task model 0.002817881
model topic 0.0027124460000000003
content model 0.002454108
previous model 0.00237946
model structure 0.0023278120000000003
target model 0.002274755
full model 0.002222953
specific model 0.002222081
latent model 0.002210184
model variables 0.002203889
model overview 0.002173424
tent model 0.0021707890000000002
model quality 0.0021638810000000003
our model 0.002157955
model posterior 0.002156368
jointcm model 0.002155546
background model 0.002152085
emission model 0.002139187
training data 0.001904183
model 0.00190258
sentence features 0.001892029
sentiment task 0.0018777709999999999
task features 0.001865521
data set 0.00182284
sentence topic 0.001751675
unlabeled data 0.001603999
sentiment corpus 0.0015566059999999999
sentiment analysis 0.001552767
other text 0.001539243
task parameters 0.00153806
parameters task 0.00153806
summarization task 0.0015172039999999999
analysis task 0.001505598
content features 0.001501748
other words 0.001488298
feature space 0.001472837
task performance 0.001472609
task label 0.001464052
feature functions 0.0014607
nlp data 0.001460109
annotated data 0.001454098
document text 0.001453322
task labels 0.001451012
entire data 0.001447972
text analysis 0.001446992
data sets 0.001446284
small data 0.001441403
joint training 0.001432689
amazon data 0.001430979
yelp data 0.001430793
sentence sequence 0.00142784
unannotated data 0.001427535
first task 0.001418611
sentence crf 0.001417029
standard sentiment 0.001411158
raw data 0.0014110120000000001
lexical features 0.0014107289999999999
same content 0.001393611
content models 0.0013923009999999999
task crf 0.001390521
sentence polarity 0.0013728310000000001
sentiment classification 0.001362663
content topic 0.0013613940000000001
regression models 0.001351428
sentiment rating 0.00133472
inference task 0.001326612
training corpus 0.001325399
sentiment ranking 0.001319741
same time 0.001314739
task labeled 0.001312986
topic sequence 0.001295897
additional features 0.001293784
task annotation 0.001288735
positive sentiment 0.001284461
sentiment ratings 0.001275986
ranking task 0.001272572
textual features 0.001255865
sentence top 0.001254593
labeled text 0.00125438
joint content 0.0012529540000000001
test words 0.001248944
original features 0.001238602
current sentence 0.00123442
unigram features 0.001228854
topic distribution 0.001224656
numeric sentiment 0.0012183839999999999
same number 0.001207459
different content 0.001207163
supervised sentiment 0.001205577
training time 0.001203919
original task 0.001203683
biguate word 0.001202165
word usage 0.001202165
level content 0.0012013290000000001
each sentence 0.0012012070000000001
task component 0.001198459
content information 0.001197989
summarization corpus 0.0011960389999999999
ith sentence 0.001195182
sentiment ranker 0.001194758
