word features 0.00277554
answer vector 0.002590813
answer pairs 0.002533811
question data 0.002487139
answer retrieval 0.002469226
candidate answer 0.002446736
first answer 0.002416522
forum answer 0.002362184
answer ranking 0.00230018
answer vectors 0.002296734
answer detection 0.002269587
visible answer 0.002255173
answer post 0.0022547640000000002
best answer 0.002225301
question vector 0.002222823
nearest answer 0.002216252
word pairs 0.002078011
language model 0.002007365
answer 0.00199445
factoid question 0.0019449839999999999
question answering 0.0019382099999999999
function word 0.00193805
question vectors 0.001928744
word distribution 0.00192206
such features 0.0019196270000000001
question detection 0.0019015969999999999
question con 0.001901557
word selection 0.001899066
rbm model 0.001873064
question thread 0.0018659179999999998
word fea 0.001862253
question threads 0.0018488579999999999
reconstructed question 0.0018452569999999999
only word 0.001841683
translation model 0.001834895
dbn model 0.001829908
corresponding word 0.001827172
low word 0.0018217939999999998
training data 0.001814568
word informa 0.001809869
graphical model 0.001793048
word frequency 0.001789324
word distributions 0.0017814999999999999
word fre 0.001763854
word occurrence 0.001758778
kth word 0.0017578539999999999
semantic information 0.0017523460000000001
textual features 0.0016528390000000001
hidden features 0.0016525020000000001
question 0.00162646
training set 0.001547505
semantic knowledge 0.001541868
model 0.00153282
training method 0.001529889
feature vector 0.001513655
position features 0.001497933
structural features 0.00146058
content words 0.0013694190000000002
similar questions 0.001357758
semantic modeling 0.001343684
cqa data 0.001338869
hidden feature 0.001332904
semantic relevance 0.001329238
semantic link 0.001323418
candidate answers 0.001315602
training algorithm 0.001312232
semantic rel 0.00130089
semantic relationship 0.001298516
semantic rele 0.0012887900000000002
factoid questions 0.001262809
entire training 0.0012435789999999999
features 0.00123689
forum data 0.001228413
function words 0.001221731
feature vectors 0.001219576
binary feature 0.001212541
opinion questions 0.001209802
corresponding feature 0.001205814
date answers 0.0011977490000000001
visible feature 0.001178015
annotated questions 0.001177884
real answers 0.0011748610000000001
input data 0.0011738360000000002
words figure 0.001173286
training object 0.001173155
manner questions 0.001171731
causation questions 0.001163852
feature sparsity 0.0011409950000000001
den feature 0.001139695
zhidao data 0.0011252760000000001
models performance 0.0011172299999999999
homogenous data 0.001108224
different sources 0.0011008419999999999
ing set 0.001087737
liver answers 0.001082847
rum data 0.001081472
rest words 0.001077203
synonymous words 0.001077054
other research 0.001068926
tent words 0.00105571
