other features 0.001949857
conversational features 0.0018084759999999998
speech data 0.0017349660000000001
same speaker 0.001679228
model performance 0.001618023
different conversations 0.001608543
model scores 0.001584608
mention features 0.001569712
same conversation 0.001556979
gap features 0.0015518209999999999
utterance topic 0.001518205
current model 0.0015135329999999998
feature set 0.001459953
our model 0.0014352739999999998
model our 0.0014352739999999998
speech corpus 0.001434166
annotation system 0.001418448
annotators model 0.0013976289999999998
generative model 0.001385907
model best 0.0013855389999999999
large conversation 0.001343
test data 0.001338801
previous utterance 0.00132803
features 0.00131508
other conversations 0.001312466
utterance con 0.001311757
average speaker 0.00129233
such corpus 0.0012748249999999998
specific utterance 0.001255979
particular speaker 0.00123698
same class 0.001236898
many conversations 0.001236242
new conversation 0.0012359390000000001
particular utterance 0.001228905
feature functions 0.001224734
other words 0.0012187600000000002
useful feature 0.001205886
single conversation 0.001202196
annotated speech 0.001190354
each utterance 0.001181413
key feature 0.001174968
conversational system 0.001174114
annotated data 0.0011670040000000001
own conversation 0.001161315
model 0.00115819
repetition feature 0.001156445
different numbers 0.0011506609999999999
sign utterance 0.001148906
utterance contents 0.001148906
same mean 0.0011207069999999999
development data 0.001113632
new conversations 0.0010968269999999999
separate conversation 0.001096044
ment data 0.001085555
annotation scheme 0.0010827060000000001
room data 0.0010769970000000001
synthetic data 0.001073325
typical conversation 0.001070847
human annotations 0.0010692969999999999
multiple conversations 0.0010646359999999999
chat corpus 0.0010487539999999998
social chat 0.001047704
parent conversation 0.001046538
various conversations 0.001044884
source annotation 0.001044157
ongoing conversation 0.001039833
ferent conversation 0.00103609
spoken conversation 0.001035374
additional annotation 0.001028781
large news 0.001025139
discourse task 0.001022524
future annotation 0.001018108
same name 0.001016808
annotation our 0.001014814
useful annotation 0.001012483
test set 0.001011813
tional system 9.89551E-4
annotation protocol 9.89323E-4
same conver 9.766479999999999E-4
new corpus 9.74146E-4
annotation schemes 9.67796E-4
related utterances 9.65395E-4
reliable annotation 9.62396E-4
artificial annotation 9.55504E-4
distinct conversations 9.52521E-4
cue words 9.50182E-4
information extraction 9.46814E-4
dialogue analysis 9.4483E-4
test performance 9.428259999999999E-4
few conversations 9.404509999999999E-4
speaker 9.3905E-4
reserve system 9.34217E-4
previous work 9.33769E-4
baseline performance 9.31607E-4
feature 9.31133E-4
utterance 9.30975E-4
min conversations 9.23501E-4
tween utterances 9.22321E-4
disentanglement system 9.11653E-4
word rep 9.08982E-4
