language model 0.002357665
bayesian model 0.002250388
supervised model 0.002133159
tion model 0.002081928
generative model 0.002077805
joint model 0.002044035
model definition 0.002000363
model description 0.00199858
model corre 0.001991975
guage model 0.001990552
topic labels 0.0017629590000000001
model 0.00175695
latent topic 0.0017504250000000001
lexical distribution 0.001736395
topic proportions 0.0017105140000000002
draw topic 0.0016960130000000001
segmentation algorithm 0.00168506
discourse segmentation 0.001579024
topic 0.00146804
word repetition 0.001420998
linear segmentation 0.0013978130000000001
joint segmentation 0.0013865750000000001
current segmentation 0.0013840740000000001
segmentation stage 0.001365541
course segmentation 0.001330232
poor segmentation 0.00132464
segmentation granularity 0.00132419
same document 0.001313279
first segment 0.0012891270000000002
lexical cohesion 0.0012503800000000002
language models 0.0012389520000000002
joint distribution 0.001215058
new segment 0.001165442
marginal distribution 0.001161892
distribution gem 0.0011608880000000001
proposal distribution 0.001158747
inference algorithm 0.0011557120000000001
level segment 0.00115448
stop words 0.001110449
same problem 0.001106419
segmentation 0.00109949
inference time 0.001094078
lexical distributions 0.001088121
single segment 0.0010854039999999999
coherent segments 0.001079949
standard alignments 0.00104036
lexical distri 0.001037631
lexical cohe 0.001035013
erate words 0.00102905
same number 0.001028337
related segment 0.001022514
different parts 9.99023E-4
previous work 9.89589E-4
entire document 9.78188E-4
document body 9.66186E-4
segment border 9.65582E-4
work discourse 9.60445E-4
such documents 9.47111E-4
distribution 9.27973E-4
second language 9.2552E-4
same order 9.17944E-4
standard lda 9.12769E-4
first baseline 9.11283E-4
discourse structure 9.08903E-4
sentence extraction 9.0176E-4
pipeline results 8.93571E-4
different levels 8.822950000000001E-4
previous research 8.773660000000001E-4
sentence alignment 8.73065E-4
specific language 8.71621E-4
competitive results 8.702650000000001E-4
favorable results 8.63253E-4
random set 8.6065E-4
same metric 8.5573E-4
first step 8.50822E-4
automatic sentence 8.50515E-4
discourse analysis 8.45405E-4
standard metrics 8.43805E-4
ing problem 8.37313E-4
coherent topics 8.328299999999999E-4
ond language 8.311460000000001E-4
discourse boundaries 8.178009999999999E-4
first stage 8.13655E-4
unsupervised base 8.104450000000001E-4
probability min 8.07781E-4
new pair 8.04793E-4
words 8.04585E-4
complex set 8.03195E-4
exact inference 8.002840000000001E-4
many documents 7.9982E-4
cohesion modeling 7.91255E-4
pairwise sentence 7.892210000000001E-4
result table 7.819330000000001E-4
linear discourse 7.778570000000001E-4
segments 7.76933E-4
bayesian segmenta 7.703110000000001E-4
joint discourse 7.666190000000001E-4
pipeline approach 7.6295E-4
related work 7.619020000000001E-4
relevant sentence 7.60071E-4
