tree kernel 0.00312611
tree fragment 0.002658091
tree fragments 0.0024304500000000002
tree structure 0.002355548
parse tree 0.002258173
different kernel 0.002244733
other tree 0.002208262
subset tree 0.002042723
partial tree 0.001981754
tree ker 0.001980171
kernel function 0.00197646
syntactic features 0.001965183
different features 0.001961053
minimal tree 0.001956845
plored tree 0.001956845
kernel space 0.001942053
feature space 0.001878493
target kernel 0.001813757
training data 0.001797331
such fragment 0.001762908
sst kernel 0.001761904
linear kernel 0.001749576
kernel product 0.0017313580000000001
original kernel 0.001714695
kernel support 0.001712865
path feature 0.001700734
structured feature 0.001688517
kernel functions 0.001688395
feature selection 0.0016825499999999999
data set 0.0016796509999999999
feature extraction 0.001676266
kernel machines 0.001669599
kernel trick 0.0016524929999999999
kernel families 0.001651915
relevant features 0.001593722
aggressive feature 0.0015891219999999999
training time 0.001570463
test data 0.001564051
syntactic fragments 0.001553203
training set 0.001548912
important features 0.001536264
several features 0.001511302
svm model 0.001503859
explicit features 0.001499937
lexical features 0.0014792070000000002
fragment space 0.001474034
structured features 0.0014683970000000002
such test 0.0014204629999999998
kernel 0.00141048
such algorithm 0.0013979029999999998
svm training 0.001394034
classification task 0.001391997
classification accuracy 0.001388155
irrelevant features 0.0013756480000000001
vant features 0.001373422
evant features 0.0013689820000000001
feature 0.00134692
data points 0.001342096
test time 0.001337183
positive data 0.001318248
test set 0.001315632
srl fragment 0.001307041
sst model 0.001294545
large data 0.00129316
base fragment 0.0012854680000000001
linear model 0.001282217
data sets 0.001272191
fragment extraction 0.001271807
space classification 0.00126984
rich fragment 0.001263697
same results 0.001261476
available data 0.00125567
whole fragment 0.001255532
other nodes 0.001250306
huge fragment 0.001245517
accuracy class 0.001245165
fragment expansion 0.001235814
fast model 0.001231249
fragment definitions 0.0012263880000000001
classifiers data 0.001226104
new set 0.001224747
fragment mining 0.001214609
training points 0.001211357
tive fragment 0.001201471
fragment spaces 0.001196988
mine model 0.001188626
positive training 0.001187509
argument nodes 0.001186505
sst training 0.00118472
fragment proliferation 0.001184531
fragment defi 0.001184531
convolution kernels 0.001183955
relevant fragments 0.001181742
language tasks 0.0011784579999999998
predicate nodes 0.001177863
structure space 0.001171491
such vectors 0.0011661829999999999
question classification 0.0011587770000000002
tion training 0.001152799
syntactic patterns 0.001152166
