topic segmentation 0.00311653
segmentation algorithm 0.002702256
segmentation results 0.002476339
new segmentation 0.002373348
segmentation method 0.002269888
topic segments 0.0022318439999999998
segmentation methods 0.002218148
probability segmentation 0.002197856
reference segmentation 0.002157867
segmentation accuracy 0.002119448
segmentation step 0.002109155
accurate segmentation 0.0020496250000000002
probable segmentation 0.0020487780000000002
different topic 0.002029378
text data 0.001961926
segment length 0.001892464
segmentation 0.00179093
lexical distribution 0.001775413
segment models 0.00171575
lexical cohesion 0.0017022299999999999
same segment 0.0016987130000000001
current topic 0.001696768
segments length 0.001689538
topic changes 0.0016808559999999999
topic segmenta 0.0016724729999999998
topic shifts 0.0016402359999999998
respective topic 0.001638741
smooth topic 0.0016318799999999998
natural topic 0.0016310559999999999
corpus figure 0.001610122
topic shift 0.001606684
tic topic 0.001597486
abrupt topic 0.0015849089999999998
segment size 0.001549747
performance algorithm 0.001537828
news corpus 0.001480783
other words 0.001477711
similarity measure 0.0014637079999999998
data set 0.001461055
various data 0.001458894
maximum segment 0.001447014
minimum segment 0.0014387290000000001
lexical disruption 0.001418445
lexical chains 0.001404935
transcripts corpus 0.001394891
few word 0.001387377
corpus statistics 0.001386599
same length 0.001372837
development data 0.0013693590000000001
segment fron 0.001367414
variable segments 0.001366787
lexical cohe 0.001366034
cosine similarity 0.00136266
lexical reoccurrence 0.001361693
lexical disruptions 0.001361693
textbook corpus 0.001360963
data sets 0.001354221
word error 0.0013540509999999998
small segments 0.0013418619999999999
word repetitions 0.001331016
topic 0.0013256
word vectors 0.0013251909999999999
results performance 0.001311911
spective word 0.001306774
long segments 0.0013051879999999999
word reoccur 0.0012982979999999998
prior model 0.001297997
different values 0.001296823
textual corpus 0.001292544
artificial data 0.001289832
transcript data 0.001284782
textual data 0.001283522
simple model 0.001281562
adjacent segments 0.001278132
challenging corpus 0.001276886
medical text 0.001276796
brown corpus 0.001275414
corpus auto 0.001267436
corresponding algorithm 0.0012645199999999999
similarity profile 0.001261138
choi data 0.001261076
tual data 0.001257971
short segments 0.0012568829999999999
data thanks 0.001256512
noisy data 0.001256512
coherent segments 0.001243981
thematic segments 0.001229123
efficient algorithm 0.0012192029999999999
incoming segments 0.0012165869999999999
art algorithm 0.0012162569999999999
tation algorithm 0.001204402
different topics 0.001198292
markov model 0.001193187
ent segments 0.001183034
tween segments 0.0011829739999999998
inal algorithm 0.001171745
algorithm cast 0.0011704699999999998
secutive segments 0.001166767
neighboring segments 0.001164572
edge length 0.0011509340000000002
