model feature 0.00338137
language model 0.00334105
model probability 0.002960987
exponential model 0.002756392
trigger model 0.002753462
model text 0.0027462679999999996
trigram model 0.002713054
model relevance 0.002577663
range model 0.002577131
adaptive model 0.0025243649999999998
model gains 0.002521734
wsj model 0.002520256
model selection 0.0025073779999999998
static model 0.002500029
mentation model 0.0024992419999999996
final model 0.002498999
guage model 0.0024989179999999997
tial model 0.002494031
gram model 0.002492088
stationary model 0.002487413
ger model 0.002486543
model ptri 0.0024859419999999997
model ratios 0.0024846499999999997
model 0.00226731
trigger word 0.002182342
word corpus 0.002177361
other word 0.00215846
current word 0.00209348
word pairs 0.002074653
next word 0.002013051
word history 0.002009532
word subset 0.001982226
such features 0.001967946
related word 0.001949807
word repetition 0.001936877
word closed 0.00192501
word posit 0.001917102
word cnn 0.001915895
language models 0.0018039689999999999
first features 0.00180018
several features 0.001751386
first words 0.00171514
other words 0.0016738
various features 0.001667085
linguistic features 0.001650686
candidate features 0.001623636
certain features 0.0016227379999999999
first feature 0.00161767
binary features 0.00156635
wsj features 0.001549516
fifteenth features 0.0015493859999999998
second feature 0.001547999
tdt features 0.001541883
certain words 0.0015376979999999999
content words 0.001533305
informative features 0.0015281609999999999
tionary features 0.0015148569999999999
features pertain 0.0015148569999999999
vance features 0.0015148569999999999
feature parameters 0.001500048
informative words 0.001443121
candidate feature 0.001441126
tent words 0.0014310389999999998
natural language 0.0014276699999999998
feature induction 0.0013883200000000002
range language 0.001383561
language processing 0.001377761
feature induct 0.0013469620000000002
sixth feature 0.001334184
feature induc 0.0013328860000000001
fifth feature 0.001332236
nineteenth feature 0.001332236
adaptive language 0.0013307949999999999
language mode 0.0013205769999999999
features 0.00129657
ral language 0.0012914519999999998
language central 0.0012909109999999998
probability distribution 0.001269849
ing data 0.0012608019999999999
linguistic data 0.001220043
exponential models 0.0012193109999999998
news corpus 0.001213863
words 0.00121153
test data 0.001194826
trigram models 0.0011759729999999999
target sentence 0.001169901
empirical probability 0.001158007
different documents 0.0011469170000000001
data stream 0.001132248
data con 0.001127994
broadcast news 0.001119624
wsj data 0.0011188729999999998
feature 0.00111406
topic similarity 0.001106749
different domains 0.001105623
various news 0.001103207
online data 0.001091044
reuters data 0.001089483
distance probability 0.001089114
multimedia data 0.001087671
