language model 0.00371257
new model 0.002776531
other sentence 0.002656427
state model 0.002610231
acoustic model 0.002577195
model performance 0.0025743009999999998
model scores 0.002557832
conversational model 0.002529535
above model 0.002525817
level model 0.0025038689999999997
complete model 0.002484458
third model 0.002478581
general model 0.0024494399999999998
model estimates 0.002441336
model parameters 0.002435821
precise model 0.002425526
language models 0.002397603
sentence test 0.0023749549999999998
first sentence 0.002342813
current sentence 0.00230857
training language 0.002228766
other words 0.002213657
previous sentence 0.002203883
model 0.0021966
overall sentence 0.002189549
word error 0.002176464
sentence boundary 0.002175748
sentence parts 0.002160709
full sentence 0.0021434369999999998
sentence boundaries 0.002124726
second sentence 0.002123913
grammatical sentence 0.002115096
same word 0.002100828
pivot sentence 0.0020978719999999998
new language 0.002095901
sentence units 0.002086178
sentence length 0.002083643
life sentence 0.002070835
preceding sentence 0.002058956
sentence connectors 0.0020553109999999998
different language 0.002032891
next word 0.001990533
new words 0.001964111
statistical language 0.001949412
recognition word 0.001948334
language modeling 0.00193408
single word 0.0019138990000000002
likely word 0.001912357
structure language 0.001891972
word unit 0.001855714
complex language 0.001837111
regular word 0.0018285
sentence 0.00182695
rank word 0.001826331
level language 0.001823239
function words 0.001818835
order language 0.0018110539999999999
trigram language 0.001798062
bigram language 0.001791088
sentences sentences 0.00177675
separate language 0.001775749
summer language 0.0017631069999999999
hurt language 0.0017483709999999999
targeted language 0.001743657
words table 0.0017351620000000002
common words 0.001719817
training data 0.001709548
total words 0.001672299
text data 0.001668869
actual words 0.001652721
few words 0.0016461730000000001
frequent words 0.001625313
partial words 0.001615128
words seriesl 0.001613057
filler words 0.001613057
training sentences 0.001601171
test data 0.001544757
training corpus 0.001533717
language 0.00151597
new models 0.001461564
data analysis 0.001407651
words 0.00138418
speech information 0.0013645369999999999
overall data 0.001359351
linguistic data 0.001353513
new information 0.001304368
testing data 0.001284615
data segmentations 0.0012651070000000001
data sets 0.001264986
data stream 0.001263865
data problems 0.001263737
test training 0.001260801
structure models 0.001257635
sparse data 0.0012269820000000002
data consortium 0.00122626
conversational models 0.001214568
complex sentences 0.0012095159999999999
baseline training 0.001208087
full sentences 0.001204862
switchboard corpus 0.0012010669999999999
