E09-1096 |
statistics of our reference data for
|
document alignment
|
. Note that although there are
|
E09-1096 |
the general architecture of our
|
document alignment
|
system . It consists of three
|
E09-1096 |
monolingual corpora to derive a set of
|
document alignments
|
that are comparable in their
|
C00-1024 |
different languages . It is similar to
|
document alignment
|
in comparable corpus . Named
|
E12-1046 |
. BiLDA takes advantage of the
|
document alignment
|
by using a single variable that
|
E09-1096 |
structural similarity . One goal of
|
document alignment
|
is for parallel sentence extraction
|
E09-1096 |
document would imply fewer possible
|
document alignment
|
pairs for the system . This increases
|
N10-1063 |
extracting parallel sentences , as the
|
document alignment
|
is already provided . Table 1
|
N10-1063 |
Since our corpus already contains
|
document alignments
|
, we sidestep this problem ,
|
E09-1096 |
features for the task of bilingual
|
document alignment
|
. Experimental results on the
|
E09-1096 |
Previous works on monolingual
|
document alignment
|
focus on automatic alignment
|
E09-1096 |
of features in our system . Our
|
document alignment
|
system consists of three stages
|
E09-1096 |
content . Fig 1 . Architecture for
|
Document Alignment
|
Model . 2.1 Candidate Generation
|
D15-1015 |
However , these models require
|
document alignments
|
as initial bilingual signals
|
E09-1096 |
In addition , they argue that
|
document alignment
|
should be done before parallel
|
C04-1151 |
quasi-comparable corpora , this
|
document alignment
|
step also serves as topic alignment
|
P09-4006 |
alignment score . After that ,
|
document alignment
|
is carried out between aligned
|
E09-1096 |
1 Introduction The problem of
|
document alignment
|
is described as the task of aligning
|
E09-1096 |
such as Tao 's ( 2005 ) . Besides
|
document alignment
|
as an end , there are many tasks
|
E06-1021 |
tablished a simple and robust
|
document alignment
|
method , we leave its application
|