word model 0.00387487
word frequency 0.002722792
common word 0.002625384
informative word 0.002621449
accent model 0.0026160899999999997
word informa 0.002612088
word informativeness 0.002578884
word changes 0.002542471
word accentuation 0.00254104
word importance 0.002533663
empty word 0.002532649
ranking word 0.002524002
model models 0.0023729809999999997
probability model 0.002340878
pos model 0.00220951
other words 0.002152671
tts model 0.002082076
baseline model 0.002061514
context words 0.002039325
semantic features 0.001981638
function words 0.0019598899999999997
idf model 0.001959518
informative words 0.001918229
content words 0.001892085
ipper model 0.001890272
prominent words 0.0018497849999999998
reference model 0.001839987
model signif 0.001839077
ttswic model 0.001839077
tween words 0.001834184
empty words 0.001829429
pitch accent 0.0018246360000000001
capitalized words 0.001822129
formative words 0.0018212839999999998
accentable words 0.0018212839999999998
speech corpus 0.001815369
accent models 0.001798871
other features 0.0016465009999999999
pos information 0.0016108630000000001
model 0.0015951
words 0.00157655
other information 0.001572574
feature correlation 0.001543826
tts accent 0.0015079660000000001
pos tag 0.001480595
text corpus 0.0014772230000000002
powerful feature 0.0014755880000000001
new information 0.001469526
context information 0.001459228
valuable feature 0.00145121
accent modeling 0.001434189
many features 0.001433474
accent status 0.001422401
accent mode 0.001393344
pos models 0.001392291
idf accent 0.001385408
new corpus 0.001385376
work information 0.001377559
discourse features 0.0013723429999999998
accent prediction 0.0013498540000000002
accent label 0.001344066
large corpus 0.001337228
global features 0.001326977
training data 0.001326312
internal features 0.001319505
information content 0.001311988
corpus table 0.00131016
tic information 0.001299423
accent assignment 0.001297356
accent placement 0.001291862
tts pitch 0.001290622
information retrieval 0.001286232
default accent 0.0012746510000000001
expert accent 0.0012727110000000002
information gain 0.001271749
comprehensive accent 0.001271341
accent indicators 0.001269869
accent patterns 0.001269815
medical corpus 0.001266597
accent assignments 0.001266386
reference accent 0.001265877
prehensive accent 0.001265002
purpose information 0.0012522240000000001
information focus 0.001250432
semantic informa 0.001243576
semantic distance 0.00124354
contrast information 0.0012414869999999999
popular information 0.0012414869999999999
hmm models 0.001233923
models hmm 0.001233923
speech synthesis 0.001217645
speech portion 0.001212915
semantic informativeness 0.001210372
markov models 0.001207406
corpus description 0.001204898
semantic similarity 0.001202004
feature 0.00119982
speech recognition 0.001190857
speech transcripts 0.001187738
entire corpus 0.001172196
