model training 0.0026844300000000002
corpus word 0.002494759
state probability 0.002491932
state distribution 0.002489197
markov model 0.0023367130000000002
state sequence 0.002209225
state transition 0.00214487
automatic model 0.002123067
initial state 0.002085119
word pair 0.002017236
state pair 0.002008066
word length 0.001996678
state values 0.00198479
word segmentation 0.001958702
target state 0.001955723
optimal state 0.001936984
word pairs 0.001933377
correct word 0.001902258
likely state 0.001874377
function words 0.001864002
probability distribution 0.001842229
word format 0.001832313
model 0.00182685
training corpus 0.001773719
training data 0.00174554
state 0.00156945
final words 0.0015417500000000002
text corpus 0.001518085
same training 0.0015120490000000001
derivative words 0.001509361
training set 0.001484681
observation probability 0.0014133190000000001
syntactic features 0.001357372
human speech 0.0013390289999999998
probability variable 0.001330392
probability distributions 0.0012953119999999999
parameter set 0.001293696
words 0.00126044
prosodic features 0.0012419
symbol probability 0.001228325
phrase sequence 0.001208677
phrase structure 0.001203784
pos sequence 0.001189824
posteriori probability 0.001177172
chinese speech 0.00117052
syntactic structure 0.0011595210000000002
observation sequence 0.001130612
hmm path 0.001127011
phonetic features 0.001109
speech synthesis 0.00107104
high accuracy 0.00107098
speech generation 0.001055822
prosodic structure 0.001044049
mandarin speech 0.001030765
statistical method 0.001026874
evaluation method 0.0010240639999999999
parameter estimation 0.001018022
estimation parameter 0.001018022
speech intonation 0.001012612
chinese text 0.0010121050000000001
contextual information 0.00101187
input information 9.90147E-4
length phrase 9.8696E-4
phrase length 9.8696E-4
hidden markov 9.82911E-4
prosodic phrase 9.78069E-4
viterbi algorithm 9.77879E-4
length sentence 9.75345E-4
sentence length 9.75345E-4
chinese sentence 9.67446E-4
test set 9.6049E-4
parameter adjustment 9.56735E-4
rich information 9.47068E-4
complete parameter 9.37547E-4
different solution 9.26867E-4
probability 9.22482E-4
distribution 9.19747E-4
pos tagging 9.17481E-4
corpus 9.16139E-4
same usage 9.10002E-4
first stage 9.09017E-4
sequential states 9.052800000000001E-4
real text 8.99242E-4
finite set 8.9391E-4
approach base 8.918649999999999E-4
phrase sequences 8.91142E-4
models 8.8665E-4
language processing 8.83047E-4
natural language 8.793130000000001E-4
distinct observation 8.761350000000001E-4
prosodic position 8.7285E-4
mandarin text 8.7235E-4
input sentence 8.700539999999999E-4
start node 8.599219999999999E-4
training 8.5758E-4
chinese parsing 8.47829E-4
present approach 8.4285E-4
average phrase 8.42328E-4
effective approach 8.402469999999999E-4
alternative approach 8.376219999999999E-4
