topic features 0.0032505800000000003
test data 0.002377217
lda features 0.002356176
baseline features 0.0023361939999999998
topic model 0.0023312840000000003
liwc features 0.002165882
topic models 0.001991844
topic distribution 0.00190136
topic modeling 0.0018406820000000002
features 0.00182518
emotion words 0.001794108
framework data 0.0017452420000000001
topic distributions 0.0017403380000000001
posterior topic 0.0017200190000000001
data col 0.0017023980000000001
tion words 0.001699534
total word 0.001698118
ordinal data 0.0016930830000000002
cognitive words 0.001687668
anger words 0.001671755
causation words 0.001663809
forward topic 0.0016579750000000001
word patterns 0.001622797
word count 0.001612666
test set 0.001550227
feature sets 0.001498785
alternative feature 0.001452685
words 0.00143586
topic 0.0014254
language complexity 0.0013297019999999999
human results 0.0013029679999999998
test sample 0.001256665
test document 0.001248924
feature 0.00122327
language use 0.001193426
other examples 0.001185166
test doc 0.0011387329999999998
other themes 0.001133532
significant time 0.001096975
other providers 0.001070839
language assessment 0.001062189
training items 0.001028384
patient language 0.00102571
human performance 0.001018231
human comparison 0.001005036
bdi training 0.001004778
significant concerns 9.70806E-4
results regression 9.54034E-4
social network 9.50364E-4
ing depression 9.403E-4
significant cost 9.35649E-4
different instruments 9.33548E-4
results table 9.22925E-4
human ratings 9.163859999999999E-4
previous work 9.10095E-4
different labels 9.09125E-4
bdi scores 9.06507E-4
model 9.05884E-4
screening results 9.05145E-4
ing documents 9.043389999999999E-4
score value 9.026959999999999E-4
linear regression 8.95699E-4
negative emotion 8.946779999999999E-4
different survey 8.936809999999999E-4
text analysis 8.890880000000001E-4
positive emotion 8.849349999999999E-4
such situations 8.55523E-4
related work 8.53664E-4
syntactic categories 8.24203E-4
svm regression 8.15075E-4
positive instances 8.048339999999999E-4
negative consequences 8.027139999999999E-4
future work 7.99678E-4
language 7.97277E-4
ing essays 7.958829999999999E-4
sophisticated models 7.94504E-4
social relationships 7.918129999999999E-4
class imbalance 7.834259999999999E-4
address class 7.834259999999999E-4
positive instance 7.796E-4
correlation improvement 7.74583E-4
negative emo 7.645729999999999E-4
lda cate 7.6239E-4
improvement result 7.60232E-4
lda theme 7.5975E-4
vanilla lda 7.5975E-4
ing rude 7.592059999999999E-4
average correlation 7.586579999999999E-4
vide baseline 7.58641E-4
screening measures 7.582590000000001E-4
high levels 7.47168E-4
first thing 7.44792E-4
average sentence 7.436249999999999E-4
social workers 7.425229999999999E-4
training 7.37766E-4
personality measure 7.30283E-4
sentence length 7.27254E-4
many people 7.25629E-4
supervised setting 7.25244E-4
shared task 7.22568E-4
