sentiment tree 0.0031329279999999997
sentiment task 0.003075915
sentiment values 0.003053755
sentiment analysis 0.0030205709999999997
sentiment information 0.0030156059999999997
compositional sentiment 0.00300434
sentiment distribution 0.002978825
sentiment value 0.002974488
sentiment score 0.002932574
prior sentiment 0.0028878439999999997
gold sentiment 0.00288758
sentiment label 0.002867602
sentiment treebank 0.0028645719999999997
sentiment scores 0.0028302559999999997
sentiment class 0.002820276
automatic sentiment 0.0028150049999999998
sentiment categories 0.0028072329999999997
stanford sentiment 0.002801981
sign sentiment 0.002798455
sentiment compositions 0.002789924
sentiment annotation 0.0027812289999999996
sentiment polarities 0.0027685939999999997
notated sentiment 0.002765878
sentiment 0.00256211
model parameters 0.002129752
composition model 0.002094681
rntn model 0.002085992
tion model 0.002074616
form model 0.002042764
pstn model 0.0020407010000000002
typical model 0.002028337
model 0.00181623
different models 0.0015796999999999999
word good 0.001563518
negation word 0.001538647
other models 0.0014895870000000001
training data 0.0014787049999999999
word dictionary 0.0014024950000000001
word senses 0.001395247
other data 0.00135536
negation models 0.001264793
negation words 0.001206901
error vector 0.001193156
ing vector 0.001180877
evaluation data 0.001115883
test data 0.001114502
node vector 0.001087037
such assumptions 0.001067317
data point 0.001056786
vector representations 0.001052812
annotated data 0.001048624
external data 0.001047359
such mod 0.001040027
such informa 0.001037257
such matching 0.001020225
semantic information 0.001016232
different negators 0.001012655
negative phrase 9.96528E-4
other approaches 9.81009E-4
negative phrases 9.70204E-4
tree nodes 9.62631E-4
external training 9.601659999999999E-4
different complex 9.52056E-4
ing error 9.44379E-4
vector represen 9.37839E-4
different hypothesis 9.239329999999999E-4
models 9.17176E-4
different constants 9.057259999999999E-4
positive value 8.89018E-4
parse tree 8.878689999999999E-4
different ways 8.87618E-4
different depths 8.72999E-4
negative poles 8.70338E-4
different purposes 8.675409999999999E-4
different granularities 8.675409999999999E-4
prediction error 8.627190000000001E-4
words 8.59284E-4
syntactic information 8.55022E-4
tive function 8.522709999999999E-4
analysis system 8.43639E-4
semantic composition 8.41187E-4
neural tensor 8.3541E-4
system performance 8.30757E-4
modeling negators 8.17581E-4
negation modeling 8.15067E-4
modeling negation 8.15067E-4
neural network 8.13282E-4
random baseline 8.0762E-4
other hand 8.07268E-4
learning inference 8.04557E-4
same negators 8.015310000000001E-4
positive sen 7.96242E-4
tree car 7.93784E-4
performance table 7.93221E-4
neural net 7.905449999999999E-4
recursive neu 7.866689999999999E-4
semantic infor 7.84057E-4
underlying function 7.83889E-4
sign function 7.81603E-4
semantic roles 7.797349999999999E-4
