word feature 0.002236907
labeled data 0.001938332
similar data 0.001931403
training data 0.00189684
different features 0.001812808
unlabeled data 0.001812011
feature vector 0.001805323
match feature 0.001661751
feature extraction 0.00165647
test data 0.001648775
data point 0.001644642
matching feature 0.001597974
semantic features 0.0015939869999999998
feature values 0.001593419
question word 0.001590635
data points 0.001588367
data increase 0.001568693
representative data 0.001564148
component feature 0.001546127
entailment features 0.001543616
testing data 0.001541156
specific data 0.0015339589999999999
feature extrac 0.001525293
data samples 0.001523332
unseen data 0.0015137759999999999
actual data 0.001506648
different sentence 0.001460853
match features 0.001426141
syntactic features 0.001372095
matching features 0.001362364
alignment features 0.001322489
feature 0.00130795
same label 0.001302009
lexsem features 0.00129218
features rep 0.001287653
labeled training 0.001265632
answer sentence 0.001262409
model mrr 0.001254589
sentence semantic 0.0012420320000000001
sentence form 0.001236281
overall model 0.001232653
classifier model 0.001219454
tive word 0.001216666
training set 0.0012044479999999999
question answer 0.001203702
sentence structure 0.001198103
new graph 0.001196757
question form 0.0011775739999999998
ssl method 0.00117545
word disambiguation 0.001169959
entailment information 0.001156779
labeled vector 0.001150935
lesk word 0.001147619
word disambigua 0.001144973
word indicators 0.001144973
unlabeled training 0.001139311
other ssl 0.0011336060000000001
first sentence 0.001132598
graph summary 0.001120848
summary graph 0.001120848
svm method 0.001119375
task performance 0.0011176699999999999
labeled dataset 0.001106699
classification problem 0.001106162
different test 0.001104473
graph summarization 0.00108564
features 0.00107234
training dataset 0.001065207
different sets 0.001059778
similarity function 0.0010590209999999998
summary information 0.001058881
algorithm let 0.0010582719999999999
learning methods 0.00105387
summarization algorithm 0.001036867
information extraction 0.001034023
performance measures 0.0010336249999999998
svm models 0.0010298809999999999
hybrid graph 0.001028633
paired sentence 0.0010284460000000001
labeled summary 0.00102694
answer text 0.001022724
such approach 0.001021947
sentence pair 0.00102161
sentence analysis 0.0010214220000000001
candidate sentence 0.001018712
different structures 0.0010171310000000001
different sample 0.001010258
sentence datasets 0.001009029
sentence pairs 0.001002314
web classification 9.998210000000001E-4
unlabeled entailment 9.98517E-4
graph laplacian 9.982749999999999E-4
entailment models 9.98269E-4
ferent sentence 9.97482E-4
sentence structures 9.97048E-4
performance increase 9.93588E-4
special graph 9.92794E-4
relation information 9.917419999999999E-4
mrr performance 9.90603E-4
feasible graph 9.86457E-4
