dialogue act 0.00543357
different dialogue 0.004937019
useful dialogue 0.0046010930000000005
dialogue acts 0.004581189
general dialogue 0.004544195
hat dialogue 0.004538126
nal dialogue 0.004511705
tain dialogue 0.004510779
irrelevant dialogue 0.004508956000000001
duce dialogue 0.004505205
ging dialogue 0.004500509000000001
undesirable dialogue 0.004500067
computing dialogue 0.004498538
propriate dialogue 0.004496426
desirable dialogue 0.004496426
dialogue 0.00423993
speech recognition 0.00220969
system gen 0.002127369
other features 0.001856681
system 0.00179446
several features 0.001704168
act cues 0.001627966
various features 0.001601252
act cue 0.001565777
act tagging 0.001553921
following features 0.0015502950000000001
logue act 0.00154704
ing information 0.001532241
different utterances 0.001525296
alogue act 0.00151239
relevant features 0.00149736
act tagg 0.001452199
abstract features 0.001447575
internal features 0.001445556
current utterance 0.0014241100000000001
training data 0.001413107
surface speech 0.001410249
many rules 0.001371818
speaker information 0.0013498569999999999
ing utterances 0.001335598
ing data 0.001327338
different set 0.001310414
training corpus 0.001307194
preceding utterance 0.0012907370000000001
data problem 0.001281049
tire utterance 0.0012728470000000001
new data 0.001242529
new feature 0.001241396
many word 0.001225352
ing corpus 0.001221425
other research 0.001216819
learning method 0.001212242
large number 0.00120669
training set 0.001206485
training time 0.0011988950000000002
machine learning 0.001190277
discourse understanding 0.001186875
tagging task 0.001177636
potential rules 0.001175514
features 0.00117387
speech 0.00115297
different models 0.001150784
speaker feature 0.001143821
new corpus 0.001136616
relevant rules 0.001128444
discourse processing 0.001124054
general rules 0.001109219
learning approaches 0.0011086569999999999
preceding utterances 0.001103204
monte data 0.001102845
this feature 0.001098185
feature interactions 0.0010951379999999998
quent utterances 0.001086618
ferent rules 0.001084334
sparse data 0.001083762
unseen data 0.001079587
data sup 0.001078179
much data 0.001078179
untagged data 0.001078179
challenging task 0.001075323
irrelevant rules 0.00107398
infinite number 0.0010663740000000001
overwhelming number 0.0010663740000000001
separate rules 0.001065471
evant rules 0.0010637469999999999
vant rules 0.0010637469999999999
fourth rules 0.0010637469999999999
recognition 0.00105672
first rule 0.0010381890000000001
ing confidence 0.0010305940000000001
information 0.00102485
test corpus 0.001021573
useful word 0.001019651
utterance 0.00101574
learning algorithm 0.00100972
different domains 0.001003936
chine learning 0.001003733
specific word 9.99332E-4
particular domain 9.97967E-4
small set 9.94603E-4
