word vector 0.0037724000000000004
word vectors 0.003207659
word vec 0.002731454
language model 0.002714242
learning model 0.002498095
baseline model 0.0023632090000000002
simple model 0.00235865
model architecture 0.002284845
model variations 0.002224615
model variants 0.002206448
model 0.00198882
vector space 0.001751711
close words 0.001601886
dimensional vector 0.0015649890000000001
language models 0.001556885
semantic features 0.0015378359999999999
vector representa 0.001535071
training data 0.0014702
words 0.0013978
training set 0.0013774850000000001
learning models 0.001340738
sentiment analysis 0.001324046
vector 0.00132002
weight vectors 0.0012412389999999999
supervised training 0.001208378
neural language 0.00119682
stanford sentiment 0.001157962
sentiment treebank 0.00114217
semantic parsing 0.001102859
hidden layer 0.001063395
cnn models 0.0010481289999999999
multiple features 0.0010360109999999999
language process 0.001026531
natural language 0.001009711
other parameters 9.91835E-4
softmax layer 9.72006E-4
excellent results 9.71411E-4
trained vectors 9.699439999999999E-4
static vectors 9.624659999999999E-4
new feature 9.585939999999999E-4
local features 9.47607E-4
remarkable results 9.44299E-4
sentiment 9.37785E-4
training 9.3745E-4
movie reviews 9.03412E-4
function 8.98488E-4
set size 8.87926E-4
sentence length 8.85519E-4
convolutional layer 8.81859E-4
penultimate layer 8.74297E-4
original task 8.71368E-4
average sentence 8.505590000000001E-4
dataset size 8.48757E-4
test set 8.472880000000001E-4
models 8.31463E-4
linear units 8.296E-4
example sentence 8.24136E-4
variable sentence 8.12201E-4
numeric information 8.10088E-4
feature map 8.088749999999999E-4
sentence modeling 8.047950000000001E-4
sentence lengths 7.871040000000001E-4
other param 7.86099E-4
deep learning 7.72784E-4
lar method 7.60897E-4
vectors 7.55279E-4
classification tasks 7.51908E-4
test time 7.43561E-4
learning meth 7.40172E-4
results 7.3746E-4
dense representations 7.26084E-4
language 7.25422E-4
neural networks 7.22662E-4
target classes 7.223819999999999E-4
feature extractors 7.157649999999999E-4
tion tasks 7.123780000000001E-4
vocabulary size 7.11768E-4
portant feature 7.103729999999999E-4
dev set 6.93859E-4
question dataset 6.92263E-4
maximum value 6.87141E-4
question types 6.78032E-4
nlp tasks 6.744100000000001E-4
features 6.72986E-4
layer 6.66087E-4
present work 6.463490000000001E-4
output figure 6.40844E-4
standard dev 6.37798E-4
mpqa dataset 6.36057E-4
subjectivity dataset 6.35762E-4
several variants 6.34825E-4
multiple channels 6.33517E-4
question classification 6.32533E-4
customer reviews 6.31017E-4
hidden units 6.281170000000001E-4
multiple filter 6.20543E-4
information 6.04358E-4
search query 6.039190000000001E-4
further improvements 6.03804E-4
softmax output 5.99857E-4
