latent features 0.002488308
distributional features 0.0022690510000000002
additional features 0.002100828
tional features 0.002100804
latent word 0.0020391280000000003
traditional features 0.002015884
unigram features 0.001994687
cal features 0.00197435
features inw 0.001972593
distributional feature 0.0019530209999999999
feature weighting 0.001821687
tional feature 0.0017847739999999998
training data 0.0017809829999999999
features 0.00175164
feature vec 0.001733337
feature sets 0.001717045
factorization feature 0.0017165169999999999
word representations 0.00169931
discriminative feature 0.0016906859999999998
unigram feature 0.0016786569999999998
set data 0.001663023
semantic similarity 0.001650366
test data 0.001630183
word rep 0.001542034
feature 0.00143561
labeled data 0.001417614
rare words 0.001382653
individual words 0.001378865
continuous data 0.001331473
unseen words 0.00132415
dividual words 0.0013192969999999999
semantic analysis 0.0012762799999999999
support vector 0.001226829
vector machine 0.001226285
training set 0.001222946
sample vector 0.001201619
latent representation 0.001193164
latent space 0.001177644
negative matrix 0.001167071
narized vector 0.001156975
binarized vector 0.001156975
semantic sim 0.0011533609999999999
lexical similarity 0.001146464
semantic relatedness 0.001141758
latent representations 0.001133518
words 0.00109841
paraphrase classification 0.001085725
test set 0.001072146
classification algorithm 0.001065743
matrix factorization 0.001052423
textual matrix 0.001038256
matrix factor 0.001037598
original training 0.001037348
tic similarity 0.001037114
unsupervised matrix 0.001011433
latent dimension 0.001007843
supervised learning 0.00100454
paraphrase corpus 0.001002789
string similarity 0.001001518
classification problem 9.94503E-4
latent represen 9.93788E-4
latent basis 9.92277E-4
matrix factorizations 9.91924E-4
nonnegative matrix 9.91924E-4
similarity metrics 9.84614E-4
latent dimensionality 9.80221E-4
supervised classification 9.78403E-4
latent rep 9.76242E-4
first sentence 9.745870000000001E-4
distributional representation 9.73907E-4
different approach 9.71341E-4
sine similarity 9.67213E-4
distributional methods 9.63487E-4
latent repre 9.62681E-4
supervised information 9.51038E-4
classification framework 9.507340000000001E-4
vector 9.35855E-4
recent work 9.24939E-4
sentence pairs 9.16096E-4
paraphrase pairs 9.15207E-4
distributional representations 9.14261E-4
important information 8.968110000000001E-4
msrpc training 8.95216E-4
classification boundary 8.905460000000001E-4
transductive learning 8.79949E-4
future work 8.75427E-4
dependency relation 8.676280000000001E-4
previous work 8.66323E-4
distributional fea 8.6604E-4
classification next 8.65675E-4
relation recall 8.61732E-4
such techniques 8.602499999999999E-4
measure accuracy 8.55264E-4
prior work 8.52507E-4
value decomposition 8.46585E-4
linear svm 8.317050000000001E-4
different configurations 8.31617E-4
classifier parameter 8.302909999999999E-4
relation precision 8.29088E-4
complementary information 8.276500000000001E-4
