0012401460000000001
error correction 0.001238277
relative performance 0.001221593
japanese learner 0.001220545
classification methods 0.0012043029999999999
article error 0.0012023210000000001
error rate 0.001194007
sequential learning 0.001191339
learning process 0.001189238
cambridge learner 0.0011876830000000001
structural learning 0.0011849109999999999
data sets 0.001178235
effective features 0.001147462
linear classifiers 0.001136034
known features 0.001131454
multiple corpora 0.001126137
base classification 0.00112407
multiple classifiers 0.001118358
grammatical error 0.001113491
native corpora 0.001093799
appropriate word 0.001089886
english text 0.001086885
word selection 0.001086456
noun phrase 0.001082029
development data 0.001077833
small corpora 0.001072538
individual corpora 0.001062188
individual classifiers 0.001054409
noun phrases 0.001049225
correct form 0.001047925
further research 0.001045772
base learners 0.001044012
jle test 0.00104377
correction task 0.001032275
overall performance 0.00102841
classifier 0.00102579
method our 0.001019671
our method 0.001019671
novel method 0.001010965
data imbalance 0.001010328
correction problem 0.001007249
development methods 0.001006054
feature 9.98771E-4
fce test 9.93034E-4
clc test 9.879609999999999E-4
respective corpora 9.861499999999999E-4
heterogeneous corpora 9.861499999999999E-4
classifiers conflicts 9.788000000000002E-4
model bias 9.77979E-4
gec model 9.755639999999999E-4
performance gain 9.653750000000001E-4
auxiliary problem 9.439730000000001E-4
integ model 9.39413E-4
several organizations 9.36168E-4
relative improvement 9.28161E-4
infocomm research 9.24252E-4
surrounding context 9.147059999999999E-4
selection task 9.106679999999999E-4
learning 9.10425E-4
same class 9.06202E-4
additional information 9.032479999999999E-4
article errors 9.02507E-4
training 8.96039E-4
common structure 8.859359999999999E-4
selection problem 8.856420000000001E-4
