semantic score 0.001687422
lexical words 0.0015828169999999998
semantic information 0.001493403
different dialogue 0.0014792529999999998
sentence tree 0.0014762080000000001
data set 0.001457595
lexical word 0.0014504610000000001
lexical score 0.001361588
dialogue system 0.0013579920000000001
recognition model 0.001348452
semantic classes 0.001312401
tion model 0.001258101
guage model 0.001223085
markov model 0.001211604
dialogue act 0.001198559
feature set 0.001195274
model assump 0.001187836
sentence trees 0.0011706590000000001
rule set 0.001161443
node set 0.00115912
speech data 0.001106577
rule score 0.001091113
word accuracy 0.0010902680000000001
word types 0.001089335
input word 0.00108371
score function 0.0010835419999999998
asr score 0.001067381
text data 0.001056111
optimal dialogue 0.001048728
system evaluation 0.0010403040000000001
output data 0.00103618
dialogue corpus 0.0010272039999999999
ing algorithm 0.001024474
word sequences 0.001020632
edge set 0.001020318
symbol set 0.00101262
second set 0.001012059
dialogue acts 0.00100902
dialogue management 0.001008922
dialogue history 0.001006902
training corpus 9.989579999999999E-4
average word 9.89776E-4
speech features 9.8492E-4
rule probability 9.77087E-4
mation data 9.74447E-4
different sources 9.71801E-4
probability function 9.69516E-4
model 9.63348E-4
third set 9.57424E-4
sentence symbol 9.57423E-4
partial sentence 9.5513E-4
fourth set 9.53259E-4
conditional probability 9.27261E-4
derivation rules 9.26755E-4
exemplar sentence 9.265650000000001E-4
sentence clustering 9.17885E-4
first type 9.14872E-4
entity score 9.01038E-4
language 8.95789E-4
tial sentence 8.908610000000001E-4
original sentence 8.90845E-4
overall system 8.894930000000001E-4
put sentence 8.8907E-4
history score 8.871899999999999E-4
vector representation 8.78656E-4
rule matrix 8.769299999999999E-4
tory score 8.75259E-4
derivation rule 8.72041E-4
words 8.70561E-4
column vector 8.62609E-4
ing results 8.538300000000001E-4
act sequence 8.43548E-4
acceleration features 8.41946E-4
system benefits 8.14885E-4
validation method 8.08661E-4
ing text 8.04679E-4
asr recognition 8.03153E-4
acoustic models 8.0004E-4
particle models 7.9541E-4
rule representation 7.93992E-4
clustering algorithm 7.91393E-4
tion rules 7.91248E-4
asr results 7.85378E-4
dialogue 7.69044E-4
first term 7.62319E-4
nition probability 7.61222E-4
vector rep 7.54226E-4
speech recognition 7.53748E-4
first row 7.46386E-4
training 7.40798E-4
tion rule 7.36534E-4
tence trees 7.32386E-4
act detection 7.30223E-4
tags information 7.30185E-4
new utterance 7.2668E-4
asr scores 7.253159999999999E-4
speech utterance 7.2068E-4
current act 7.19648E-4
asr output 7.16296E-4
input speech 7.141490000000001E-4
