different features 0.002198989
feature value 0.001978877
important feature 0.001916425
important features 0.001896655
feature values 0.00188923
feature figure 0.00182832
feature weighting 0.001786284
training data 0.00178308
matching feature 0.0017563350000000001
feature vectors 0.00174923
feature relevance 0.0017392430000000001
third feature 0.001733174
noisy features 0.001713675
informative features 0.001711488
irrelevant features 0.001710664
symbolic features 0.001710664
feature 0.00151409
ing data 0.001511504
features 0.00149432
data problem 0.0014279190000000001
second function 0.0013926540000000001
word form 0.001353783
voting function 0.001349616
different information 0.0013468
new data 0.0013432980000000001
training set 0.001301785
classifier algorithm 0.001272869
sparse data 0.001272104
tive method 0.0012409180000000001
form information 0.001215855
polation method 0.001213432
voting method 0.001203272
naive model 0.001193688
smoothing algorithm 0.0011713539999999999
anguage model 0.001169348
training corpus 0.0011598749999999999
function 0.00111947
test set 0.0010740329999999998
training cases 0.001070152
machine learning 0.001063567
training patterns 0.001047839
other methods 0.001039252
class label 0.001021485
unknown word 0.001014843
iterative training 0.001011859
context fea 9.960160000000002E-4
learning literature 9.89352E-4
contextual information 9.84417E-4
method 9.73126E-4
conditional distribution 9.64054E-4
training material 9.63485E-4
such cases 9.48768E-4
first noun 9.43812E-4
statistical information 9.41234E-4
such applications 9.35851E-4
model 9.34866E-4
information gain 9.32819E-4
same results 9.28063E-4
selection task 9.27549E-4
various information 9.18669E-4
information sources 9.18098E-4
class labels 9.09864E-4
conditioning information 9.09767E-4
ing parameters 8.90409E-4
diverse information 8.87529E-4
conditional probability 8.87187E-4
rich information 8.75671E-4
other weighting 8.745140000000001E-4
suffix information 8.72097E-4
relevant information 8.70969E-4
affix information 8.60037E-4
many fea 8.53694E-4
right context 8.463870000000001E-4
test cases 8.424E-4
language modeling 8.36835E-4
ture set 8.306399999999999E-4
algorithm 8.25854E-4
other neigh 8.19321E-4
diverse set 8.011229999999999E-4
important problem 7.93234E-4
noun phrase 7.930319999999999E-4
rect class 7.91739E-4
schematic form 7.91635E-4
distribution assumptions 7.87435E-4
smoothing methods 7.82432E-4
language processing 7.80284E-4
separate test 7.79943E-4
same type 7.7137E-4
first letter 7.711630000000001E-4
learning 7.68274E-4
value patterns 7.66566E-4
possible condition 7.62358E-4
language mod 7.597960000000001E-4
complex language 7.59696E-4
large number 7.59205E-4
few parameters 7.58271E-4
first place 7.56087E-4
linguistic tasks 7.550899999999999E-4
unknown words 7.52713E-4
language process 7.463960000000001E-4
