features word 0.00252246
language model 0.002366019
kernel word 0.002070305
word sequence 0.002001958
class model 0.0019705729999999998
entropy model 0.001947744
test sentences 0.001941262
markov model 0.0019155519999999998
negative sentences 0.001909674
correct sentences 0.001890027
different features 0.001835239
incorrect sentences 0.001814254
model sample 0.0018123009999999999
classifier word 0.001809976
sample sentences 0.001795991
positive sentences 0.0017806520000000002
word string 0.001739764
correct sentence 0.001735107
sentences return 0.001734471
guage model 0.001725514
sentence classification 0.001724101
sentence sampling 0.001720185
training data 0.0017016169999999999
inative model 0.001685336
rect sentences 0.001675417
word occurrences 0.001672652
word strings 0.001672652
many features 0.0016694639999999998
ined sentences 0.001668861
language models 0.001662892
sentence maximum 0.0016497360000000002
such training 0.0016017359999999999
training corpus 0.001586194
whole sentence 0.001544082
sentence discrimination 0.0015288250000000001
cient sentence 0.00151776
such models 0.0014877010000000001
preceding words 0.001474622
distinct words 0.001460073
compound words 0.001457811
mon words 0.001457811
model 0.00144894
sentences 0.00143263
few features 0.001391314
training examples 0.0013880709999999998
candidate features 0.001385128
new feature 0.0013667660000000002
other sampling 0.001362814
feature vector 0.0013462180000000002
smcm features 0.001342146
relevant features 0.0013394890000000001
explicit features 0.001337677
learning method 0.001334495
training set 0.001330238
latent features 0.001327746
tent features 0.001325484
distinct features 0.001325463
kernel method 0.001323827
training time 0.00131629
data set 0.001312159
language modeling 0.001307046
evaluation data 0.001290493
learning algorithm 0.001287398
sentence 0.00127771
training samples 0.001257483
entropy models 0.001244617
other applications 0.001226507
words 0.0012206
native language 0.001219023
probabilistic language 0.001212495
discriminative language 0.001194682
bilistic language 0.001189509
feature item 0.001182755
local information 0.00118146
other hand 0.001174977
training sets 0.001162267
information retrieval 0.0011589909999999998
inative language 0.001153475
criminative language 0.001153377
feature weighting 0.001140182
batch training 0.001136915
sampling method 0.0011324669999999998
overlapping information 0.001131453
input training 0.0011286199999999999
latent information 0.001123899
tent information 0.001121637
essential information 0.001119375
correlation information 0.001119375
learning methods 0.0011164780000000002
feature functions 0.001116177
many examples 0.001111697
training exam 0.001108243
classification learning 0.001090894
likelihood method 0.00108622
features 0.00108599
online learning 0.001072006
online algorithm 0.001070398
ing examples 0.0010636539999999998
different sets 0.001051668
possible number 0.0010398809999999999
