language model 0.00315466
other language 0.00271033
language models 0.002549437
language modeling 0.002318868
statistical language 0.002276919
second language 0.002251992
language processing 0.002178542
other features 0.00214584
model training 0.0021431849999999997
trigram language 0.002128374
natural language 0.00210683
language mod 0.00209909
gram language 0.002088969
probabilistic language 0.0020867999999999998
language learners 0.002085738
ond language 0.002083072
sri language 0.00208118
language competency 0.002080717
training data 0.002051035
other data 0.00205068
word frequency 0.001958937
test data 0.0019334780000000002
vocabulary words 0.0019102939999999999
complex words 0.001872912
unique words 0.001861517
language 0.00186108
discriminative words 0.001846094
data set 0.00184462
words britannica 0.0018308119999999998
unknown words 0.001821666
classifier features 0.001804442
model classifier 0.001801432
num words 0.001794995
ual words 0.001769101
perplexity features 0.00174348
model perplexity 0.00174047
word sequence 0.001719713
word usage 0.001702352
word lists 0.0016971920000000001
word instances 0.001696489
following features 0.001682678
ular word 0.001677307
word fre 0.0016772150000000001
static word 0.0016747180000000002
other text 0.001644365
only features 0.001642495
model scores 0.001622793
svm model 0.0016008399999999999
rate features 0.0016007249999999999
traditional features 0.001593959
feature selection 0.0015830660000000002
selection feature 0.0015830660000000002
ing data 0.001579446
particular features 0.001576461
parse features 0.001563591
standard data 0.0015563340000000001
words 0.00154734
other models 0.0015376069999999999
data sets 0.0015323370000000002
training set 0.0014927949999999999
reader data 0.00146897
feature space 0.001464238
training corpus 0.001458008
treebank data 0.0014526220000000002
level text 0.0014505310000000001
development data 0.001446545
data points 0.001434093
text classification 0.001433297
feature selec 0.0014189930000000001
explicit feature 0.001418007
different grade 0.0014150950000000001
test set 0.001375238
different texts 0.001319938
other detection 0.001315641
different categories 0.0013023050000000001
features 0.00129659
model 0.00129358
grade training 0.001286547
detection error 0.001279026
web text 0.001248149
different formats 0.0012366950000000001
other methods 0.001230018
different thresh 0.0011993190000000001
feature 0.0011972
training texts 0.00119139
classification work 0.0011840140000000002
other classes 0.0011770959999999999
level classifier 0.001163268
text document 0.001158937
svm training 0.001156865
negative training 0.001153846
other fea 0.001150003
level classifiers 0.001149645
many classification 0.001139108
large corpus 0.0011370920000000001
news text 0.001121319
cost error 0.001118507
error rate 0.00111677
first set 0.001105728
separate training 0.001099913
