word segmentation 0.0032631780000000003
language model 0.003259183
standard word 0.0031591230000000002
word forms 0.003129865
word type 0.003096974
word boundary 0.003094903
unknown word 0.003087223
word form 0.003086999
word sequence 0.0030853720000000003
surface word 0.003081998
phonetic model 0.003075497
word boundaries 0.0030741650000000002
new word 0.0030740010000000003
same word 0.003055097
word tokens 0.0030270180000000002
intended word 0.003011925
gold word 0.002992218
word frequencies 0.0029700530000000003
word types 0.002966952
next word 0.002958159
known word 0.002956743
rare word 0.0029280160000000003
clusters word 0.002926419
word segmenta 0.0029261
unlabeled word 0.002924278
mic word 0.002921157
first model 0.002847692
model features 0.002837253
acoustic model 0.002813981
bayesian model 0.002751485
model scores 0.0026736950000000002
model posterior 0.002646877
generative model 0.002639106
guage model 0.002613758
realistic model 0.002602791
mentation model 0.0026004260000000003
noise model 0.002590495
model deficient 0.0025785450000000002
context words 0.002362949
model 0.00235825
other words 0.002225064
surface words 0.0020725780000000003
frequent words 0.002052086
separate words 0.002013223
intended words 0.002002505
segment words 0.001933014
rare words 0.001918596
ticular words 0.001912965
words 0.00169161
phonetic corpus 0.001563802
phonetic pronunciation 0.0013979679999999999
first language 0.0013903750000000001
language modeling 0.001326471
bigram language 0.001265697
training data 0.001254485
early language 0.001205506
gram language 0.001156276
context set 0.001144482
complex language 0.001133137
context parameters 0.001130764
language acquisition 0.001128887
phonetic categories 0.001128003
artificial corpus 0.001122029
infant language 0.001121416
our corpus 0.001119717
probability state 0.001118064
common pronunciation 0.001093896
prior distribution 0.00109298
phonetic changes 0.0010840679999999999
buckeye corpus 0.0010825700000000001
ralistic corpus 0.001080149
context information 0.001072318
inference process 0.0010692710000000001
transducer distribution 0.001068865
ratner corpus 0.00106814
cial corpus 0.00106814
different distributions 0.001065669
same distribution 0.001049204
phonetic input 0.001047199
phonetic side 0.001043064
phonetic knowledge 0.001039946
previous models 0.00103758
phonetic con 0.001033313
different parts 0.0010308729999999999
transducer probability 0.001022361
theory models 0.001020784
phonetic variation 0.00101607
other features 0.001012457
speech recognition 0.0010022
intended pronunciation 9.91616E-4
marginal distribution 9.83273E-4
pronunciation variation 9.79544E-4
phonetic version 9.795E-4
right context 9.759739999999999E-4
joint probability 9.754500000000001E-4
real speech 9.722629999999999E-4
unique pronunciation 9.70534E-4
phonetic transcription 9.68653E-4
phonetic variability 9.685049999999999E-4
first set 9.62585E-4
