lexical features 0.0028933020000000004
prosodic features 0.002888685
speech data 0.0026110300000000003
prosody features 0.002537264
binary features 0.002504257
parse features 0.002432093
local features 0.002386104
ical features 0.002378565
ngram features 0.002371962
independent features 0.002363467
gether features 0.002363467
training data 0.002349625
lexical model 0.002258042
text data 0.002217281
different feature 0.002199437
features 0.00213025
speech training 0.002000175
unlabeled data 0.001952823
adaptation data 0.0018679170000000002
model selection 0.0018378650000000002
mrda data 0.001831498
wiki data 0.001810792
source data 0.001803005
target data 0.001801403
feature vector 0.001780002
speech act 0.001764758
single feature 0.001758539
various feature 0.001757046
feature sets 0.001755694
model predictions 0.0017492340000000001
textual data 0.001745659
original feature 0.001720959
data the 0.001716896
speech recognition 0.001701161
feature representation 0.001686768
dimensional feature 0.001661749
feature types 0.001638481
feature vec 0.001618839
unlabeled speech 0.001603373
speech domain 0.001586144
act corpus 0.001568748
lexical models 0.0015477009999999999
prosodic models 0.0015430840000000001
news speech 0.001542141
training set 0.001526658
model 0.00149499
labeled speech 0.0014812900000000001
automatic speech 0.001479074
labeling speech 0.001471684
word sequence 0.0014423740000000002
speech corpora 0.001431796
ent speech 0.001428304
conversational speech 0.001415038
different question 0.0014094490000000001
first word 0.001401398
spontaneous speech 0.001384664
feature 0.0013839
tional speech 0.001383241
unannotated speech 0.0013780160000000001
speech waveforms 0.001364354
initial word 0.001354582
training utterances 0.001312897
new training 0.001307456
mrda corpus 0.001286038
text sentences 0.001279315
original corpus 0.0012718389999999999
word fragments 0.0012672200000000001
other sentences 0.001260542
test set 0.001240396
tag question 0.0012251739999999999
mrda training 0.001220643
act classification 0.001201454
wiki training 0.0011999369999999999
other cases 0.001198555
text domain 0.001192395
correspondence learning 0.001189773
work question 0.001189504
lexical system 0.0011846299999999999
initial words 0.00117081
news text 0.001148392
classifier lexical 0.001145192
question detection 0.001143174
baseline results 0.0011425530000000001
speech 0.00113079
prosodic fea 0.001122212
previous work 0.001112762
different languages 0.001108539
other adaptation 0.001105945
detection problem 0.00110172
additional text 0.00109537
scribe words 0.001090166
labeled text 0.001087541
act tags 0.00106667
regression models 0.001065969
different portuguese 0.001062841
text source 0.001059806
different authors 0.001053943
adaptation set 0.00104495
question type 0.001040371
other punctuation 0.001040018
