coherence model 0.0036
same model 0.003277336
grid model 0.003060763
speaker model 0.002973338
baseline model 0.002966377
good model 0.002950583
single model 0.002830003
model combination 0.00279366
topical model 0.00279343
ibm model 0.002781099
time model 0.002772722
comparison model 0.002769935
model score 0.002756094
linear model 0.002753714
gle model 0.002729612
combined model 0.002726052
translation model 0.0027254189999999998
supervised model 0.002722211
degenerate model 0.002711318
weak model 0.002711318
parison model 0.002711318
coherence models 0.00256535
model 0.00247759
different models 0.002269633
lexical models 0.002174218
other models 0.002157305
syntactic models 0.002080159
training data 0.002067545
grid models 0.002026113
such models 0.001988481
disentanglement models 0.001882398
other coherence 0.001836775
individual models 0.001771593
topic models 0.001758428
discourse coherence 0.001750743
generative models 0.001733337
ence models 0.001717082
popular models 0.001708873
lexicalized models 0.001708843
herence models 0.001697289
combined models 0.001691402
mixture models 0.001679889
ing data 0.001571029
local coherence 0.001558987
chat data 0.001539065
sentence task 0.001531013
swbd data 0.0015141970000000001
synthetic data 0.001459543
entity grid 0.001451611
models 0.00144294
fisher data 0.001420417
lexical features 0.001413489
coherence transfer 0.0013969909999999998
global coherence 0.0013958439999999998
coherence mod 0.0013913909999999999
irc data 0.001384431
extra data 0.00137563
tanglement data 0.001374956
cal coherence 0.0013726049999999998
sophisticated coherence 0.0013700329999999999
training algorithm 0.001360961
tomodel coherence 0.001356234
same set 0.001351746
previous sentence 0.00133717
speaker information 0.001334746
good performance 0.001333482
other words 0.001319943
syntactic features 0.00131943
role information 0.0012985919999999999
same speaker 0.001295494
standard entity 0.001286508
different domain 0.001283348
training sets 0.001274632
training example 0.001223897
single entity 0.001220851
prior information 0.001202544
performance increase 0.001198596
training transcripts 0.001194591
second feature 0.001187383
irc training 0.001186836
topical entity 0.001184278
disentanglement task 0.001172338
mixture training 0.001171924
training regime 0.001169111
generative entity 0.001158835
single sentence 0.0011505460000000001
lexical parameters 0.001143432
different topic 0.001142181
poor performance 0.001140853
previous sentences 0.001137553
rst feature 0.001130949
coherence 0.00112241
same topic 0.001115234
salience feature 0.001112663
history information 0.001109495
combined performance 0.001108951
news text 0.0011064
entity grids 0.001103837
same conversation 0.001096923
syntactic role 0.0010968129999999999
