D08-1042 |
the leaf projection paths using
|
string kernels
|
. A few kernels based on dependency
|
D09-1029 |
the generic definition of the
|
string kernel
|
in order to take into account
|
D09-1142 |
representation generated by the
|
string kernel
|
is used to find a hyperplane
|
D10-1023 |
similarity between two examples . The
|
string kernel
|
we described in the previous
|
D09-1142 |
the substrings generated by the
|
string kernel
|
have no position information
|
D10-1023 |
al. , 2002 ) . Informally , a
|
string kernel
|
aims to efficiently compute the
|
C04-1109 |
quite well for the task . Although
|
string kernels
|
can capture common word subsequences
|
D10-1023 |
it uses a Regression SVM with a
|
string kernel
|
to predict essay scores . 6.3
|
D09-1029 |
feature space associated with the
|
string kernel
|
of length n is indexed by a set
|
D09-1029 |
syntagmatic kernels as a combination of
|
string kernels
|
applied to sequences of words
|
D09-1029 |
sequences of words . Therefore , the
|
string kernel
|
( Shawe-Taylor and Cristianini
|
D08-1082 |
built on top of SVMstruct with
|
string kernels
|
. Additionally , there is substantial
|
D10-1023 |
alignment , alignment kernels , and
|
string kernels
|
. 2 Corpus Information We use
|
D09-1029 |
subsequences . The associated
|
string kernel
|
is defined by Kn ( si , sj )
|
D10-1023 |
learning -- sequence alignment and
|
string kernels
|
, as well as the introduction
|
D09-1112 |
will be referred as SK1 and SK2 (
|
String Kernel
|
1 and 2 ) . They differer in
|
D09-1142 |
kernel ridge regression and the
|
string kernel
|
( KRR-SK ) presented in Section
|
D09-1142 |
whether that is the case , we use a
|
string kernel
|
, which for a given string (
|
D09-1112 |
between si and sj . Since the
|
string kernel
|
skips some elements of the target
|
C04-1109 |
, 2002 ) compared a wordbased
|
string kernel
|
and n-gram kernels at the sequence
|