features words 0.00401002
top features 0.002456794
boolean features 0.002431368
weight features 0.002429465
word frequency 0.002414134
ditional features 0.002387573
word distribution 0.002292742
word use 0.0022823369999999997
many words 0.002207917
word types 0.002204629
features 0.00218919
word frequencies 0.002147571
word fre 0.00213987
certain words 0.002125525
word recurrence 0.002125268
available words 0.002077673
vocabulary words 0.002074895
feature set 0.00204476
stigma words 0.002030905
taboo words 0.0020187070000000002
banner words 0.0020187070000000002
lary words 0.0020187070000000002
words 0.00182083
feature weighting 0.001772346
feature quality 0.001766365
tfidf feature 0.001723617
feature redundancy 0.001721509
full feature 0.001717931
training data 0.001659343
test data 0.001577703
feature 0.00150366
testing data 0.001293642
data parameter 0.001290891
pba data 0.001272058
corpora data 0.001249952
data nbest 0.001245477
good model 0.001243569
same classification 0.0011754299999999999
full model 0.0011651769999999999
urn model 0.001147964
other perspective 0.0010971050000000001
discriminative models 0.0010648979999999999
mutual information 0.0010273090000000001
perspective classification 0.001022678
text clas 0.001018368
bayes models 0.001015815
classification accuracy 0.001004933
different issues 9.997299999999999E-4
text segmentation 9.95813E-4
such tasks 9.88521E-4
generative models 9.877269999999999E-4
text categorization 9.85282E-4
topic classification 9.78525E-4
textual information 9.76346E-4
text cate 9.75827E-4
text catego 9.74672E-4
text categoriza 9.73671E-4
model 9.50906E-4
event models 9.489009999999999E-4
models rou 9.489009999999999E-4
sider models 9.489009999999999E-4
different websites 9.47505E-4
classification task 9.47284E-4
set performance 9.42268E-4
linear svm 9.36079E-4
other hand 9.064790000000001E-4
other vocabu 8.92669E-4
lexical consolida 8.893239999999999E-4
classification accuracies 8.73787E-4
penalty corpus 8.73645E-4
such perspectives 8.612069999999999E-4
results table 8.366539999999999E-4
party classification 8.33716E-4
same types 8.31054E-4
opinion classification 8.268139999999999E-4
curate classification 8.17285E-4
timent classification 8.17285E-4
same speaker 8.14747E-4
frequency distributions 8.06807E-4
information 7.69858E-4
entire corpus 7.61086E-4
bitterlemons corpus 7.56963E-4
full set 7.55371E-4
same fold 7.53668E-4
models 7.51808E-4
large vocabulary 7.18639E-4
future work 7.17541E-4
low performance 7.068529999999999E-4
analysis studies 6.99386E-4
frequency composition 6.946529999999999E-4
many times 6.931870000000001E-4
discourse analysis 6.88322E-4
ment frequency 6.86925E-4
low power 6.85786E-4
svm bool 6.85253E-4
frequency distri 6.84341E-4
vocabulary selection 6.7932E-4
small number 6.78258E-4
multiple times 6.67864E-4
large discrepancy 6.638519999999999E-4
