crf model 0.002717726
model accuracy 0.002616236
markov model 0.002476106
baseline model 0.002473616
model search 0.002450235
tagging model 0.002445429
prosody model 0.002416371
bidirectional model 0.002393719
discriminative model 0.002378411
class model 0.002353648
model order 0.0022908629999999998
enriched model 0.002289277
third model 0.002282809
line model 0.0022806149999999997
side model 0.002273995
model parame 0.002246492
ond model 0.002245643
pcfg model 0.002244282
product model 0.002244096
rectional model 0.002241369
model configurations 0.002241369
different models 0.002189838
other models 0.002069357
training data 0.002020146
model 0.00201681
different speech 0.001948788
different data 0.001871824
several models 0.00183783
baseline models 0.001710046
training set 0.001706592
other features 0.0017058870000000001
tagging models 0.001681859
word sequence 0.001675953
hmm models 0.0016746110000000001
prosodic feature 0.001671735
rescoring models 0.0016399750000000001
feature set 0.001635672
variable models 0.001630771
bidirectional models 0.001630149
discriminative models 0.001614841
ing word 0.0016139169999999999
word error 0.0016054619999999998
generative models 0.001564089
prosodic features 0.001547505
word values 0.001545852
enriched models 0.001525707
tive models 0.001520438
trained models 0.0015154040000000001
native models 0.00149452
right models 0.001491221
accurate models 0.001487978
models ranges 0.001485875
ging models 0.001481246
riched models 0.0014785800000000002
different pos 0.0014771659999999998
speech recognition 0.001463327
additional training 0.0014444360000000001
ing data 0.0013996030000000001
conversational speech 0.001381845
english speech 0.001374645
speech processing 0.001372984
feature functions 0.0013709879999999999
speech segmentation 0.001355513
sentence boundary 0.001331539
weight feature 0.0013268079999999999
combined training 0.001326357
primary training 0.0013229370000000002
speech transcriptions 0.0013071810000000001
prosodic information 0.001302583
speech transcript 0.001293691
prosody features 0.001289331
speech transcription 0.001288318
different sizes 0.001287545
other speaker 0.001278953
raw speech 0.0012760830000000001
feature templates 0.001260876
sentence boundaries 0.001259633
speech transcripts 0.0012577550000000002
sive feature 0.001257741
speech synthesis 0.001256774
speech category 0.001255926
models 0.00125324
data size 0.001250883
sentence segmentation 0.001248899
sational speech 0.0012448560000000001
spontaneous speech 0.0012383820000000001
speech tran 0.001237444
newswire data 0.0012304540000000002
segmented data 0.00121538
annotated sentence 0.00120712
automatic sentence 0.001206525
data segmentations 0.001201507
oraclebreak data 0.001196954
unpunctuated words 0.001190264
test set 0.001188939
same tag 0.001177501
large performance 0.00117215
side sentence 0.001162761
different seed 0.001162652
tagging performance 0.001161121
