answer retrieval 0.003746328
answer candidate 0.003736492
ranking answer 0.003633646
correct answer 0.0036196220000000003
answer sentence 0.003554074
answer pairs 0.003543633
answer pair 0.003527394
answer sentences 0.003450613
answer candidates 0.0033994050000000003
answer terms 0.003388118
top answer 0.0033574060000000003
rect answer 0.003320844
answer candi 0.003313831
sequent answer 0.00330488
perfect answer 0.0033036560000000003
acceptable answer 0.003301933
answer 0.00306579
features word 0.00243278
semantic word 0.00237967
question term 0.002333204
semantic features 0.0022757899999999998
question terms 0.002182338
question answering 0.002177628
opinion question 0.002155804
representative question 0.002139462
relation features 0.001961052
question 0.00186001
word class 0.001777897
japanese word 0.0017658180000000002
word pairs 0.001746173
content word 0.001732651
analysis features 0.0017185500000000001
wordnet features 0.001686552
word phrase 0.0016678650000000001
polarity word 0.001652391
word polarity 0.001652391
word phrases 0.0016268720000000001
word classes 0.001622543
class semantic 0.001620907
semantic class 0.001620907
word level 0.001619925
semantic content 0.001575661
statistical features 0.001572298
polarity features 0.001548511
negative word 0.001547048
mantic word 0.001523193
retrieval system 0.001521719
head word 0.001519585
word polarities 0.001516646
respective word 0.001508593
semantic classes 0.0014655529999999999
net features 0.001452951
tic features 0.001423084
msa features 0.001405823
semantic orientation 0.001403549
training data 0.0014021160000000001
many questions 0.001393319
ranking answers 0.001386325
correct answers 0.001372301
semantic cat 0.0013488089999999999
different systems 0.0013461509999999999
same feature 0.001325194
factoid questions 0.001321194
first type 0.001296877
ing data 0.001267963
such sentiment 0.00123876
opinion questions 0.001231024
syntactic analysis 0.0012237609999999999
such pair 0.001217546
retrieval module 0.0012090339999999999
feature set 0.001202786
sentiment information 0.00120037
unseen questions 0.001173658
syntactic dependency 0.001173581
probabilistic model 0.001171084
retrieval method 0.0011703619999999999
test data 0.0011680100000000001
features 0.00116445
similar type 0.001150902
model parameters 0.001141791
feature sets 0.001133003
new type 0.001131505
japanese words 0.001128635
second type 0.001120978
semantic 0.00111134
baseline feature 0.001109148
trained model 0.001102927
content words 0.001095468
rect answers 0.001073523
significant performance 0.001073164
first work 0.0010618989999999998
performance improvement 0.0010564189999999998
candidate sentences 0.001055525
relation pattern 0.001054158
relation instance 0.001050184
same sentiment 0.001049636
causal relation 0.00104951
relation instances 0.00104548
similar words 0.001039065
grammatical relation 0.001038237
