opinion word 0.00565212
opinion target 0.00522335
opinion words 0.005083173
opinion relations 0.004905363
candidate opinion 0.004865683
opinion relation 0.004769494
same opinion 0.004702874
several opinion 0.004676559
opinion mining 0.004644054
opinion relevance 0.004634247
opinion targets 0.004610419
specific opinion 0.004567657
similar opinion 0.004553765
opinion expressions 0.004511564
correct opinion 0.004490623
potential opinion 0.0044872900000000005
infrequent opinion 0.0044430450000000005
opinion 0.00417393
other word 0.0021537600000000002
word candidates 0.002054507
target extraction 0.0020335320000000002
word alignment 0.00201336
word position 0.0019571140000000002
different information 0.00195693
monolingual word 0.001833669
models word 0.0018309210000000001
word pairs 0.0018228200000000002
performance results 0.001792693
different methods 0.0017722319999999999
word positions 0.0017596640000000002
bilingual word 0.0017464940000000001
target candidate 0.001741173
alignment model 0.00162776
target candidates 0.001625737
feature extraction 0.0015945590000000002
other words 0.001584813
propagation algorithm 0.001568426
syntactic information 0.001565985
domain sentiment 0.0015544769999999999
alignment algorithm 0.00153646
different relation 0.001533534
crf model 0.001527138
alignment results 0.0015102319999999998
other methods 0.001509832
position information 0.0014978839999999999
mining algorithm 0.0014714139999999999
different language 0.001465479
translation model 0.001452282
statistical information 0.001440002
mutual information 0.0014374729999999999
same graph 0.001427758
extraction task 0.0014222450000000001
model this 0.001406984
discriminative model 0.001404642
frequency information 0.001402936
dependency relations 0.001402441
sentiment lexicon 0.0013961759999999998
full model 0.0013862890000000002
detailed information 0.001383932
extraction process 0.0013834860000000002
hmm model 0.0013729340000000001
alignment methods 0.0013694319999999999
experimental results 0.0013681499999999998
different ranking 0.001342839
previous methods 0.001307874
useful information 0.00130164
syntactic relations 0.0012784580000000001
propagation method 0.001274159
different domains 0.001263109
extraction procedure 0.0012600430000000002
extraction benefits 0.001252576
collocation extraction 0.001252576
comparable results 0.001247592
finial results 0.0012447679999999998
chinese language 0.0012442310000000002
same method 0.0012359670000000001
unsupervised methods 0.001233525
different products 0.001230838
large methods 0.0012298769999999999
review sentences 0.001227894
different sizes 0.001225442
different languages 0.00121861
different wtms 0.0012116359999999999
unified graph 0.001209647
traditional methods 0.001202883
bipartite graph 0.001201412
adjacent methods 0.001193834
sentence patterns 0.001183895
null words 0.001178852
irrelevant words 0.0011778700000000001
identification methods 0.001165929
overall performance 0.001157846
methods hotel 0.001157803
chinese reviews 0.001154372
chinese web 0.001144412
syntactic relation 0.001142589
methods camera 0.0011367999999999999
supervised methods 0.001133335
pervious methods 0.00111772
chinese texts 0.001112916
