feature set 0.00444065
different feature 0.004258278
first feature 0.004119863
feature values 0.004065926
feature vector 0.004013742
feature selection 0.004012154
active feature 0.003948448
feature types 0.003921863
feature sets 0.003914599
additional feature 0.003906319
feature prediction 0.003896971
particular feature 0.003878045
feature pruning 0.00382619
third feature 0.003822231
ond feature 0.003820332
bias feature 0.003819433
heavy feature 0.003815007
feature 0.00357215
same features 0.002807446
features standard 0.002577945
individual features 0.002566869
active features 0.002561228
certain features 0.002532144
additional features 0.002519099
distance features 0.0025078599999999998
binary features 0.002469992
sparse features 0.002466332
rich features 0.002446682
bag features 0.0024430679999999996
contextual features 0.0024327479999999998
extracted features 0.002432475
specific features 0.002429956
sic features 0.0024279339999999996
model accuracy 0.002260858
language model 0.002240368
features 0.00218493
model comparison 0.002058069
mixture model 0.001993814
scalable model 0.001972245
training data 0.001970139
model parametrization 0.001943954
training time 0.001798674
model 0.00170097
training dataset 0.001539154
training datasets 0.001523803
training sets 0.001485189
decrease training 0.001424117
data the 0.001167002
other words 0.001164841
target word 0.00114538
training 0.00114274
development set 0.001120974
entropy classifier 0.001120809
reasonable set 0.001111328
optimization method 0.001091208
partition function 0.001086233
word vocabulary 0.001082386
development data 0.001079873
data sparsity 0.001070248
linear models 0.001062112
parameter estimation 0.001048172
probabilistic classifier 0.001038621
other smoothing 0.001020028
parameter vector 9.94633E-4
novel approach 9.63139E-4
generative language 9.57787E-4
maximum entropy 9.560770000000001E-4
different sizes 9.45209E-4
train models 9.42294E-4
other classi 9.40476E-4
conditional probabilities 9.315650000000001E-4
same order 9.23346E-4
large datasets 9.19289E-4
language modeling 9.18735E-4
interpolation models 9.149939999999999E-4
large number 9.074740000000001E-4
classification problem 9.026749999999999E-4
possible words 9.011679999999999E-4
probability dis 8.891039999999999E-4
same conditioning 8.88603E-4
input fea 8.77598E-4
entropy perceptron 8.76751E-4
natural language 8.65337E-4
context words 8.53165E-4
polated form 8.497349999999999E-4
contextual information 8.40728E-4
estimation the 8.34734E-4
active fea 8.32982E-4
language applications 8.2548E-4
classifier consis 8.242920000000001E-4
function 8.22536E-4
regular words 8.10322E-4
parameter estimates 8.04825E-4
first stage 8.002580000000001E-4
able parameter 7.972789999999999E-4
unigram distribution 7.95613E-4
relative performance 7.84168E-4
the experiments 7.72772E-4
standard comparison 7.50114E-4
method 7.4858E-4
