question data 0.001439625
evaluation results 0.0014065430000000001
unseen data 0.001305174
answer set 0.00127787
trec evaluation 0.001266839
training set 0.001261004
different question 0.001222428
statistical answer 0.001183226
human language 0.001157043
ing system 0.001143933
other trec 0.001137069
aquaint corpus 0.001102093
system seal 0.001100791
other methods 0.001086156
small set 0.001076383
additional answer 0.001063027
system ephyra 0.0010346040000000002
hybrid system 0.001033675
other strings 0.001029566
answering system 0.001028334
system overview 0.001023808
other source 0.001017982
trec questions 0.001017511
answer candidates 0.001016756
baseline scores 0.001013455
other candidates 0.001004097
brid system 9.870690000000001E-4
proprietary system 9.870690000000001E-4
question words 9.86138E-4
semantic approach 9.78097E-4
experimental results 9.73114E-4
date set 9.700139999999999E-4
answer type 9.68481E-4
top results 9.56824E-4
candidate sentences 9.543360000000001E-4
answer candidate 9.51145E-4
top answer 9.38102E-4
distance measures 9.234590000000001E-4
threshold trec 9.1862E-4
complete set 9.16504E-4
list questions 9.1416E-4
test fold 9.12598E-4
answer validation 9.103189999999999E-4
answer keys 9.07307E-4
same page 9.05696E-4
extraction rules 9.03767E-4
hybrid method 8.9686E-4
tive set 8.966569999999999E-4
uation results 8.87274E-4
tial set 8.86845E-4
inverse relation 8.85199E-4
relation edge 8.85199E-4
new list 8.83221E-4
answer types 8.8068E-4
answer candi 8.78188E-4
particular question 8.779339999999999E-4
approach seal 8.77753E-4
set expansion 8.7728E-4
answer selec 8.77001E-4
set expan 8.74722E-4
confidence measures 8.741150000000001E-4
other languages 8.73712E-4
answer extrac 8.73282E-4
initial answer 8.712279999999999E-4
ural language 8.703529999999999E-4
markup language 8.703529999999999E-4
poor answer 8.69057E-4
merging answer 8.671099999999999E-4
answer extractors 8.671099999999999E-4
effective method 8.666819999999999E-4
chinese entities 8.596189999999999E-4
hint words 8.53872E-4
low table 8.50985E-4
classification approach 8.48158E-4
confidence scores 8.45868E-4
corpus 8.4576E-4
information 8.45041E-4
top trec 8.171200000000001E-4
trec evaluations 8.15176E-4
similar structures 8.151300000000001E-4
tion features 8.13231E-4
average precision 8.12894E-4
matching approach 8.11747E-4
hybrid approach 8.10637E-4
additional candidates 8.09073E-4
complex questions 8.07411E-4
future experiments 8.07379E-4
bilistic scores 8.0546E-4
confidence score 8.04509E-4
small sets 7.99595E-4
trec participants 7.96436E-4
linguistic analysis 7.94739E-4
past trec 7.79903E-4
list question 7.78057E-4
total number 7.75706E-4
proposed approach 7.65456E-4
approach gener 7.65456E-4
future work 7.62684E-4
particular page 7.61621E-4
trec eval 7.603460000000001E-4
