different word 0.003303492
word error 0.0031974119999999997
other word 0.003103813
class word 0.00295065
word probability 0.002930912
function word 0.002891523
first word 0.002890508
final word 0.002847359
average word 0.002822431
full word 0.0028095029999999997
reference word 0.002808362
word level 0.002794819
individual word 0.002787207
word types 0.002779657
repeated word 0.0027685749999999997
word identity 0.002736097
word duration 0.002733311
word length 0.002729562
other features 0.002704283
lexical features 0.0027019039999999998
language model 0.002612003
features figure 0.002517409
other words 0.002515193
prosodic features 0.002467847
following features 0.002460607
various features 0.002417723
intensity features 0.002404835
individual features 0.002387677
numeric features 0.00236397
above features 0.0023637149999999997
class words 0.00236203
categorical features 0.002334826
disfluency features 0.002323378
meric features 0.002311923
single words 0.002253646
joint model 0.002244579
separate words 0.0022314920000000003
full words 0.002220883
reference words 0.0022197420000000002
correct words 0.002203837
single model 0.002202836
individual words 0.002198587
repeated words 0.002179955
last words 0.002179106
full model 0.002170073
nearby words 0.0021579850000000003
adjacent words 0.00214999
oov words 0.002144113
tial model 0.002127327
ual words 0.002126826
infrequent words 0.0021249570000000002
vidual words 0.002123813
erence words 0.002123813
disfluent words 0.002123813
feature values 0.002116483
initial model 0.002112487
features 0.00210678
regression model 0.002104872
combined model 0.002092106
sion model 0.002083996
model fit 0.002077406
model neg 0.002072181
model controls 0.002072181
tistical model 0.002072181
binary feature 0.002063465
numeric feature 0.00197765
feature weights 0.001935193
words 0.00191769
model 0.00186688
feature 0.00172046
data set 0.001602428
different error 0.0014882839999999999
other models 0.001374054
data analysis 0.001344992
data sets 0.001333809
tic data 0.0012702859999999998
entire data 0.0012305109999999999
exploratory data 0.001221574
contractions data 0.001214695
different baseline 0.001158364
different speakers 0.001099883
various models 0.001087494
different types 0.001070529
speech recognition 0.0010618469999999999
standard errors 0.0010546779999999999
baseline error 0.001052284
regression models 0.001014543
same effect 0.0010078650000000001
adaptation models 9.87862E-4
asr error 9.8738E-4
linear combination 9.5052E-4
error rates 9.49428E-4
error rate 9.45035E-4
possible values 9.417200000000001E-4
deletion error 9.36278E-4
other speaker 9.290819999999999E-4
syntactic class 9.09454E-4
error dra 8.95805E-4
other asr 8.93781E-4
average speech 8.85413E-4
