training data 0.00294414
such feature 0.002742341
feature set 0.002657893
standard feature 0.0025071390000000002
feature sets 0.002460228
small feature 0.002459672
feature values 0.0024433099999999998
data set 0.002415193
feature types 0.002368044
data size 0.002281975
test data 0.002280723
ing data 0.0022496499999999997
medium feature 0.002243295
many model 0.0022276600000000002
data error 0.002226095
data entropy 0.002212899
model form 0.002203214
language model 0.0021444430000000002
maxent model 0.002137443
original model 0.002079897
exact model 0.002070515
data times 0.00205987
model types 0.002056804
data sizes 0.00205555
data structure 0.002054929
data structures 0.002051854
model type 0.002030033
data matrix 0.002025008
data increases 0.002004497
single model 0.002002261
ferent data 0.002001158
data correctness 0.001997685
feature 0.00199402
level model 0.001971737
training algorithm 0.0018320760000000002
other models 0.001831769
same training 0.001766634
model 0.00168278
training entropy 0.001654399
maxent training 0.001647483
conditional models 0.0015589560000000002
ent training 0.001541963
training instance 0.0015385630000000002
features sets 0.001530728
joint models 0.001524474
entropy models 0.001519479
maxent models 0.0015125630000000002
typical training 0.001500215
training instances 0.001483314
training matrix 0.001466508
tag features 0.001439949
other words 0.001412621
ent models 0.0014070430000000002
ditional models 0.001330563
correlated features 0.0013126869999999999
dense features 0.0013126869999999999
word selection 0.001297391
word sense 0.001282202
indicator function 0.00121352
objective function 0.001198718
word clustering 0.001193707
training 0.00119282
standard set 0.001176992
function measure 0.001166012
same size 0.0011044689999999999
same test 0.001103217
jective function 0.0010948849999999999
other hand 0.0010884179999999999
other improvements 0.00108281
similar words 0.001071234
features 0.00106452
models 0.0010579
many output 0.0010511309999999999
simple algorithm 0.001050174
same results 0.001045805
such applications 0.001040642
set selection 0.0010406
conditional maximum 0.001013886
test error 0.001004178
same number 0.0010032399999999999
performance level 0.00100146
same value 9.99622E-4
prior probability 9.98672E-4
parameter optimization 9.941239999999999E-4
test entropy 9.90982E-4
learning problems 9.90065E-4
new algorithm 9.80048E-4
test corpus 9.75708E-4
maximum entropy 9.744090000000001E-4
many problems 9.72535E-4
words train 9.72454E-4
many experiments 9.712919999999999E-4
ing language 9.599929999999999E-4
conditional maxent 9.557190000000001E-4
output values 9.55541E-4
gis algorithm 9.442270000000001E-4
maximum number 9.42256E-4
space results 9.3901E-4
problem domain 9.37722E-4
ing sentence 9.324229999999999E-4
