tree kernel 0.00422727
tree features 0.00395828
tree kernels 0.003372869
parse tree 0.003259598
tree fragments 0.003127814
other tree 0.003099641
accuracy tree 0.003056074
dependency tree 0.0029600250000000002
tree pairs 0.002881835
subset tree 0.0028731850000000003
tree frag 0.002865105
tree patterns 0.002862478
tree ker 0.002843607
possible tree 0.0028367040000000002
specific tree 0.002831718
penn tree 0.002815204
shared tree 0.002812297
tree substitution 0.002811544
bayesian tree 0.00279613
mantic tree 0.002793378
tree rerank 0.0027901650000000003
conclusion tree 0.0027880360000000002
syntactic features 0.00245127
kernel learning 0.002386797
syntactic feature 0.0023809
kernel training 0.00228567
kernel method 0.002154829
kernel system 0.002112882
kernel approaches 0.002097296
such features 0.002066864
kernel methods 0.002063096
kernel cache 0.001939819
kernel meth 0.001939819
approximate kernel 0.001939819
tsg features 0.001864466
feature set 0.001817115
explicit features 0.001814193
relevant features 0.0017970459999999998
reranking feature 0.001780231
right features 0.001765528
explicit feature 0.001743823
irrelevant features 0.0017331129999999999
feature representation 0.0017170319999999998
implicit features 0.001695748
kernel 0.00168945
language model 0.00168539
length feature 0.001675626
features balance 0.0016719559999999998
feature sets 0.001662935
language classification 0.001638811
training data 0.001509865
features 0.00142046
parse trees 0.001355395
feature 0.00135009
grammatical classification 0.0013251629999999999
data settings 0.001316074
classification accuracy 0.001301437
relation extraction 0.0013006559999999999
discussion syntactic 0.001287215
syntactic tool 0.001282743
native language 0.00127126
classification tasks 0.001262123
domain data 0.001255202
positive data 0.001240363
exact data 0.001207348
language identification 0.001202074
model size 0.0011958709999999998
question classification 0.001182257
same manner 0.001165766
test accuracy 0.0011631
text classification 0.001158587
language modeling 0.001156268
language prediction 0.0011344179999999999
test set 0.001111871
trigram language 0.001108953
training time 0.001103897
gram model 0.001100297
subtree kernels 0.0010912650000000001
training set 0.001063245
cal classification 0.0010465980000000001
other tasks 0.001040761
classification arities 0.001035179
syntactic 0.00103081
test coarse 0.001004199
large models 9.96533E-4
berkeley parse 9.84815E-4
first dataset 9.79598E-4
single parse 9.77912E-4
test scores 9.66022E-4
future work 9.58804E-4
test sets 9.576910000000001E-4
cfg models 9.56103E-4
many settings 9.53944E-4
other words 9.4983E-4
many errors 9.46532E-4
system accuracy 9.41686E-4
relation 9.37817E-4
international corpus 9.33062E-4
many text 9.26919E-4
first version 9.21997E-4
