opinion word 0.00552388
opinion target 0.00512571
opinion words 0.00502027
opinion relations 0.0047188410000000005
opinion relation 0.004657346000000001
opinion mining 0.004528316
opinion targets 0.004521557
ing opinion 0.004474849
strong opinion 0.004425551
models opinion 0.00441664
opinion associations 0.004409617
opinion tar 0.004388633
potential opinion 0.004371628000000001
opinion association 0.004366256000000001
opinion rela 0.004350812
ith opinion 0.004350789
jth opinion 0.0043415450000000005
opinion 0.00408527
target word 0.00247905
word extraction 0.002331161
such word 0.002062047
topic model 0.002053892
word candidate 0.001999152
word candidates 0.0019952530000000002
sentiment words 0.00198009
target extraction 0.0019329909999999998
word alignment 0.0018960110000000002
words extraction 0.001827551
word number 0.0017833340000000001
ion word 0.001778082
monolingual word 0.0017654040000000001
extraction results 0.001764788
word preference 0.00176202
word pairs 0.001756455
word candi 0.0017465900000000001
word position 0.0017464660000000001
modified word 0.0017385540000000002
word alignments 0.0017088440000000002
extraction performance 0.0017061440000000002
word preferences 0.0016879640000000001
word vjt 0.001686997
word vio 0.001686997
sider word 0.001686997
ployed word 0.001686997
alignment model 0.0016238110000000002
target candidate 0.001600982
target candidates 0.001597083
other words 0.001555139
lda model 0.001478942
unified model 0.0014571520000000002
different relations 0.001450901
model seman 0.001417211
sentiment lexicon 0.001413839
different relation 0.0013894060000000001
same words 0.001386257
different candidate 0.001377872
target list 0.0013546259999999998
target candi 0.00134842
alignment results 0.001329638
targets extraction 0.001328838
implicit sentiment 0.001328096
ranking algorithm 0.001303083
sentiment associ 0.001294029
extraction precision 0.001293004
other methods 0.00128511
dataset methods 0.001282811
ing extraction 0.00128213
ion words 0.001274472
ing graph 0.001270842
graph ranking 0.00126804
ranking results 0.001259014
such relations 0.001257008
level extraction 0.001254861
extraction step 0.001246317
extraction process 0.001243691
tional extraction 0.00123234
walk algorithm 0.001231937
preferred words 0.001227587
topic distribution 0.001226637
homogeneous words 0.001223942
previous methods 0.001218449
extraction strategy 0.001203604
experimental results 0.0011949460000000001
identification results 0.0011944759999999999
walking algorithm 0.001190844
useless words 0.001185344
global graph 0.001182366
other candidates 0.001176782
heterogeneous graph 0.001174751
model 0.00116641
algorithm ongtt 0.001166047
preference information 0.001165539
results figure 0.001164716
review dataset 0.001164086
form extraction 0.00116189
set domain 0.001157087
right graph 0.001152375
topic modeling 0.0011516389999999999
occurrence information 0.0011444559999999999
collective extraction 0.001142357
