I
n
t
e
r
n
at
i
on
a
l
Jou
r
n
al
of
A
d
van
c
e
s
i
n
A
p
p
l
i
e
d
S
c
i
e
n
c
e
s
(
I
JA
A
S
)
V
ol
.
1
5
, N
o.
2
,
J
une
20
2
6
, pp.
532
~
540
I
S
S
N
:
2252
-
8814
,
D
O
I
:
10.11591/
i
j
a
a
s
.
v
1
5
.
i
2
.
pp
532
-
540
532
Jou
r
n
al
h
om
e
page
:
h
t
t
p
:
/
/
i
j
aa
s
.i
ae
s
c
o
r
e
.c
om
S
e
m
an
t
i
c
c
l
u
s
t
e
r
i
n
g of
s
c
i
e
n
t
i
f
i
c
ab
s
t
r
ac
t
s
w
i
t
h
t
r
an
s
f
or
m
e
r
e
m
b
e
d
d
i
n
gs
an
d
t
r
ad
i
t
i
on
al
t
e
x
t
r
e
p
r
e
s
e
n
t
at
i
on
s
M
u
s
t
h
of
a G
a
l
i
h
P
r
ad
an
a
1
, P
u
j
o
H
ar
i
S
ap
u
t
r
o
2
, A
r
d
h
yan
s
yah
M
u
al
o
3
, A
r
b
i
at
i
F
ai
z
ah
4
,
Wah
yu
n
i
F
i
t
h
r
at
u
l
Z
al
m
i
2
1
D
e
pa
r
t
m
e
nt
of
D
a
t
a
S
c
i
e
nc
e
, F
a
c
ul
t
y of
C
om
put
e
r
S
c
i
e
nc
e
,
U
P
N
V
e
t
e
r
a
n
J
a
ka
r
t
a
,
J
a
ka
r
t
a
,
I
n
d
one
s
i
a
2
D
e
pa
r
t
m
e
nt
of
I
nf
or
m
a
t
i
c
s
E
ngi
ne
e
r
i
ng
, F
a
c
ul
t
y of
E
ngi
ne
e
r
i
ng, S
a
m
R
a
t
ul
a
ngi
U
ni
ve
r
s
i
t
y, M
a
na
d
o, I
n
d
one
s
i
a
3
D
e
pa
r
t
m
e
nt
of
I
nf
or
m
a
t
i
c
s
M
a
na
ge
m
e
nt
,
P
ol
i
t
e
kni
k N
e
ge
r
i
F
a
kF
a
k
, F
a
kf
a
k, I
nd
one
s
i
a
4
D
e
pa
r
t
m
e
nt
of
I
nf
or
m
a
t
i
on
S
ys
t
e
m
s
,
F
a
c
ul
t
y
of
I
nf
or
m
a
t
i
c
s
,
I
ns
t
i
t
u
t
T
e
knol
ogi
d
a
n B
i
s
ni
s
P
G
R
I
D
e
w
a
n
t
a
r
a
, J
om
ba
ng, I
nd
one
s
i
a
A
r
t
i
c
l
e
I
n
f
o
A
B
S
T
R
A
C
T
A
r
t
i
c
l
e
h
i
s
t
o
r
y
:
R
e
c
e
i
ve
d
J
un 23, 2025
R
e
vi
s
e
d
A
pr
10, 2026
A
c
c
e
pt
e
d
A
pr
22, 2026
Th
e
large
a
n
d
diverse
q
u
a
nt
i
t
y
of
scie
nt
ific
doc
um
e
nt
s
i
n
th
e
world
,
a
n
d
specifically
i
n
I
n
do
n
esia
,
m
akes
th
e
process
of
processi
n
g
scie
nt
ific
docum
e
nt
da
t
a
a
n
i
nt
eres
t
i
n
g
s
tu
dy
.
O
n
e
th
a
t
represe
nt
s
th
e
e
nt
ire
scie
nt
ific
docum
e
nt
is
th
ro
u
g
h
abs
t
rac
t
s
.
T
h
e
approac
h
th
a
t
ca
n
be
do
n
e
for
th
e
process
of
processi
n
g
a
n
d
gro
u
pi
n
g
doc
um
e
nt
s
is
t
o
apply
c
l
u
s
t
eri
n
g
.
I
n
th
is
case
,
t
ex
t
-
based
cl
u
s
t
eri
n
g
is
c
u
rre
nt
ly
h
e
avily
i
n
fl
u
e
n
ced
by
th
e
fea
tu
re
represent
a
t
io
n
of
th
e
t
ex
t
da
t
a
u
sed
.
So
m
e
pop
u
lar
represe
nt
a
t
io
n
s
of
fea
tu
res
are
t
er
m
freq
u
e
n
cy
-
i
n
verse
doc
um
e
nt
freq
u
e
n
cy
(
TF
-
IDF
)
a
n
d
co
unt
vect
orizer
,
b
ut
th
ey
s
t
il
l
h
ave
sig
n
ifica
nt
weak
n
esses
i
n
th
e
co
nt
ex
t
of
un
ders
t
a
n
di
n
g
th
e
m
ea
n
i
n
g
of
n
a
tu
ral
la
n
g
u
age
.
To
cover
th
e
drawbacks
,
i
t
can
u
se
t
ra
n
sfor
m
ers
or
e
m
beddi
n
g
t
ypes.
I
n
th
is
s
tu
dy
,
several
t
es
t
sce
n
arios
will
be
carried
o
ut
t
o
ob
t
ai
n
i
n
for
m
a
t
io
n
a
n
d
a
n
overview
of
h
ow
t
o
co
m
pare
th
e
effec
t
ive
n
ess
of
th
e
t
radi
t
io
n
al
TF
-
IDF
m
odel
a
n
d
th
e
bidirec
t
io
n
al
en
coder
represe
nt
a
t
io
n
s
fro
m
t
ra
n
sfor
m
ers
(
BERT
)
a
n
d
se
nt
e
n
ce
bidirect
io
n
al
e
n
coder
represe
nt
a
t
io
n
s
fro
m
t
ra
n
sfor
m
ers
(
SBERT
)
em
beddi
n
g
m
odels
i
n
I
n
do
n
esia
n
-
la
n
g
u
a
ge
scie
nt
ific
abs
t
rac
t
cl
u
s
t
eri
n
g
wi
th
several
cl
u
s
t
eri
n
g
m
odels
,
s
u
c
h
as
k
-
m
ea
n
s
a
n
d
agglo
m
era
t
ive
.
T
h
e
res
u
l
t
s
of
th
e
s
tu
dy
s
h
owed
th
a
t
th
e
m
os
t
effec
t
ive
cl
u
s
t
eri
n
g
ob
t
ai
n
ed
was
by
u
si
n
g
a
n
em
beddi
n
g
m
odel
of
a
co
m
bi
n
a
t
io
n
of
BERT
a
n
d
k
-
m
ea
n
s
,
w
h
ic
h
w
as
th
e
m
os
t
co
n
siste
nt
wi
th
the
m
os
t
op
t
i
m
al
num
ber
of
cl
u
s
t
ers
bei
n
g
2
cl
u
s
t
ers.
K
e
y
w
or
d
s
:
A
bs
t
r
a
c
t
C
l
us
t
e
r
i
ng
E
m
be
ddi
ng
T
e
xt
T
r
a
di
t
i
ona
l
This
is
an
open
access
article
under
the
CC
BY
-
SA
license
.
C
or
r
e
s
pon
di
n
g A
u
t
h
or
:
M
us
t
hof
a
G
a
l
i
h P
r
a
da
na
D
e
pa
r
t
m
e
n
t
of
D
a
t
a
S
c
i
e
nc
e
, F
a
c
u
l
t
y of
C
om
pu
t
e
r
S
c
i
e
nc
e
, U
P
N
V
e
t
e
r
a
n J
a
ka
r
t
a
R
S
F
a
t
m
a
w
a
t
i
R
oa
d, P
ondok L
a
bu, C
i
l
a
nda
k,
J
a
ka
r
t
a
, I
ndone
s
i
a
E
m
a
i
l
:
m
us
t
hof
a
ga
l
i
hp
r
a
da
na
@
upnvj
.a
c
.
i
d
1.
I
N
T
R
O
D
U
C
T
I
O
N
S
c
i
e
nt
i
f
i
c
doc
um
e
nt
s
t
oda
y
ha
ve
a
l
a
r
ge
a
nd
di
ve
r
s
e
qua
nt
i
t
y,
a
nd
t
he
e
a
s
e
of
a
c
c
e
s
s
t
o
s
c
i
e
n
t
i
f
i
c
doc
um
e
nt
s
ha
s
a
l
s
o
be
c
o
m
e
e
a
s
i
e
r
be
c
a
us
e
m
a
ny t
e
c
hnol
og
i
e
s
c
a
n
he
l
p
t
o
ga
i
n
a
c
c
e
s
s
,
a
nd m
a
ke
a
s
c
i
e
n
t
i
f
i
c
doc
um
e
nt
m
or
e
a
c
c
e
s
s
i
b
l
e
.
T
hi
s
i
s
s
how
n
by
a
s
i
gni
f
i
c
a
n
t
i
nc
r
e
a
s
e
i
n
t
he
num
be
r
of
s
c
i
e
nt
i
f
i
c
publ
i
c
a
t
i
ons
.
I
ndone
s
i
a
i
t
s
e
l
f
i
n
2024
w
i
l
l
be
r
a
nke
d
19t
h
i
n
t
he
w
or
l
d
a
nd
5
t
h
i
n
t
h
e
A
s
i
a
n
r
e
g
i
on
w
i
t
h
a
t
o
t
a
l
of
58,224
i
nde
xe
d
s
c
i
e
nt
i
f
i
c
doc
um
e
n
t
s
.
T
h
i
s
c
ondi
t
i
on
s
how
s
t
ha
t
na
t
i
ona
l
r
e
s
e
a
r
c
h
a
nd
pub
l
i
c
a
t
i
on
a
c
t
i
vi
t
i
e
s
c
ont
i
nue
t
o
gr
ow
r
a
p
i
dl
y,
i
n
l
i
ne
w
i
t
h
t
he
e
a
s
e
of
a
c
c
e
s
s
t
o
i
n
f
or
m
a
t
i
on
a
nd
t
he
i
m
pl
e
m
e
nt
a
t
i
on
of
ope
n
publ
i
c
a
t
i
on
pol
i
c
i
e
s
.
H
ow
e
ve
r
,
t
he
m
a
s
s
i
ve
i
n
c
r
e
a
s
e
i
n
t
he
vo
l
um
e
of
s
c
i
e
n
t
i
f
i
c
da
t
a
a
l
s
o
pos
e
s
ne
w
c
ha
l
l
e
nge
s
i
n t
he
pr
oc
e
s
s
of
i
n
f
or
m
a
t
i
on
m
a
na
ge
m
e
nt
a
nd a
na
l
y
s
i
s
, e
s
pe
c
i
a
l
l
y i
n t
h
e
gr
oupi
ng or
c
l
us
t
e
r
i
ng of
doc
um
e
nt
s
ba
s
e
d on s
i
m
i
l
a
r
i
t
y
i
n c
ont
e
n
t
a
nd c
ont
e
xt
,
ba
s
e
d on n
a
t
ur
a
l
l
a
ngu
a
ge
pr
oc
e
s
s
i
ng
[
1]
, [
2
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
dv A
ppl
S
c
i
I
S
S
N
:
2252
-
8814
Se
m
ant
i
c
c
l
us
t
e
r
i
ng of
s
c
i
e
nt
i
f
i
c
abs
t
r
ac
t
s
w
i
t
h
t
r
ans
f
o
r
m
e
r
e
m
be
ddi
ngs
…
(
M
us
t
hof
a G
al
i
h P
r
adana
)
533
T
he
qua
l
i
t
y
of
t
he
c
l
us
t
e
r
i
ng
r
e
s
ul
t
s
i
s
h
i
ghl
y
de
pe
nd
e
nt
on
t
he
t
e
xt
r
e
pr
e
s
e
nt
a
t
i
on
t
e
c
hni
qu
e
us
e
d.
C
l
a
s
s
i
c
a
ppr
oa
c
he
s
l
i
ke
a
c
ount
ve
c
t
o
r
i
z
e
r
[
3
]
–
[
6]
a
nd
t
e
r
m
f
r
e
que
nc
y
-
i
nve
r
s
e
doc
um
e
n
t
f
r
e
que
nc
y
(
TF
-
I
D
F
)
[
7]
–
[
12]
ha
ve
l
ong
be
e
n
u
t
i
l
i
z
e
d
t
o e
xt
r
a
c
t
w
or
d
f
e
a
t
ur
e
s
,
e
a
c
h
w
i
t
h
i
t
s
a
dv
a
nt
a
ge
s
a
nd l
i
m
i
t
a
t
i
on
s
[
13
]
–
[
16]
.
A
l
t
hough
e
f
f
e
c
t
i
ve
i
n
m
ode
l
i
ng
w
or
d
f
r
e
que
nc
y,
t
he
m
e
t
hod
i
s
l
e
s
s
a
bl
e
t
o
unde
r
s
t
a
nd
s
e
m
a
n
t
i
c
m
e
a
ni
ng
i
n
de
pt
h. T
he
s
e
l
i
m
i
t
a
t
i
on
s
c
a
n now
be
ove
r
c
om
e
w
i
t
h a
t
r
a
ns
f
o
r
m
e
r
-
ba
s
e
d a
ppr
o
a
c
h
[
17]
–
[
19
]
, w
hi
c
h i
s
a
bl
e
t
o
r
e
pr
e
s
e
n
t
t
h
e
c
on
t
e
xt
ua
l
r
e
l
a
t
i
ons
h
i
p
be
t
w
e
e
n
w
or
ds
i
n
t
w
o
d
i
r
e
c
t
i
ons
.
T
hi
s
c
a
pa
bi
l
i
t
y
m
a
k
e
s
e
m
be
ddi
ng
t
r
a
ns
f
o
r
m
e
r
s
s
u
c
h
a
s
b
i
di
r
e
c
t
i
ona
l
e
nc
ode
r
r
e
pr
e
s
e
nt
a
t
i
ons
f
r
o
m
t
r
a
ns
f
o
r
m
e
r
s
(
B
E
R
T
)
s
upe
r
i
or
i
n
c
o
m
pl
e
x
t
e
xt
c
l
us
t
e
r
i
ng
pr
oc
e
s
s
e
s
,
i
nc
l
ud
i
ng
s
c
i
e
n
t
i
f
i
c
doc
um
e
n
t
s
[
11]
,
[
12]
,
c
o
m
pa
r
e
d
w
i
t
h
t
he
t
r
a
di
t
i
ona
l
t
e
xt
r
e
pr
e
s
e
n
t
a
t
i
on
, e
ve
n w
he
n c
om
b
i
ne
d w
i
t
h
i
m
a
ge
p
r
oc
e
s
s
i
ng i
npu
t
s
[
20
]
, [
21]
.
V
a
r
i
ous
pr
e
v
i
ous
s
t
udi
e
s
ha
ve
s
how
n
va
r
i
ous
a
ppr
o
a
c
he
s
t
o
doc
um
e
n
t
c
l
u
s
t
e
r
i
ng,
s
t
a
r
t
i
ng f
r
om
d
a
t
a
m
i
ni
ng
m
e
t
hods
[
22]
,
opt
i
m
i
z
a
t
i
on
m
e
t
a
-
he
ur
i
s
t
i
c
s
[
23]
,
a
u
t
oe
nc
ode
r
[
24
]
,
t
o
de
e
p
m
ul
t
i
-
l
a
be
l
f
r
a
m
e
w
or
ks
s
uc
h
a
s
de
e
p
e
xt
r
e
m
e
m
ul
t
i
-
l
a
be
l
l
e
a
r
ni
ng
(
D
e
e
pX
M
L
)
[
25]
.
T
he
a
ppl
i
c
a
t
i
on
of
t
he
B
E
R
T
m
ode
l
t
o
a
l
gor
i
t
hm
s
s
uc
h
a
s
k
-
m
e
a
ns
,
F
uz
z
y
-
C
-
M
e
a
ns
,
a
nd
hi
e
r
a
r
c
hi
c
a
l
c
l
us
t
e
r
i
ng
i
n
ge
ne
r
a
l
r
e
s
ul
t
s
i
n
i
m
pr
ove
d
pe
r
f
or
m
a
nc
e
ove
r
T
F
-
I
D
F
[
26]
,
[
27
]
.
S
om
e
a
dva
nc
e
d
m
ode
l
de
ve
l
op
m
e
nt
s
,
s
uc
h
a
s
W
E
C
l
us
t
e
r
i
ng+
+
[
28]
a
nd
C
l
us
T
op
[
29]
,
a
s
w
e
l
l
a
s
t
he
us
e
of
e
m
b
e
ddi
ng
f
r
o
m
l
a
r
g
e
l
a
ngua
ge
m
ode
l
s
s
uc
h
a
s
G
P
T
-
3.5
a
nd
s
e
nt
e
nc
e
bi
d
i
r
e
c
t
i
ona
l
e
nc
ode
r
r
e
pr
e
s
e
nt
a
t
i
ons
f
r
om
t
r
a
ns
f
o
r
m
e
r
s
(
S
B
E
R
T
)
,
a
l
s
o
s
how
e
d
i
nc
r
e
a
s
e
d
a
c
c
ur
a
c
y
i
n doc
um
e
n
t
gr
oupi
ng.
I
n
de
t
e
r
m
i
ni
ng
t
he
num
be
r
of
c
l
u
s
t
e
r
s
t
ha
t
a
r
e
m
os
t
s
u
i
t
a
b
l
e
, a
nu
m
be
r
o
f
e
va
l
ua
t
i
on m
e
t
r
i
c
s
s
uc
h
a
s
t
he
s
i
l
houe
t
t
e
s
c
or
e
[
30
]
a
nd D
a
vi
e
s
-
B
ou
l
di
n i
nde
x
[
31]
a
r
e
o
f
t
e
n a
pp
l
i
e
d t
o
m
e
a
s
u
r
e
t
he
qua
l
i
t
y o
f
s
e
pa
r
a
t
i
on
be
t
w
e
e
n
c
l
us
t
e
r
s
.
S
om
e
s
t
ud
i
e
s
us
e
a
c
o
m
bi
na
t
i
on
of
m
e
t
hods
,
s
uc
h
a
s
ga
p
s
t
a
t
i
s
t
i
c
s
[
32]
o
r
t
he
e
l
bow
m
e
t
hod
,
t
o
s
t
r
e
ngt
he
n
t
he
r
e
s
ul
t
s
of
de
t
e
r
m
i
n
i
ng
opt
i
m
a
l
k
va
l
ue
s
.
M
e
a
nw
hi
l
e
,
c
l
a
s
s
i
c
m
ode
l
s
s
uc
h
a
s
TF
-
I
D
F
a
nd
l
a
t
e
nt
D
i
r
i
c
h
l
e
t
a
l
l
o
c
a
t
i
on
(
L
D
A
)
r
e
m
a
i
n
u
s
e
d
a
s
be
nc
hm
a
r
ks
t
o
a
s
s
e
s
s
t
he
e
x
t
e
nt
t
o
w
hi
c
h
t
r
a
ns
f
o
r
m
e
r
-
ba
s
e
d a
ppr
oa
c
he
s
c
a
n i
m
pr
ove
c
l
us
t
e
r
i
ng ou
t
c
om
e
s
[
33]
–
[
36]
.
I
n ge
ne
r
a
l
, va
r
i
ous
s
t
ud
i
e
s
ha
ve
s
how
n t
ha
t
t
r
a
n
s
f
or
m
e
r
e
m
be
ddi
ng
-
b
a
s
e
d r
e
pr
e
s
e
nt
a
t
i
ons
a
r
e
a
bl
e
t
o
pr
ovi
de
be
t
t
e
r
r
e
s
ul
t
s
t
ha
n
t
r
a
di
t
i
ona
l
a
pp
r
oa
c
he
s
.
H
ow
e
ve
r
,
s
i
m
i
l
a
r
r
e
s
e
a
r
c
h
w
i
t
h
a
f
oc
us
on
I
ndone
s
i
a
n
s
c
i
e
nt
i
f
i
c
doc
um
e
nt
s
i
s
s
t
i
l
l
l
i
m
i
t
e
d,
a
l
t
hough
t
he
num
be
r
of
publ
i
c
a
t
i
on
s
c
ont
i
nue
s
t
o
i
nc
r
e
a
s
e
.
T
he
r
e
f
or
e
,
t
hi
s
s
t
udy
w
i
l
l
c
o
m
pa
r
e
t
he
e
f
f
e
c
t
i
ve
ne
s
s
of
T
F
-
I
D
F
,
c
ount
ve
c
t
or
i
z
e
r
,
a
nd
B
E
R
T
or
S
B
E
R
T
i
n
t
he
gr
oupi
ng
of
I
ndone
s
i
a
n
s
c
i
e
n
t
i
f
i
c
a
b
s
t
r
a
c
t
s
.
T
he
r
e
s
ul
t
s
of
t
hi
s
r
e
s
e
a
r
c
h
a
r
e
e
xpe
c
t
e
d
t
o
be
a
f
oo
t
hol
d
f
o
r
t
he
de
ve
l
opm
e
n
t
of
s
ys
t
e
m
s
s
uc
h a
s
a
c
a
de
m
i
c
s
e
a
r
c
h e
ng
i
ne
s
, t
he
m
a
t
i
c
m
a
pp
i
ng, a
ut
o
m
a
t
i
c
l
i
t
e
r
a
t
u
r
e
r
e
vi
e
w
, a
nd
r
e
s
e
a
r
c
h d
a
s
hboa
r
ds
i
n t
he
f
ut
u
r
e
.
2.
M
E
T
H
O
D
T
hi
s
r
e
s
e
a
r
c
h
i
s
c
a
r
r
i
e
d
out
t
hr
ough
s
e
ve
r
a
l
s
t
a
g
e
s
.
T
he
s
e
s
t
a
ge
s
a
r
e
a
r
r
a
nge
d
i
n
a
s
t
r
u
c
t
ur
e
d
w
a
y,
s
t
a
r
t
i
ng f
r
om
pr
ob
l
e
m
f
or
m
ul
a
t
i
on
t
o dr
a
w
i
ng c
onc
l
us
i
on
s
. T
he
r
e
s
e
a
r
c
h f
l
ow
i
s
vi
s
u
a
l
i
z
e
d i
n F
i
gur
e
1.
F
i
gur
e
1. R
e
s
e
a
r
c
h
s
t
a
ge
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8814
I
nt
J
A
dv A
ppl
S
c
i
,
V
ol
. 15, N
o. 2, J
une
2026
:
532
-
540
534
T
he
de
t
a
i
l
s
of
t
h
e
i
m
p
l
e
m
e
nt
a
t
i
on of
t
he
r
e
s
e
a
r
c
h a
r
e
a
s
f
ol
l
ow
s
:
i)
P
r
obl
e
m
i
de
n
t
i
f
i
c
a
t
i
on:
t
he
f
i
r
s
t
s
t
e
p
of
t
he
r
e
s
e
a
r
c
h
i
s
t
o r
e
c
ogni
z
e
t
h
e
pr
ob
l
e
m
s
be
hi
nd
t
hi
s
s
t
udy.
T
he
i
nc
r
e
a
s
i
ng
num
b
e
r
o
f
s
c
i
e
nt
i
f
i
c
publ
i
c
a
t
i
ons
de
m
a
nds
a
n
a
ut
o
m
a
t
e
d
m
e
c
ha
n
i
s
m
t
o
gr
oup
a
bs
t
r
a
c
t
s
t
o
m
a
ke
t
he
m
e
a
s
i
e
r
t
o s
e
a
r
c
h.
ii)
S
t
udy
l
i
t
e
r
a
t
u
r
e
:
t
he
ne
x
t
s
t
a
ge
i
s
t
o
r
e
v
i
e
w
pr
e
vi
ous
s
t
ud
i
e
s
r
e
l
e
va
nt
t
o
t
hi
s
t
he
m
e
.
T
he
f
oc
us
of
t
he
s
t
udy i
s
di
r
e
c
t
e
d t
ow
a
r
ds
t
e
x
t
r
e
pr
e
s
e
nt
a
t
i
on a
pp
r
oa
c
he
s
s
uc
h a
s
T
F
-
I
D
F
, B
E
R
T
,
a
nd S
B
E
R
T
, a
s
w
e
l
l
a
s
unl
a
be
l
e
d g
r
oupi
ng t
e
c
hni
que
s
s
uc
h a
s
k
-
m
e
a
n
s
a
nd
a
ggl
o
m
e
r
a
t
i
ve
c
l
us
t
e
r
i
ng
.
i
i
i
)
D
a
t
a
c
ol
l
e
c
t
i
on
a
nd
p
r
e
pa
r
a
t
i
on
:
t
he
r
e
s
e
a
r
c
h
d
a
t
a
i
s
i
n
t
h
e
f
or
m
of
a
c
ol
l
e
c
t
i
on
of
s
c
i
e
nt
i
f
i
c
a
bs
t
r
a
c
t
s
obt
a
i
ne
d f
r
om
d
i
gi
t
a
l
pub
l
i
c
a
t
i
on ba
s
e
s
. A
l
l
t
e
xt
i
s
t
h
e
n pr
oc
e
s
s
e
d f
i
r
s
t
s
o t
ha
t
i
t
ha
s
a
uni
f
or
m
f
or
m
a
t
.
i
v)
T
e
xt
e
m
be
dd
i
ng:
i
n
o
r
de
r
f
or
t
h
e
a
l
go
r
i
t
h
m
t
o
be
pr
oc
e
s
s
e
d,
t
he
t
e
xt
i
s
c
onve
r
t
e
d i
nt
o
a
nu
m
e
r
i
c
ve
c
t
o
r
t
hr
ough
t
w
o
a
ppr
oa
c
he
s
.
I
n
t
he
t
r
a
di
t
i
ona
l
m
e
t
hod,
T
F
-
I
D
F
i
s
us
e
d,
w
hi
c
h
c
a
l
c
ul
a
t
e
s
t
he
w
e
i
ght
o
f
w
or
ds
ba
s
e
d on t
he
f
r
e
que
nc
y o
f
t
he
i
r
a
ppe
a
r
a
nc
e
i
n do
c
um
e
n
t
s
.
v)
C
l
us
t
e
r
i
ng
a
ppl
i
c
a
t
i
on
:
a
f
t
e
r
t
he
t
e
x
t
i
s
r
e
p
r
e
s
e
nt
e
d
i
nt
o
ve
c
t
o
r
s
,
t
he
gr
oup
i
ng
pr
oc
e
s
s
i
s
c
a
r
r
i
e
d
out
us
i
ng t
w
o uns
upe
r
v
i
s
e
d a
l
gor
i
t
hm
s
.
vi
)
E
va
l
ua
t
i
on
:
t
he
qua
l
i
t
y
of
t
he
c
l
u
s
t
e
r
s
w
a
s
a
na
l
yz
e
d
us
i
ng
t
w
o
m
a
i
n
m
e
a
s
ur
e
s
,
na
m
e
l
y
t
he
s
i
l
houe
t
t
e
s
c
or
e
a
nd
t
he
D
a
vi
e
s
–
B
oul
d
i
n
i
nde
x
.
S
i
l
houe
t
t
e
s
c
or
e
i
s
us
e
d
t
o
a
s
s
e
s
s
t
he
pr
oxi
m
i
t
y
of
a
doc
um
e
n
t
t
o
i
t
s
c
l
us
t
e
r
c
om
pa
r
e
d t
o o
t
he
r
c
l
us
t
e
r
s
.
vi
i
)
C
onc
l
us
i
on
:
t
he
f
i
na
l
s
t
a
ge
i
s
t
o
p
r
e
s
e
n
t
t
he
r
e
s
u
l
t
s
o
f
t
he
c
l
us
t
e
r
v
i
s
ua
l
i
z
a
t
i
on
o
f
t
h
e
va
r
i
ous
c
onf
i
gur
a
t
i
ons
a
nd c
om
pa
r
e
t
he
pe
r
f
or
m
a
nc
e
o
f
e
a
c
h m
e
t
hod.
3.
R
E
S
U
L
T
S
A
N
D
D
I
S
C
U
S
S
I
O
N
3.1.
R
aw
d
at
a
T
he
da
t
a
us
e
d i
n t
hi
s
s
t
udy
a
r
e
a
bs
t
r
a
c
t
da
t
a
f
r
om
s
c
i
e
n
t
i
f
i
c
r
e
s
e
a
r
c
h doc
um
e
n
t
s
. T
he
a
m
ount
of
da
t
a
i
s
a
s
m
a
ny
a
s
1
,
000
a
bs
t
r
a
c
t
da
t
a
po
i
nt
s
.
T
hi
s
a
b
s
t
r
a
c
t
da
t
a
i
s
obt
a
i
ne
d
f
r
om
t
he
da
t
a
r
e
po
s
i
t
o
r
y
of
a
bs
t
r
a
c
t
s
t
ude
nt
t
he
s
i
s
da
t
a
a
t
t
he
un
i
ve
r
s
i
t
y
by t
a
ki
ng s
pe
c
i
a
l
s
pe
c
i
f
i
c
a
t
i
ons
a
t
t
h
e
F
a
c
ul
t
y
of
C
om
pu
t
e
r
S
c
i
e
nc
e
.
T
he
pr
e
-
pr
oc
e
s
s
i
ng
s
t
a
ge
s
a
r
e
d
i
vi
de
d
i
n
t
o
2
pa
r
t
s
,
a
c
c
o
r
di
ng
t
o
t
h
e
ne
e
ds
of
e
a
c
h
t
ype
of
a
ppr
oa
c
h.
T
he
t
r
a
di
t
i
ona
l
a
pp
r
oa
c
h
r
e
qui
r
e
s
s
e
ve
r
a
l
s
t
e
ps
a
nd
s
t
a
ge
s
t
h
a
t
a
r
e
m
or
e
de
t
a
i
l
e
d
a
nd
i
n
-
d
e
pt
h,
w
hi
l
e
us
i
ng
t
he
e
m
be
ddi
ng
a
ppr
oa
c
h,
it
onl
y
us
e
s
c
a
s
e
f
ol
d
i
ng
a
nd
l
i
gh
t
c
l
e
a
n
i
ng
w
i
t
hout
s
t
e
m
m
i
ng,
be
c
a
us
e
e
m
be
ddi
ng
r
e
qui
r
e
s
c
om
p
l
e
t
e
ne
s
s
a
nd
c
o
m
pl
e
t
e
ne
s
s
o
f
s
e
nt
e
nc
e
s
t
o
g
e
t
t
he
c
ont
e
x
t
a
s
a
w
hol
e
.
I
n
t
he
t
r
a
d
i
t
i
ona
l
m
ode
l
,
t
he
c
a
s
e
f
ol
d
i
ng
pr
oc
e
s
s
i
s
by
c
onve
r
t
i
ng
t
e
x
t
t
o
l
ow
e
r
c
a
s
e
.
T
he
ne
xt
pr
o
c
e
s
s
i
s
c
l
e
a
ni
ng f
or
punc
t
ua
t
i
on,
a
s
w
e
l
l
a
s
t
he
t
oke
n
i
z
a
t
i
on
pr
oc
e
s
s
.
S
t
opw
or
d
r
e
m
ova
l
r
e
m
ove
s
m
e
a
n
i
ngl
e
s
s
w
or
ds
a
nd
c
on
t
i
nue
s
t
h
e
s
t
e
m
m
i
ng
or
l
e
m
m
a
t
i
z
a
t
i
on
pr
oc
e
s
s
f
o
r
a
r
e
t
ur
n
t
o
t
he
r
oot
w
or
d.
E
m
be
ddi
ng
i
s
di
vi
de
d
i
n
t
o
2
s
t
a
ge
s
,
na
m
e
l
y
,
t
oke
ni
z
a
t
i
on
, w
hi
c
h
i
s
c
a
r
r
i
e
d out
by t
he
B
E
R
T
or
S
R
E
R
T
t
oke
ni
z
e
r
. A
l
t
hough t
he
pr
oc
e
s
s
r
e
m
a
i
ns
t
h
e
s
a
m
e
,
onl
y
t
he
pr
oc
e
s
s
i
s
d
i
r
e
c
t
l
y
pe
r
f
or
m
e
d
by t
he
B
E
R
T
or
S
R
E
R
T
m
ode
l
.
S
t
a
ge
2
w
i
t
h W
or
dP
i
e
c
e
t
oke
ni
z
a
t
i
on.
F
ur
t
he
r
t
oke
ni
z
a
t
i
on
u
s
e
s
t
h
e
B
E
R
T
voc
a
b
by
r
e
f
e
r
r
i
ng
t
o a
bout
30,000
s
ubw
or
ds
.
I
f
t
he
w
or
d
i
s
no
t
pr
e
s
e
nt
i
n t
he
voc
a
b,
i
t
i
s
br
oke
n dow
n i
n
t
o s
ubw
or
d pi
e
c
e
s
w
i
t
h
t
he
## pr
e
f
i
x.
3.2.
E
m
b
e
d
d
i
n
g
p
i
p
e
l
i
n
e
s
an
d
c
l
u
s
t
e
r
i
n
g c
on
f
i
gu
r
at
i
on
s
A
bs
t
r
a
c
t
da
t
a
m
us
t
go
t
h
r
ough
t
h
e
pr
e
-
pr
oc
e
s
s
i
ng
s
t
a
ge
s
de
s
c
r
i
be
d
e
a
r
l
i
e
r
.
I
n
t
he
t
r
a
d
i
t
i
ona
l
a
ppr
oa
c
h,
i
t
i
s
qui
t
e
l
ong
f
r
o
m
t
he
t
ok
e
ni
z
a
t
i
on
s
t
a
ge
,
t
he
r
e
m
ova
l
of
s
t
opw
or
ds
,
a
nd
s
t
e
m
m
i
ng
be
f
o
r
e
t
he
pr
oc
e
s
s
c
ont
i
nue
s
on
doc
um
e
nt
c
l
us
t
e
r
i
ng.
M
e
a
nw
hi
l
e
,
on
t
he
s
t
a
ge
w
i
t
h
a
t
r
a
ns
f
o
r
m
e
r
-
ba
s
e
d
e
m
be
ddi
ng
m
ode
l
, a
f
t
e
r
t
oke
ni
z
a
t
i
on,
t
he
ne
xt
s
t
e
p i
s
a
l
i
gh
t
e
r
pr
e
-
pr
oc
e
s
s
i
ng w
i
t
h l
ow
e
r
c
a
s
e
nor
m
a
l
i
z
a
t
i
on a
nd c
ha
r
a
c
t
e
r
r
e
m
ova
l
s
t
a
ge
s
by
a
ppl
y
i
ng
W
or
dP
i
e
c
e
.
T
he
r
e
s
ul
t
w
i
l
l
be
f
or
m
e
d
f
r
om
t
he
out
put
of
t
he
l
a
s
t
hi
dde
n
s
t
a
t
e
w
i
t
h m
e
a
n poo
l
i
ng, w
hi
c
h
t
he
n ne
e
ds
t
o be
nor
m
a
l
i
z
e
d w
i
t
h L
2.
3.3.
C
l
u
s
t
e
r
vi
s
u
a
l
i
z
a
t
i
on
T
he
c
l
us
t
e
r
i
ng
r
e
s
u
l
t
s
ob
t
a
i
n
e
d
w
i
t
h
a
ggl
om
e
r
a
t
i
ve
c
l
us
t
e
r
i
ng
s
how
t
ha
t
t
he
do
c
um
e
n
t
on
t
h
e
l
e
f
t
s
i
de
f
or
m
s
a
s
m
a
l
l
c
l
us
t
e
r
a
nd i
s
l
ow
i
n s
pa
c
i
ng. T
hi
s
s
how
s
hi
gh uni
f
or
m
i
t
y. O
ve
r
a
l
l
, t
h
i
s
s
a
l
t
s
how
s
t
ha
t
t
he
da
t
a
s
e
t
c
a
n
be
gr
oupe
d
i
nt
o
s
e
ve
r
a
l
m
a
i
n
gr
oups
.
T
hi
s
di
v
i
s
i
on
i
s
qui
t
e
c
l
e
a
r
,
s
how
i
ng
t
ha
t
t
he
r
e
a
r
e
di
f
f
e
r
e
nc
e
s
i
n c
l
us
t
e
r
s
. T
he
r
e
s
u
l
t
s
o
f
t
he
de
nd
r
ogr
a
m
a
r
e
s
how
n
in
F
i
gur
e
2.
I
n
a
ddi
t
i
on
t
o
goi
ng
t
hr
ough
t
he
de
ndr
og
r
a
m
,
a
no
t
he
r
gr
a
ph
i
s
a
ke
yw
or
d
he
a
t
m
a
p.
F
r
om
t
he
ke
yw
or
d
c
l
us
t
e
r
he
a
t
m
a
p,
i
t
i
s
s
how
n
t
ha
t
s
e
ve
r
a
l
ke
yw
or
ds
f
or
e
a
c
h
c
l
u
s
t
e
r
,
b
a
s
e
d
on
t
he
T
F
-
I
D
F
m
a
ppi
ng
,
s
how
t
he
d
i
s
t
r
i
but
i
on
o
f
w
or
ds
i
n t
he
a
bs
t
r
a
c
t
doc
um
e
n
t
on
t
he
l
e
f
t
.
T
he
c
ol
o
r
on t
he
he
a
t
m
a
p
a
l
s
o
s
how
s
t
he
f
r
e
que
nc
y
of
w
or
ds
i
n
t
h
e
da
t
a
s
e
t
;
t
he
da
r
ke
r
t
he
c
o
l
or
,
t
he
m
or
e
dom
i
na
nt
t
h
e
w
or
d
t
ha
t
a
ppe
a
r
s
i
n
t
he
gr
oupi
ng
o
f
a
bs
t
r
a
c
t
doc
u
m
e
nt
s
.
T
hi
s
w
or
d
r
e
f
e
r
s
t
o
t
he
I
ndone
s
i
a
n
w
or
d
t
ha
t
a
ppe
a
r
s
on
s
t
ud
e
nt
a
bs
t
r
a
c
t
doc
um
e
nt
s
,
s
u
c
h
a
s
s
e
r
ve
r
s
a
nd
w
e
bs
i
t
e
s
t
ha
t
a
ppe
a
r
qui
t
e
of
t
e
n.
T
hi
s
he
a
t
m
a
p
i
s
e
xp
e
c
t
e
d
t
o
be
a
b
l
e
t
o
pr
ovi
de
a
n
ove
r
vi
e
w
of
t
he
t
h
e
m
a
t
i
c
unde
r
s
t
a
ndi
ng
o
f
e
a
c
h
ke
yw
or
d
di
s
t
r
i
but
i
on
i
n
c
l
us
t
e
r
i
ng
,
de
t
a
i
l
s
on F
i
gur
e
3.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
dv A
ppl
S
c
i
I
S
S
N
:
2252
-
8814
Se
m
ant
i
c
c
l
us
t
e
r
i
ng of
s
c
i
e
nt
i
f
i
c
abs
t
r
ac
t
s
w
i
t
h
t
r
ans
f
o
r
m
e
r
e
m
be
ddi
ngs
…
(
M
us
t
hof
a G
al
i
h P
r
adana
)
535
I
n
a
ddi
t
i
on
t
o
goi
ng
t
hr
ough
t
he
de
ndr
og
r
a
m
,
a
no
t
he
r
gr
a
ph
i
s
a
ke
yw
or
d
he
a
t
m
a
p.
F
r
om
t
he
ke
yw
or
d
c
l
us
t
e
r
he
a
t
m
a
p,
i
t
i
s
s
how
n
t
ha
t
s
e
ve
r
a
l
ke
yw
or
ds
f
or
e
a
c
h
c
l
u
s
t
e
r
,
b
a
s
e
d
on
t
he
T
F
-
I
D
F
m
a
ppi
ng,
s
how
t
he
d
i
s
t
r
i
but
i
on
o
f
w
or
ds
i
n t
he
a
bs
t
r
a
c
t
doc
um
e
n
t
on
t
he
l
e
f
t
.
T
he
c
ol
o
r
on t
he
he
a
t
m
a
p
a
l
s
o
s
how
s
t
he
f
r
e
que
nc
y
of
w
or
ds
i
n
t
he
da
t
a
s
e
t
.
T
he
da
r
ke
r
t
he
c
o
l
or
,
t
he
m
o
r
e
dom
i
na
nt
t
he
w
or
d
t
h
a
t
a
ppe
a
r
s
i
n
t
he
gr
oupi
ng
o
f
a
bs
t
r
a
c
t
doc
u
m
e
nt
s
.
T
hi
s
w
or
d
r
e
f
e
r
s
t
o
t
he
I
ndone
s
i
a
n
w
or
d
t
ha
t
a
ppe
a
r
s
on
s
t
ud
e
nt
a
bs
t
r
a
c
t
doc
um
e
nt
s
,
s
u
c
h
a
s
s
e
r
ve
r
s
a
nd
w
e
bs
i
t
e
s
t
ha
t
a
ppe
a
r
qui
t
e
of
t
e
n.
T
hi
s
he
a
t
m
a
p
i
s
e
xp
e
c
t
e
d
t
o
be
a
b
l
e
t
o
pr
ovi
de
a
n
ove
r
v
i
e
w
of
t
he
t
he
m
a
t
i
c
unde
r
s
t
a
nd
i
ng
o
f
e
a
c
h
ke
yw
or
d
d
i
s
t
r
i
bu
t
i
on
i
n
c
l
us
t
e
r
i
ng.
T
hi
s
s
e
pa
r
a
t
i
on
i
s
ba
s
e
d
on
t
he
s
e
m
a
nt
i
c
r
e
l
a
t
i
ons
h
i
p
be
t
w
e
e
n
w
or
ds
[
37]
,
no
t
onl
y
ba
s
e
d
on
t
h
e
f
r
e
que
nc
y
of
oc
c
ur
r
e
nc
e
[
38]
.
F
i
gur
e
2.
D
e
ndr
ogr
a
m
c
l
us
t
e
r
i
ng
F
i
gur
e
3. H
e
a
t
m
a
p
ke
yw
or
d
3.4.
T
e
s
t
s
c
e
n
ar
i
os
T
he
c
l
us
t
e
r
i
ng m
ode
l
s
t
ha
t
w
i
l
l
be
us
e
d
a
r
e
k
-
m
e
a
n
s
a
nd
a
ggl
om
e
r
a
t
i
ve
by
a
pp
l
yi
ng
t
he
B
E
R
T
a
nd
S
B
E
R
T
e
m
be
ddi
ng
m
ode
l
s
,
a
s
w
e
l
l
a
s
a
t
r
a
di
t
i
ona
l
T
F
-
I
D
F
m
ode
l
.
R
e
s
ul
t
s
w
i
l
l
be
pr
e
s
e
nt
e
d
ba
s
e
d
on
s
c
e
na
r
i
os
w
i
t
h
t
he
de
t
e
r
m
i
na
t
i
on
of
t
he
num
be
r
of
c
l
us
t
e
r
s
,
s
t
a
r
t
i
ng
f
r
om
2
t
o
6
,
t
o
f
i
nd
t
he
opt
i
m
a
l
c
l
us
t
e
r
.
R
e
f
e
r
s
t
o
t
he
t
e
s
t
r
e
s
ul
t
s
i
n
bot
h
s
c
e
na
r
i
os
f
o
r
t
r
a
d
i
t
i
ona
l
m
ode
l
s
a
nd
e
m
be
ddi
ng
m
ode
l
s
w
i
t
h
2
c
l
us
t
e
r
i
ng
a
l
gor
i
t
hm
s
.
S
o,
t
he
m
os
t
opt
i
m
a
l
va
l
ue
obt
a
i
ne
d i
s
t
o us
e
B
E
R
T
a
nd
k
-
m
e
a
ns
w
i
t
h a
num
be
r
of
c
l
us
t
e
r
s
of
2.
T
hi
s
i
s
s
how
n
b
y
t
he
s
i
l
houe
t
t
e
s
c
or
e
va
l
ue
of
0.2592
be
i
ng
t
he
hi
ghe
s
t
,
a
nd
t
he
D
a
vi
e
s
-
B
oul
di
n
i
nde
x
va
l
ue
of
2.5416
,
w
i
t
h
a
f
a
i
r
l
y
l
ow
va
l
ue
or
a
n
opt
i
m
a
l
va
l
ue
.
T
hi
s
s
t
a
t
e
s
t
ha
t
B
E
R
T
i
s
a
bl
e
t
o
c
a
pt
ur
e
s
e
m
a
nt
i
c
r
e
l
a
t
i
ons
hi
ps
be
t
w
e
e
n
s
e
nt
e
nc
e
s
be
t
t
e
r
[
39]
.
T
he
r
e
s
ul
t
s
of
t
he
vi
s
ua
l
i
z
a
t
i
on
of
t
he
c
om
pa
r
i
s
on
of
t
he
s
e
t
w
o
m
ode
l
s
a
r
e
s
how
n
i
n
F
i
gur
e
4
,
w
he
r
e
F
i
gur
e
4(
a
)
s
how
s
t
he
s
i
l
houe
t
t
e
a
nd
F
i
gur
e
4(
b)
s
how
s
t
he
D
a
vi
e
s
–
B
oul
di
n i
nde
x
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8814
I
nt
J
A
dv A
ppl
S
c
i
,
V
ol
. 15, N
o. 2, J
une
2026
:
532
-
540
536
(
a
)
(
b)
F
i
gur
e
4. C
om
pa
r
i
s
on m
od
e
l
of
(
a
)
s
i
l
houe
t
t
e
s
c
or
e
a
nd
(
b)
D
a
vi
e
s
–
B
ou
l
di
n i
nd
e
x
s
c
or
e
A
s
f
or
t
he
ove
r
a
l
l
s
c
e
n
a
r
i
o
w
i
t
hou
t
c
a
t
e
gor
i
z
a
t
i
on,
i
t
i
s
s
how
n
i
n
F
i
gur
e
5.
F
i
gur
e
5
s
how
s
t
he
r
e
s
u
l
t
of
t
he
s
i
l
houe
t
t
e
s
c
or
e
, w
hi
l
e
F
i
gur
e
6 e
xp
l
a
i
ns
t
he
D
a
vi
e
s
-
B
ou
l
di
n i
nde
x s
c
or
e
. T
he
s
e
t
w
o m
e
t
r
i
c
s
ne
e
d t
o be
i
nt
e
r
p
r
e
t
e
d
i
n m
o
r
e
de
pt
h.
F
i
gur
e
5. A
l
l
s
c
e
na
r
i
o s
i
l
houe
t
t
e
s
c
o
r
e
F
i
gur
e
6. A
l
l
s
c
e
na
r
i
o
D
a
vi
e
s
-
B
ou
l
di
n
i
nd
e
x s
c
or
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
dv A
ppl
S
c
i
I
S
S
N
:
2252
-
8814
Se
m
ant
i
c
c
l
us
t
e
r
i
ng of
s
c
i
e
nt
i
f
i
c
abs
t
r
ac
t
s
w
i
t
h
t
r
ans
f
o
r
m
e
r
e
m
be
ddi
ngs
…
(
M
us
t
hof
a G
al
i
h P
r
adana
)
537
T
he
opt
i
m
a
l
nu
m
be
r
o
f
c
l
us
t
e
r
s
i
s
i
n
t
he
2nd
c
l
us
t
e
r
.
M
or
e
num
be
r
s
l
ow
e
r
t
he
qua
l
i
t
y
of
t
he
c
l
us
t
e
r
.
T
he
m
os
t
opt
i
m
a
l
r
e
pr
e
s
e
n
t
a
t
i
on
m
ode
l
i
s
t
he
e
m
b
e
ddi
ng
m
ode
l
w
i
t
h
B
E
R
T
a
s
t
he
t
e
x
t
r
e
pr
e
s
e
nt
a
t
i
on
f
or
s
c
i
e
nt
i
f
i
c
a
bs
t
r
a
c
t
c
l
u
s
t
e
r
i
ng
,
c
o
m
bi
ne
d
w
i
t
h
k
-
m
e
a
ns
,
w
hi
c
h
i
s
ove
r
a
l
l
m
o
r
e
s
t
a
bl
e
t
ha
n
a
gg
l
om
e
r
a
t
i
ve
i
n
c
om
bi
na
t
i
on
w
i
t
h
B
E
R
T
e
m
be
ddi
ng.
T
he
S
B
E
R
T
e
m
be
ddi
ng
m
ode
l
ne
e
d
s
f
i
ne
-
t
un
i
ng
i
f
i
t
i
s
t
o
be
us
e
d
i
n
t
he
dom
a
i
n
of
s
c
i
e
nt
i
f
i
c
publ
i
c
a
t
i
ons
(
e
s
pe
c
i
a
l
l
y
i
n
I
ndone
s
i
a
n
or
m
ul
t
i
d
i
s
c
i
p
l
i
na
r
y)
[
40]
.
T
he
t
r
a
d
i
t
i
ona
l
TF
-
I
D
F
m
ode
l
i
s
not
r
e
c
o
m
m
e
nde
d f
or
t
e
xt
r
e
pr
e
s
e
nt
a
t
i
on i
n
c
ont
e
xt
ua
l
c
l
us
t
e
r
i
ng
,
s
uc
h a
s
s
c
i
e
nt
i
f
i
c
j
ou
r
na
l
a
bs
t
r
a
c
t
s
[
41]
. T
he
m
ode
l
a
l
s
o s
how
s
t
ha
t
s
e
m
a
nt
i
c
c
l
us
t
e
r
i
ng i
s
not
onl
y t
e
c
hn
i
c
a
l
l
y us
e
f
ul
but
c
a
n a
l
s
o ha
ve
a
pr
a
c
t
i
c
a
l
i
m
p
a
c
t
.
T
he
p
r
a
c
t
i
c
a
l
i
m
p
a
c
t
i
s
t
o
be
a
b
l
e
t
o s
uppor
t
r
e
s
e
a
r
c
h
di
s
c
ove
r
y, t
he
m
a
t
i
c
m
a
pp
i
ng,
a
nd
a
m
or
e
e
f
f
i
c
i
e
nt
l
i
t
e
r
a
t
ur
e
s
yn
t
he
s
i
s
pr
oc
e
s
s
.
T
hi
s
pot
e
nt
i
a
l
c
a
n be
e
xp
l
or
e
d f
u
r
t
he
r
i
n a
di
g
i
t
a
l
l
i
b
r
a
r
y or
i
n a
r
e
s
e
a
r
c
h da
s
hboa
r
d. T
hi
s
i
s
e
xpe
c
t
e
d
t
o
be
a
bl
e
t
o
be
de
ve
l
ope
d i
n
m
ul
t
i
p
l
e
l
a
ngua
ge
s
a
nd
i
n
t
e
gr
a
t
e
d
w
i
t
h
va
r
i
ous
ot
he
r
da
t
a
s
ou
r
c
e
s
.
T
he
r
e
s
u
l
t
s
of
t
hi
s
s
t
udy
s
how
t
ha
t
s
e
m
a
nt
i
c
g
r
oupi
ng i
s
no
t
onl
y t
e
c
hn
i
c
a
l
l
y
be
ne
f
i
c
i
a
l
f
or
i
m
pr
ov
i
ng
c
l
us
t
e
r
qua
l
i
t
y
but
a
l
s
o
ha
s
a
s
i
gni
f
i
c
a
nt
pr
a
c
t
i
c
a
l
i
m
pa
c
t
.
T
he
c
l
u
s
t
e
r
s
f
or
m
e
d
c
a
n
s
uppor
t
r
e
s
e
a
r
c
h
di
s
c
ove
r
y,
t
he
m
a
t
i
c
m
a
ppi
ng,
a
nd
m
or
e
e
f
f
i
c
i
e
n
t
a
ut
o
m
a
t
e
d
l
i
t
e
r
a
t
u
r
e
s
ynt
he
s
i
s
.
W
i
t
h
t
he
po
t
e
nt
i
a
l
f
o
r
i
nt
e
g
r
a
t
i
on
i
n
t
o
di
gi
t
a
l
l
i
br
a
r
i
e
s
,
c
i
t
a
t
i
on
ne
t
w
or
ks
,
or
r
e
s
e
a
r
c
h
da
s
hbo
a
r
ds
,
t
h
i
s
a
ppr
o
a
c
h
c
ont
r
i
but
e
s
t
o
a
c
c
e
l
e
r
a
t
i
ng
a
c
c
e
s
s
t
o
a
nd
unde
r
s
t
a
nd
i
ng
of
s
c
i
e
nt
i
f
i
c
l
i
t
e
r
a
t
ur
e
.
G
oi
ng
f
or
w
a
r
d,
r
e
s
e
a
r
c
h
d
i
r
e
c
t
i
ons
c
a
n
be
de
ve
l
ope
d
t
hr
ough
m
ul
t
i
l
i
ngua
l
gr
oup
i
ng, dyna
m
i
c
t
op
i
c
e
vol
u
t
i
on
t
r
a
c
k
i
ng, a
nd i
nt
e
g
r
a
t
i
on w
i
t
h c
i
t
a
t
i
on
-
b
a
s
e
d m
e
t
r
i
c
s
t
o e
n
r
i
c
h
a
na
l
ys
i
s
a
nd e
xpa
nd t
h
e
s
c
ope
of
a
pp
l
i
c
a
t
i
on
s
.
4.
C
O
N
C
L
U
S
I
O
N
T
he
r
e
s
ul
t
s
s
how
e
d
t
h
a
t
t
he
m
os
t
opt
i
m
a
l
m
ode
l
w
a
s
t
he
e
m
be
ddi
ng
m
ode
l
w
i
t
h
t
he
be
s
t
r
e
s
u
l
t
s
i
n
B
E
R
T
a
nd
k
-
m
e
a
n
s
,
w
i
t
h
t
he
m
os
t
opt
i
m
a
l
num
b
e
r
of
c
l
us
t
e
r
s
i
n
c
l
us
t
e
r
2
.
T
hi
s
r
e
s
ul
t
w
a
s
s
t
r
e
ngt
h
e
ne
d
by
t
he
r
e
s
ul
t
s
of
t
he
l
a
r
ge
s
t
s
i
l
houe
t
t
e
s
c
or
e
a
nd
t
h
e
s
m
a
l
l
e
s
t
D
a
vi
e
s
-
B
oul
di
n
i
nde
x
v
a
l
ue
.
T
he
s
e
f
i
ndi
ngs
s
t
a
t
e
t
ha
t
t
r
a
n
s
f
or
m
e
r
r
e
pr
e
s
e
nt
a
t
i
on
s
c
a
n
pr
oduc
e
opt
i
m
a
l
c
l
us
t
e
r
s
a
nd
c
a
n
be
u
s
e
d
t
o
s
uppo
r
t
f
ut
u
r
e
r
e
s
e
a
r
c
h
on
t
opi
c
m
a
ppi
ng.
F
U
N
D
I
N
G
I
N
F
O
R
M
A
T
I
O
N
T
he
a
ut
hor
s
s
t
a
t
e
t
ha
t
no
f
und
i
ng
w
a
s
i
nvol
ve
d
i
n
t
he
de
ve
l
opm
e
n
t
of
t
h
i
s
r
e
s
e
a
r
c
h.
T
he
w
or
k
w
a
s
c
onduc
t
e
d i
nde
p
e
nde
nt
l
y w
i
t
hout
f
i
na
nc
i
a
l
s
uppor
t
f
r
o
m
a
ny e
xt
e
r
na
l
a
g
e
nc
y, gr
a
n
t
, or
i
n
s
t
i
t
ut
i
ona
l
f
und.
A
U
T
H
O
R
C
O
N
T
R
I
B
U
T
I
O
N
S
S
T
A
T
E
M
E
N
T
T
hi
s
j
our
n
a
l
u
s
e
s
t
he
C
ont
r
i
but
o
r
R
ol
e
s
T
a
xonom
y
(
C
R
e
di
T
)
t
o
r
e
c
ogn
i
z
e
i
nd
i
vi
dua
l
a
u
t
hor
c
ont
r
i
bu
t
i
ons
,
r
e
duc
e
a
u
t
hor
s
h
i
p di
s
pu
t
e
s
, a
nd
f
a
c
i
l
i
t
a
t
e
c
o
l
l
a
bo
r
a
t
i
on.
N
am
e
of
A
u
t
h
or
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
M
us
t
ho
f
a
G
a
l
i
h P
r
a
da
na
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
P
uj
o H
a
r
i
S
a
put
r
o
✓
✓
✓
✓
A
r
dhya
ns
ya
h M
ua
l
o
✓
✓
✓
✓
A
r
bi
a
t
i
F
a
i
z
a
h
✓
✓
✓
✓
W
a
hyun
i
F
i
t
hr
a
t
ul
Z
a
l
m
i
✓
✓
✓
✓
✓
✓
C
:
C
onc
e
pt
ua
l
i
z
a
t
i
on
M
:
M
e
t
ho
d
ol
ogy
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
d
a
t
i
on
Fo
:
Fo
r
m
a
l
a
na
l
ys
i
s
I
:
I
nve
s
t
i
ga
t
i
on
R
:
R
e
s
our
c
e
s
D
:
D
a
t
a
C
ur
a
t
i
on
O
:
W
r
i
t
i
ng
-
O
r
i
g
i
na
l
D
r
a
f
t
E
:
W
r
i
t
i
ng
-
R
e
vi
e
w
&
E
d
i
t
i
ng
Vi
:
Vi
s
ua
l
i
z
a
t
i
on
Su
:
Su
pe
r
vi
s
i
on
P
:
P
r
oj
e
c
t
a
d
m
i
ni
s
t
r
a
t
i
on
Fu
:
Fu
nd
i
ng a
c
qui
s
i
t
i
on
C
O
N
F
L
I
C
T
O
F
I
N
T
E
R
E
S
T
S
T
A
T
E
M
E
N
T
T
he
a
ut
hor
s
s
t
a
t
e
no
c
onf
l
i
c
t
o
f
i
n
t
e
r
e
s
t
.
I
N
F
O
R
M
E
D
C
O
N
S
E
N
T
N
ot
a
ppl
i
c
a
bl
e
.
T
hi
s
s
t
udy
d
i
d
no
t
i
nvo
l
ve
hu
m
a
n
pa
r
t
i
c
i
pa
n
t
s
,
hu
m
a
n
da
t
a
,
o
r
a
ny
pe
r
s
ona
l
l
y
i
de
nt
i
f
i
a
b
l
e
i
nf
o
r
m
a
t
i
on.
A
l
l
da
t
a
us
e
d
w
e
r
e
e
i
t
he
r
publ
i
c
l
y
a
va
i
l
a
bl
e
,
f
u
l
l
y
a
nonym
i
z
e
d,
or
de
r
i
v
e
d
f
r
om
non‑
hum
a
n s
our
c
e
s
, a
nd t
h
e
r
e
f
o
r
e
no i
n
f
or
m
e
d c
ons
e
nt
w
a
s
r
e
qui
r
e
d f
r
o
m
i
nd
i
vi
dua
l
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8814
I
nt
J
A
dv A
ppl
S
c
i
,
V
ol
. 15, N
o. 2, J
une
2026
:
532
-
540
538
E
T
H
I
C
A
L
A
P
P
R
O
V
A
L
N
ot
a
ppl
i
c
a
bl
e
.
T
hi
s
r
e
s
e
a
r
c
h
di
d
not
i
nvo
l
ve
hum
a
n
s
ub
j
e
c
t
s
,
hum
a
n
bi
o
l
ogi
c
a
l
m
a
t
e
r
i
a
l
s
,
or
e
xpe
r
i
m
e
nt
a
l
pr
oc
e
dur
e
s
on
a
n
i
m
a
l
s
.
T
he
w
or
k
w
a
s
c
onduc
t
e
d
s
o
l
e
l
y
on
c
o
m
put
a
t
i
ona
l
m
ode
l
s
,
publ
i
c
l
y
a
va
i
l
a
b
l
e
da
t
a
s
e
t
s
,
or
non
‑
s
e
ns
i
t
i
ve
da
t
a
t
ha
t
di
d
not
r
e
qui
r
e
i
n
t
e
r
ve
n
t
i
on
w
i
t
h
l
i
vi
ng
o
r
ga
ni
s
m
s
.
T
he
r
e
f
or
e
,
e
t
hi
c
a
l
a
pp
r
ova
l
f
r
om
a
n
i
n
s
t
i
t
ut
i
ona
l
r
e
vi
e
w
boa
r
d
or
a
n
i
m
a
l
e
t
hi
c
s
c
om
m
i
t
t
e
e
w
a
s
no
t
ne
c
e
s
s
a
r
y
f
or
t
h
i
s
s
t
udy.
D
A
T
A
A
V
A
I
L
A
B
I
L
I
T
Y
T
he
da
t
a
t
ha
t
s
uppor
t
t
he
f
i
ndi
ngs
of
t
h
i
s
s
t
udy a
r
e
a
va
i
l
a
b
l
e
f
r
o
m
t
he
c
or
r
e
s
pond
i
ng a
ut
hor
,
[
M
G
P
]
,
upon r
e
a
s
ona
bl
e
r
e
que
s
t
.
R
E
F
E
R
E
N
C
E
S
[
1]
A
.
K
ul
ka
r
ni
a
nd
A
.
S
hi
va
na
nd
a
,
N
at
ur
al
l
anguage
pr
oc
e
s
s
i
ng
r
e
c
i
pe
s
:
unl
oc
k
i
ng
t
e
x
t
dat
a
w
i
t
h
m
ac
hi
ne
l
e
a
r
ni
ng
and
de
e
p
l
e
ar
ni
ng us
i
ng P
y
t
hon
. B
e
r
ke
l
e
y,
C
a
l
i
f
or
ni
a
:
A
pr
e
s
s
, 2021
, d
oi
:
10.1007/
978
-
1
-
4842
-
7351
-
7.
[
2]
R
. S
. T
. L
e
e
,
N
at
ur
al
l
anguage
pr
oc
e
s
s
i
ng:
a t
e
x
t
book
w
i
t
h
P
y
t
hon
i
m
pl
e
m
e
nt
at
i
on
. S
i
nga
po
r
e
:
S
pr
i
nge
r
, 2023
,
d
oi
:
10.1007/
978
-
981
-
99
-
1999
-
4.
[
3]
S
.
K
ha
n,
M
.
A
nw
a
r
,
H
.
Q
a
yyum
,
F
.
A
l
i
,
a
nd
M
. N
a
w
a
z
,
“
F
a
ke
ne
w
s
c
l
a
s
s
i
f
i
c
a
t
i
on
us
i
ng
m
a
c
hi
ne
l
e
a
r
n
i
ng:
c
ount
ve
c
t
or
i
z
e
r
a
nd
s
uppor
t
ve
c
t
or
m
a
c
h
i
ne
,”
J
ou
r
nal
o
f
C
om
put
i
ng
and
B
i
om
e
di
c
a
l
I
n
f
or
m
at
i
c
s
,
vo
l
.
4,
no.
1,
pp.
54
–
63,
J
a
n.
2022,
d
oi
:
10.56979/
401/
2022/
85.
[
4]
M
.
M
.
D
a
ny
a
l
,
S
.
S
.
K
ha
n,
M
.
K
ha
n,
S
.
U
l
l
a
h,
M
.
B
.
G
ha
f
f
a
r
,
a
n
d
W
.
K
ha
n,
“
S
e
n
t
i
m
e
nt
a
na
l
ys
i
s
of
m
ovi
e
r
e
vi
e
w
s
ba
s
e
d
on
N
B
a
ppr
oa
c
he
s
us
i
n
g T
F
–
I
D
F
a
n
d
c
oun
t
ve
c
t
or
i
z
e
r
,”
Soc
i
al
N
e
t
w
or
k
A
nal
y
s
i
s
and M
i
ni
ng
, vol
. 1
4, no
. 1,
A
pr
.
2024,
d
oi
:
10.1
007/
s
132
78
-
024
-
01250
-
9.
[
5]
G
. M
. R
a
z
a
, Z
. S
. B
ut
t
, S
. L
a
t
i
f
, a
n
d
A
. W
a
hi
d
, “
S
e
nt
i
m
e
nt
a
na
l
ys
i
s
on
C
O
V
I
D
T
w
e
e
t
s
:
a
n e
xpe
r
i
m
e
nt
a
l
a
na
l
y
s
i
s
on
t
he
i
m
pa
c
t
of
c
ount
ve
c
t
or
i
z
e
r
a
nd
T
F
-
I
D
F
on
s
e
nt
i
m
e
nt
pr
e
d
i
c
t
i
ons
us
i
ng d
e
e
p
l
e
a
r
ni
ng
m
od
e
l
s
,”
i
n
2021
I
nt
e
r
nat
i
onal
C
onf
e
r
e
nc
e
on
D
i
gi
t
al
F
ut
ur
e
s
and T
r
ans
f
or
m
at
i
v
e
T
e
c
hnol
ogi
e
s
, I
C
oD
T
2 2021
, M
a
y 2021, pp. 1
–
6
, d
oi
:
10.1109/
I
C
oD
T
252288.2021.9441508.
[
6]
H
.
D
.
A
buba
ka
r
a
n
d
M
.
U
m
a
r
,
“
S
e
nt
i
m
e
nt
c
l
a
s
s
i
f
i
c
a
t
i
on:
r
e
v
i
e
w
of
t
e
xt
ve
c
t
or
i
z
a
t
i
on
m
e
t
ho
d
s
:
ba
g
o
f
w
or
d
s
,
TF
-
I
D
F
,
W
or
d
2ve
c
a
nd
D
oc
2ve
c
,”
SL
U
J
our
nal
of
Sc
i
e
nc
e
and T
e
c
hnol
ogy
, vo
l
. 4, no. 1&
2, pp. 27
–
33, A
ug. 2022, d
oi
:
10.56471/
s
l
uj
s
t
.v4i
.266.
[
7]
M
.
R
.
A
.
P
r
a
s
e
t
ya
a
n
d
A
.
M
.
P
r
i
ya
t
no,
“
D
i
c
e
s
i
m
i
l
a
r
i
t
y
a
nd
T
F
-
I
D
F
f
or
ne
w
s
t
u
d
e
nt
a
d
m
i
s
s
i
ons
c
ha
t
b
ot
,”
R
I
G
G
S:
J
our
nal
of
A
r
t
i
f
i
c
i
a
l
I
n
t
e
l
l
i
ge
nc
e
and D
i
g
i
t
a
l
B
us
i
ne
s
s
, vol
. 1, no. 1, pp. 13
–
18, J
u
l
. 2022, d
oi
:
10.31004/
r
i
ggs
.v1
i
1.5.
[
8]
D
. M
e
i
d
e
l
f
i
, Y
ul
he
r
ni
w
a
t
i
, I
. R
a
hm
a
yuni
, T
. H
i
d
a
ya
t
, a
nd
D
.
C
ha
nd
r
a
, “
T
F
-
I
D
F
i
m
pl
e
m
e
nt
a
t
i
on f
or
s
i
m
i
l
a
r
i
t
y c
he
c
ke
r
on t
he
f
i
na
l
pr
oj
e
c
t
t
i
t
l
e
,”
I
nt
e
r
na
t
i
ona
l
J
our
na
l
o
f
A
dv
anc
e
d
Sc
i
e
nc
e
C
om
put
i
ng
and
E
ngi
ne
e
r
i
ng
,
vo
l
.
3,
no.
1,
pp.
40
–
52,
2021,
d
oi
:
10.30630/
i
j
a
s
c
e
.3.1.3.
[
9]
G
.
Y
una
nd
a
,
D
.
N
ur
j
a
na
h,
a
nd
S
.
M
e
l
i
a
na
,
“
R
e
c
o
m
m
e
nd
a
t
i
on
s
ys
t
e
m
f
r
om
M
i
c
r
os
of
t
N
e
w
s
d
a
t
a
us
i
ng
T
F
-
I
D
F
a
nd
c
os
i
ne
s
i
m
i
l
a
r
i
t
y
m
e
t
ho
d
s
,”
B
ui
l
di
ng
of
I
nf
or
m
at
i
c
s
,
T
e
c
hnol
ogy
and
Sc
i
e
n
c
e
,
vol
.
4,
no.
1,
pp.
277
–
284,
2022,
d
oi
:
10.47065/
bi
t
s
.v4i
1.1670.
[
10]
F
.
L
a
n,
“
R
e
s
e
a
r
c
h
on
t
e
x
t
s
i
m
i
l
a
r
i
t
y
m
e
a
s
ur
e
m
e
nt
hybr
i
d
a
l
gor
i
t
hm
w
i
t
h
t
e
r
m
s
e
m
a
nt
i
c
i
nf
or
m
a
t
i
on
a
nd
T
F
-
I
D
F
m
e
t
ho
d
,”
A
dv
anc
e
s
i
n M
u
l
t
i
m
e
d
i
a
, vol
. 2022, pp. 1
–
11, A
pr
. 2022, d
oi
:
10.1155/
2022
/
7923262.
[
11]
A
.
W
e
nd
l
a
nd
,
M
.
Z
e
ne
r
e
,
a
nd
J
.
N
i
e
m
a
nn,
“
I
nt
r
od
uc
t
i
on
t
o
t
e
xt
c
l
a
s
s
i
f
i
c
a
t
i
on:
i
m
pa
c
t
of
s
t
e
m
m
i
ng
a
nd
c
om
pa
r
i
n
g
T
F
-
I
D
F
a
nd
c
ount
ve
c
t
or
i
z
a
t
i
on
a
s
f
e
a
t
ur
e
e
xt
r
a
c
t
i
on
t
e
c
hni
que
s
,”
i
n
C
om
m
uni
c
at
i
ons
i
n
C
om
put
e
r
and
I
nf
or
m
at
i
on
Sc
i
e
nc
e
,
2021,
vol
.
1442,
pp. 289
–
300
, d
oi
:
10.1007/
978
-
3
-
030
-
85521
-
5_19.
[
12]
A
.
G
up
t
a
a
n
d
U
.
S
ha
r
m
a
,
“
M
a
c
hi
ne
l
e
a
r
n
i
ng
-
ba
s
e
d
a
s
pe
c
t
c
a
t
e
gor
y
d
e
t
e
c
t
i
on
f
or
H
i
nd
i
d
a
t
a
us
i
ng
TF
-
I
D
F
a
n
d
c
oun
t
ve
c
t
or
i
z
a
t
i
on,”
i
n
2nd
I
E
E
E
I
nt
e
r
na
t
i
ona
l
C
onf
e
r
e
n
c
e
on
D
e
v
i
c
e
I
nt
e
l
l
i
ge
nc
e
,
C
om
put
i
ng
and
C
om
m
uni
c
at
i
on
T
e
c
hno
l
ogi
e
s
,
D
I
C
C
T
2024
, M
a
r
. 2024, pp. 39
–
44
, d
oi
:
10.1109/
D
I
C
C
T
61038.2024.10532960.
[
13]
M
.
S
i
ngha
l
,
N
.
S
i
ngha
l
,
S
.
K
he
r
a
,
A
.
U
pm
a
nyu,
a
n
d
P
.
N
a
gr
a
t
h,
“
I
m
pr
ovi
s
a
t
i
on
of
R
e
dd
i
t
f
l
a
i
r
d
e
t
e
c
t
i
on
us
i
ng
T
F
-
I
D
F
a
n
d
c
ount
ve
c
t
or
i
z
e
r
,”
i
n
A
I
P
C
onf
e
r
e
n
c
e
P
r
oc
e
e
di
ng
s
, 2023,
d
oi
:
10.1063/
5.0181369.
[
14]
K
.
M
.
S
ur
ya
ni
ng
r
um
,
“
C
om
pa
r
i
s
on
of
t
he
T
F
-
I
D
F
m
e
t
hod
w
i
t
h
t
he
c
ount
ve
c
t
or
i
z
e
r
t
o
c
l
a
s
s
i
f
y
ha
t
e
s
pe
e
c
h
,”
E
ngi
n
e
e
r
i
ng,
M
at
he
m
at
i
c
s
and C
om
put
e
r
Sc
i
e
nc
e
J
our
nal
, vo
l
. 5, no. 2, pp. 79
–
83, M
a
y 2023, d
oi
:
10.21512/
e
m
a
c
s
j
our
na
l
.v5i
2.9978.
[
15]
T
. A
hm
e
d
, S
. F
. M
ukt
a
, T
. A
l
M
a
hm
ud
, S
. A
l
H
a
s
a
n, a
nd
M
. G
.
H
us
s
a
i
n, “
B
a
ngl
a
t
e
xt
e
m
ot
i
on c
l
a
s
s
i
f
i
c
a
t
i
on us
i
ng L
R
, M
N
B
a
nd
M
L
P
w
i
t
h
T
F
-
I
D
F
&
C
ount
V
e
c
t
or
i
z
e
r
,”
i
n
I
C
SE
C
2022
-
I
n
t
e
r
nat
i
onal
C
om
put
e
r
S
c
i
e
nc
e
and
E
ng
i
ne
e
r
i
ng
C
onf
e
r
e
nc
e
2022
,
D
e
c
. 2022, pp. 275
–
280
, d
oi
:
10.1109/
I
C
S
E
C
56337.2022.10049341.
[
16]
K
.
K
r
a
m
e
r
,
“
C
om
pa
r
a
t
i
ve
a
na
l
ys
i
s
of
d
oc
um
e
nt
-
l
e
ve
l
e
m
be
d
d
i
ng
m
e
t
ho
d
s
f
or
s
i
m
i
l
a
r
i
t
y
s
c
or
i
ng
on
S
ha
ke
s
pe
a
r
e
s
onne
t
s
a
nd
T
a
yl
o
r
S
w
i
f
t
l
yr
i
c
s
,”
M
oon
l
i
gh
t
:
A
I
C
ol
l
e
ague
f
o
r
R
e
s
e
ar
c
h P
ape
r
s
,
D
e
c
. 2024
.
[
17]
I
. S
ol
d
e
vi
l
l
a
a
nd
N
. F
l
or
e
s
, “
N
a
t
ur
a
l
l
a
ngua
ge
p
r
oc
e
s
s
i
ng t
hr
ough B
E
R
T
f
or
i
d
e
nt
i
f
yi
ng ge
nd
e
r
-
ba
s
e
d
vi
ol
e
nc
e
m
e
s
s
a
g
e
s
on s
oc
i
a
l
m
e
d
i
a
,”
i
n
2021
I
E
E
E
I
nt
e
r
nat
i
ona
l
C
onf
e
r
e
n
c
e
on
I
nf
o
r
m
at
i
on
C
om
m
uni
c
at
i
on
and
So
f
t
w
ar
e
E
ng
i
ne
e
r
i
ng,
I
C
I
C
SE
2021
,
M
a
r
. 2021, pp. 204
–
208
, d
oi
:
10.1109/
I
C
I
C
S
E
52190.2021.9404127.
[
18]
M
.
G
.
P
r
a
d
a
na
, N
.
I
r
z
a
vi
ka
,
N
.
M
a
ul
a
na
,
J
.
M
u,
a
nd
V
.
K
.
W
a
r
i
,
“
P
e
r
f
or
m
a
nc
e
i
m
p
r
ove
m
e
nt
of
c
os
i
ne
s
i
m
i
l
a
r
i
t
y
a
l
go
r
i
t
hm
w
i
t
h
bi
d
i
r
e
c
t
i
ona
l
e
nc
od
e
r
r
e
pr
e
s
e
nt
a
t
i
ons
f
r
om
t
r
a
ns
f
or
m
e
r
s
on
a
bs
t
r
a
c
t
d
oc
um
e
nt
s
i
m
i
l
a
r
i
t
y
d
e
t
e
c
t
i
on,”
J
O
I
V
:
I
nt
e
r
na
t
i
ona
l
J
our
nal
on I
nf
or
m
at
i
c
s
V
i
s
ua
l
i
z
a
t
i
on
, vo
l
. 9, no. 2, M
a
r
. 2025, d
oi
:
10.62527
/
j
o
i
v.9.2.2853.
[
19]
A
. B
ha
va
ni
a
n
d
B
. S
.
K
um
a
r
, “
A
r
e
vi
e
w
of
s
t
a
t
e
a
r
t
o
f
t
e
x
t
c
l
a
s
s
i
f
i
c
a
t
i
on a
l
gor
i
t
hm
s
,”
i
n
5t
h I
nt
e
r
nat
i
onal
C
onf
e
r
e
nc
e
on C
om
put
i
ng
M
e
t
hodol
ogi
e
s
and C
om
m
u
ni
c
at
i
on, I
C
C
M
C
2021
,
A
pr
. 202
1, pp. 148
4
–
1490
,
d
oi
:
10.110
9/
I
C
C
M
C
510
19.2021.94
18262.
[
20]
A
.
D
ha
r
,
H
.
M
ukhe
r
j
e
e
,
N
.
S
.
D
a
s
h,
a
nd
K
. R
oy,
“
T
e
xt
c
a
t
e
gor
i
z
a
t
i
on:
pa
s
t
a
nd
pr
e
s
e
nt
,”
A
r
t
i
f
i
c
i
al
I
n
t
e
l
l
i
ge
nc
e
R
e
v
i
e
w
,
vol
.
54,
no. 4, pp. 3007
–
3054, A
pr
. 2021, d
oi
:
10.1007/
s
10462
-
020
-
09919
-
1.
[
21]
A
.
A
.
J
a
l
a
l
a
nd
B
.
H
.
A
l
i
,
“
T
e
xt
d
oc
um
e
nt
s
c
l
us
t
e
r
i
ng
us
i
ng
d
a
t
a
m
i
ni
ng
t
e
c
hni
que
s
,”
I
nt
e
r
na
t
i
ona
l
J
our
na
l
of
E
l
e
c
t
r
i
c
al
and
C
om
put
e
r
E
ngi
ne
e
r
i
ng
, vo
l
. 11, no. 1, pp. 664
–
670, F
e
b. 2021, d
oi
:
10.11591/
i
j
e
c
e
.v11
i
1.pp664
-
670.
[
22]
L
.
A
bua
l
i
ga
h
e
t
al
.
,
“
A
d
va
nc
e
s
i
n
m
e
t
a
-
he
u
r
i
s
t
i
c
opt
i
m
i
z
a
t
i
on
a
l
gor
i
t
hm
s
i
n
bi
g
d
a
t
a
t
e
xt
c
l
us
t
e
r
i
ng,”
E
l
e
c
t
r
on
i
c
s
,
vol
.
10,
no.
2,
pp. 1
–
29, J
a
n. 2021, d
oi
:
10.3390/
e
l
e
c
t
r
on
i
c
s
10020101.
[
23]
B
.
D
i
a
l
l
o,
J
.
H
u,
T
.
L
i
,
G
.
A
.
K
ha
n,
X
.
L
i
a
ng,
a
nd
Y
.
Z
ha
o,
“
D
e
e
p
e
m
be
dd
i
ng
c
l
us
t
e
r
i
ng
ba
s
e
d
on
c
ont
r
a
c
t
i
ve
a
ut
oe
nc
od
e
r
,”
N
e
ur
oc
om
put
i
ng
, vol
. 433, pp. 96
–
107, A
pr
. 2021, d
oi
:
10.1016/
j
.ne
uc
om
.2020.12.094.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
dv A
ppl
S
c
i
I
S
S
N
:
2252
-
8814
Se
m
ant
i
c
c
l
us
t
e
r
i
ng of
s
c
i
e
nt
i
f
i
c
abs
t
r
ac
t
s
w
i
t
h
t
r
ans
f
o
r
m
e
r
e
m
be
ddi
ngs
…
(
M
us
t
hof
a G
al
i
h P
r
adana
)
539
[
24]
B
.
V
.
V
.
S
.
P
r
a
s
a
d
,
G
.
S
uc
ha
r
i
t
ha
,
K
.
G
.
S
.
V
e
nka
t
e
s
a
n,
T
.
R
.
P
a
t
na
l
a
,
T
.
M
ur
a
r
i
,
a
nd
S
.
R
.
K
a
r
a
na
m
,
“
O
pt
i
m
i
s
a
t
i
on
of
t
he
e
xe
c
ut
i
on
t
i
m
e
us
i
ng
H
a
d
oop
-
ba
s
e
d
pa
r
a
l
l
e
l
m
a
c
hi
ne
l
e
a
r
ni
ng
on
c
om
put
i
ng
c
l
us
t
e
r
s
,”
i
n
C
om
put
e
r
N
e
t
w
or
k
s
,
B
i
g
D
at
a
and
I
oT
,
S
i
nga
por
e
:
S
p
r
i
nge
r
,
2022, pp. 233
–
244
, d
oi
:
10.1007
/
978
-
981
-
19
-
0898
-
9_18.
[
25]
K
.
D
a
hi
ya
e
t
al
.
,
“
D
e
e
pX
M
L
:
a
d
e
e
p
e
xt
r
e
m
e
m
ul
t
i
-
l
a
be
l
l
e
a
r
n
i
ng
f
r
a
m
e
w
or
k
a
ppl
i
e
d
t
o
s
hor
t
t
e
xt
d
oc
u
m
e
nt
s
,”
i
n
W
SD
M
2021
-
P
r
oc
e
e
di
ngs
o
f
t
he
14
t
h
A
C
M
I
n
t
e
r
nat
i
onal
C
on
f
e
r
e
nc
e
on
W
e
b
Se
ar
c
h
and
D
at
a
M
i
ni
ng
,
M
a
r
.
2021,
pp.
31
–
39.
d
oi
:
10.1145/
3437963.3441810.
[
26]
A
.
S
uba
kt
i
, H
.
M
ur
f
i
,
a
nd
N
.
H
a
r
i
a
d
i
,
“
T
he
pe
r
f
or
m
a
nc
e
of
B
E
R
T
a
s
d
a
t
a
r
e
pr
e
s
e
nt
a
t
i
on
o
f
t
e
xt
c
l
us
t
e
r
i
ng,”
J
our
nal
of
B
i
g
D
at
a
,
vol
. 9, no. 1, D
e
c
. 2022, d
oi
:
10.1186/
s
40537
-
022
-
00564
-
9.
[
27]
R
. S
a
ha
, “
I
nf
l
ue
nc
e
of
va
r
i
ous
t
e
xt
e
m
be
dd
i
ngs
on c
l
us
t
e
r
i
ng pe
r
f
or
m
a
nc
e
i
n N
L
P
,”
M
a
y 2023
,
ar
X
i
v
:
2305.03144
.
[
28]
V
.
M
e
ht
a
,
S
.
B
a
w
a
,
a
nd
J
.
S
i
ngh,
“
W
E
C
l
us
t
e
r
i
ng:
w
or
d
e
m
be
dd
i
ngs
ba
s
e
d
t
e
xt
c
l
us
t
e
r
i
ng
t
e
c
hni
que
f
or
l
a
r
ge
d
a
t
a
s
e
t
s
,”
C
om
pl
e
x
and I
nt
e
l
l
i
ge
n
t
Sy
s
t
e
m
s
, vol
. 7, no. 6, pp. 3211
–
3224, D
e
c
. 2021, d
oi
:
10.1007/
s
40747
-
021
-
00512
-
9.
[
29]
Z
.
C
he
n,
C
.
M
i
,
S
.
D
uo,
J
.
H
e
,
a
n
d
Y
.
Z
hou,
“
C
l
us
T
op:
a
n
uns
up
e
r
v
i
s
e
d
a
nd
i
nt
e
gr
a
t
e
d
t
e
x
t
c
l
us
t
e
r
i
ng
a
nd
t
opi
c
e
xt
r
a
c
t
i
on
f
r
a
m
e
w
or
k.”
J
a
n. 2023
,
ar
X
i
v
:
2301.00818
.
[
30]
M
.
S
hut
a
yw
i
a
nd
N
.
N
.
K
a
c
houi
e
,
“
S
i
l
houe
t
t
e
a
na
l
ys
i
s
f
or
pe
r
f
or
m
a
nc
e
e
va
l
ua
t
i
on
i
n
m
a
c
h
i
ne
l
e
a
r
ni
ng
w
i
t
h
a
ppl
i
c
a
t
i
ons
t
o
c
l
us
t
e
r
i
ng,”
E
nt
r
opy
, vol
. 23, no. 6, J
un. 2021, d
oi
:
10.3390
/
e
23060759.
[
31]
F
.
R
os
,
R
.
R
i
a
d
,
a
nd
S
.
G
ui
l
l
a
um
e
,
“
P
D
B
I
:
a
pa
r
t
i
t
i
oni
ng
D
a
vi
e
s
-
B
oul
d
i
n
i
nd
e
x
f
or
c
l
us
t
e
r
i
ng
e
va
l
ua
t
i
o
n,”
N
e
ur
oc
om
put
i
ng
,
vol
. 528, pp. 178
–
199, A
pr
. 2023, d
oi
:
10.1016/
j
.ne
uc
o
m
.2023.01.043.
[
32]
P
.
Y
.
H
s
u
a
nd
P
.
A
.
H
.
N
guye
n,
“
A
f
a
s
t
m
e
t
hod
f
or
d
i
s
c
ove
r
i
ng
a
s
u
i
t
a
bl
e
nu
m
be
r
of
c
l
us
t
e
r
s
f
or
f
uz
z
y
c
l
u
s
t
e
r
i
ng,”
I
nt
e
l
l
i
ge
nt
D
at
a A
nal
y
s
i
s
, vo
l
. 26, no. 6, pp. 1523
–
1538, N
ov. 2022, d
oi
:
10.3233/
I
D
A
-
200511.
[
33]
D
.
A
r
i
yus
a
nd
A
r
d
i
a
ns
ya
h,
“
O
pt
i
m
i
z
a
t
i
on
s
ubs
t
i
t
ut
i
on
c
i
phe
r
a
nd
hi
dd
e
n
pl
a
i
nt
e
xt
i
n
i
m
a
ge
d
a
t
a
us
i
n
g
L
S
B
m
e
t
h
o
d
,”
J
our
nal
of
P
hy
s
i
c
s
:
C
onf
e
r
e
nc
e
Se
r
i
e
s
, vo
l
. 1201, no. 1, 2019, d
oi
:
10.1088/
1742
-
6596
/
1201/
1
/
012033.
[
34]
M
.
A
ks
oy,
S
.
Y
a
nı
k,
a
nd
M
.
F
.
A
m
a
s
ya
l
i
,
“
A
c
o
m
pa
r
a
t
i
ve
a
na
l
ys
i
s
of
t
e
xt
r
e
pr
e
s
e
nt
a
t
i
on,
c
l
a
s
s
i
f
i
c
a
t
i
on
,
a
nd
c
l
us
t
e
r
i
ng
m
e
t
hod
s
ove
r
r
e
a
l
pr
o
j
e
c
t
pr
opos
a
l
s
,”
I
n
t
e
r
nat
i
onal
J
our
na
l
of
I
nt
e
l
l
i
ge
nt
C
om
put
i
ng
and
C
y
be
r
ne
t
i
c
s
,
vol
.
16,
no.
3,
pp.
595
–
628,
J
ul
. 2023, d
oi
:
10.1108/
I
J
I
C
C
-
11
-
2022
-
0289.
[
35]
Z
.
L
i
,
Q
.
S
u,
S
.
S
i
,
a
nd
J
.
Y
u,
“
L
e
ve
r
a
gi
ng
B
E
R
T
a
nd
T
F
I
D
F
f
e
a
t
ur
e
s
f
or
s
hor
t
t
e
xt
c
l
us
t
e
r
i
ng
vi
a
a
l
i
gn
m
e
nt
-
pr
o
m
ot
i
ng
c
o
-
t
r
a
i
ni
ng,”
i
n
E
M
N
L
P
2024
-
2024
C
onf
e
r
e
nc
e
on
E
m
pi
r
i
c
al
M
e
t
hod
s
i
n
N
a
t
ur
al
L
anguage
P
r
oc
e
s
s
i
ng,
2024,
pp.
14897
–
14913
,
d
oi
:
10.18653/
v1/
2024.e
m
n
l
p
-
m
a
i
n.828.
[
36]
M
.
A
s
a
d
i
,
A
.
H
e
i
d
a
r
i
,
a
nd
N
.
J
.
N
a
vi
m
i
pour
,
“
A
ne
w
f
l
ow
-
ba
s
e
d
a
ppr
oa
c
h
f
or
e
nha
nc
i
ng
bot
ne
t
d
e
t
e
c
t
i
on
e
f
f
i
c
i
e
nc
y
us
i
ng
c
onvol
ut
i
ona
l
ne
u
r
a
l
n
e
t
w
or
ks
a
nd
l
ong s
hor
t
-
t
e
r
m
m
e
m
o
r
y
,”
K
now
l
e
dge
and
I
nf
or
m
at
i
on Sy
s
t
e
m
s
, vol
.
67, pp.
6139
–
6170
, 2025,
d
oi
:
10.1007/
s
10115
-
025
-
02410
-
9.
[
37]
Q
.
X
u,
H
.
G
u,
a
nd
S
.
W
.
J
i
,
“
T
e
xt
c
l
us
t
e
r
i
ng
ba
s
e
d
on
pr
e
-
t
r
a
i
ne
d
m
od
e
l
s
a
nd
a
ut
oe
nc
od
e
r
s
,”
F
r
ont
i
e
r
s
i
n
C
om
put
at
i
onal
N
e
ur
os
c
i
e
nc
e
, vo
l
. 17, J
a
n. 2023, d
oi
:
10.3389/
f
nc
om
.2023.1334436.
[
38]
A
.
M
our
a
,
P
.
L
i
m
a
,
F
.
M
e
nd
onç
a
,
S
.
S
.
M
os
t
a
f
a
,
a
nd
F
.
M
.
-
D
i
a
s
,
“
O
n
t
he
us
e
of
t
r
a
ns
f
or
m
e
r
-
ba
s
e
d
m
od
e
l
s
f
or
i
nt
e
nt
d
e
t
e
c
t
i
on
us
i
ng c
l
us
t
e
r
i
ng a
l
go
r
i
t
hm
s
,”
A
pp
l
i
e
d Sc
i
e
nc
e
s
, vo
l
. 13, no. 8, A
pr
. 2023, d
oi
:
10.3390/
a
pp13085178.
[
39]
L
.
G
e
or
ge
a
n
d
P
.
S
um
a
t
hy,
“
A
n
i
nt
e
gr
a
t
e
d
c
l
us
t
e
r
i
ng
a
nd
B
E
R
T
f
r
a
m
e
w
or
k
f
or
i
m
pr
ove
d
t
opi
c
m
od
e
l
i
ng,”
I
nt
e
r
nat
i
onal
J
our
nal
of
I
nf
or
m
at
i
on T
e
c
hnol
og
y
, vol
. 15, no. 4, pp. 2187
–
2195, A
pr
. 2023, d
oi
:
10.1007/
s
41870
-
023
-
01268
-
w.
[
40]
A
.
P
e
t
ukhova
,
J
.
P
.
M
.
-
C
a
r
va
l
ho,
a
nd
N
.
F
a
c
ha
d
a
,
“
T
e
xt
c
l
us
t
e
r
i
ng
w
i
t
h
l
a
r
ge
l
a
ngua
ge
m
od
e
l
e
m
be
d
d
i
ngs
,”
I
nt
e
r
nat
i
onal
J
our
nal
of
C
ogni
t
i
v
e
C
om
put
i
ng i
n E
ngi
ne
e
r
i
ng
, vo
l
. 6. pp. 100
–
108, D
e
c
. 2025
, d
oi
:
10.1016/
j
.
i
j
c
c
e
.2024.11.004.
[
41]
S
ut
r
i
a
w
a
n,
S
.
R
us
t
a
d
, G
.
F
.
S
hi
d
i
k,
a
n
d
P
uj
i
ono,
“
P
e
r
f
or
m
a
nc
e
e
va
l
ua
t
i
on
of
t
e
xt
e
m
be
dd
i
ng
m
od
e
l
s
f
or
a
m
bi
gui
t
y
c
l
a
s
s
i
f
i
c
a
t
i
on
in
I
nd
one
s
i
a
n
ne
w
s
c
or
pus
:
a
c
om
pa
r
a
t
i
ve
s
t
ud
y
o
f
T
F
-
I
D
F
,
W
or
d
2V
e
c
,
F
a
s
t
T
e
xt
B
E
R
T
,
a
nd
G
P
T
,”
I
nge
ni
e
r
i
e
de
s
Sy
s
t
e
m
e
s
d’
I
nf
or
m
at
i
on
, vol
. 30, no. 6, pp. 1469
–
1482, J
un. 2025, d
oi
:
10.18280
/
i
s
i
.300606.
B
I
O
G
R
A
P
H
I
E
S
O
F
A
U
T
H
O
R
S
Musthofa
Galih
Pradana
ob
t
ai
n
ed
h
is
b
ac
h
elor'
s
degree
i
n
I
n
for
m
a
t
ics
fro
m
Un
iversi
t
as
A
m
iko
m
Yogyakar
t
a
i
n
2017,
followed
by
a
m
as
t
er'
s
degree
i
n
th
e
sa
m
e
field
fro
m
th
e
sa
m
e
i
n
s
t
i
tut
io
n
i
n
2019
.
I
n
th
e
sa
m
e
year,
h
e
bega
n
h
is
acade
m
ic
career
as
a
lec
tu
rer
a
t
Un
iversi
t
as
Al
m
a
A
t
a
Yogyakar
t
a,
w
h
ere
h
e
served
unt
il
2022.
Si
n
ce
2022,
h
e
h
as
bee
n
a
fu
ll
-
t
i
m
e
lec
tu
rer
i
n
th
e
un
dergrad
u
a
t
e
progra
m
i
n
Da
t
a
Scie
n
ce
,
Fac
u
l
t
y
of
Co
m
p
uter
Scie
n
ce,
UPN
Ve
t
era
n
Jakar
t
a
.
His
acade
m
ic
a
n
d
researc
h
i
nt
eres
t
s
i
n
cl
u
de
co
m
p
ut
a
t
io
n
al
li
n
g
u
is
t
ics,
n
a
tu
ral
la
n
g
u
age
processi
n
g
(NLP)
,
a
n
d
i
m
age
process
i
n
g
.
He
is
ac
t
ively
i
n
volved
i
n
t
eac
h
i
n
g,
research,
a
n
d
sc
h
olarly
develop
m
e
nt
i
n
th
ese
fields
.
He
ca
n
be
co
nt
ac
t
ed
a
t
e
m
ail:
mu
s
th
ofagali
h
prada
n
a@
u
p
n
vj
.
ac
.
id
.
Pujo
Hari
Saputro
is
a
lec
tu
rer
i
n
th
e
Depar
tm
e
nt
of
I
n
for
m
a
t
ics
E
n
gi
n
eeri
n
g
,
Facu
l
t
y
of
E
n
gi
n
eeri
n
g
,
Sa
m
Ra
tu
la
n
gi
U
n
iversi
t
y.
His
areas
of
exper
t
ise
a
n
d
researc
h
int
eres
t
i
n
cl
u
de
deep
lear
n
i
n
g
,
i
m
age
processi
n
g
,
n
a
tu
ral
la
n
g
u
age
processi
n
g
,
th
e
i
nt
er
n
e
t
of
th
i
n
gs
(IoT),
a
n
d
e
-
gover
nm
e
nt
.
He
h
olds
a
Bac
h
elor'
s
degree
(S.
Ko
m.
)
i
n
I
n
for
m
a
t
ics
En
gi
n
eeri
n
g
fro
m
AMKOM
U
n
iversi
t
y
Yogyakar
t
a
a
n
d
a
m
as
t
er'
s
degree
(M.
T
.
)
i
n
th
e
sa
m
e
field
fro
m
A
tm
a
Jaya
U
n
iversi
t
y
Yogyakar
t
a
.
He
ca
n
be
co
nt
ac
t
ed
a
t
e
m
ail:
pu
jo
h
arisap
ut
ro@
un
sra
t.
ac.
id
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8814
I
nt
J
A
dv A
ppl
S
c
i
,
V
ol
. 15, N
o. 2, J
une
2026
:
532
-
540
540
Ardhyansyah
Mualo
received
h
is
bac
h
elor'
s
degree
i
n
I
n
for
m
a
t
ics
E
n
gi
n
eeri
n
g
at
Un
iversi
t
as
Sa
t
ria
Makassar
in
2013
,
a
n
d
Mas
t
er
of
I
n
for
m
a
t
ics
E
n
gi
n
eeri
n
g
at
U
n
iversi
t
as
Atm
a
Jaya
,
Yogyakar
t
a
in
2016
.
He
h
as
bee
n
a
lec
tu
rer
at
a
priva
t
e
un
iversi
t
y
in
Makassar
from
2014
to
2018,
a
n
d
fro
m
2018
unt
il
n
ow
,
he
is
ac
t
ively
t
eac
h
i
n
g
at
th
e
Depar
tm
e
nt
of
In
for
m
a
t
ics
Ma
n
age
m
e
nt
at
th
e
Poli
t
ek
n
ik
Negeri
FakFak
in
Wes
t
Pap
u
a
P
rovi
n
ce.
Ac
t
ive
in
th
e
m
e
m
bers
h
ip
of
th
e
Associa
t
io
n
of
Hig
h
er
Ed
u
ca
t
io
n
in
I
n
for
m
a
t
ics
a
n
d
Co
m
p
ut
ers
(APTIKOM).
Ac
t
ively
co
nt
rib
ut
i
n
g
to
researc
h
a
n
d
co
mmun
i
t
y
service
,
bo
th
locally
p
u
blis
h
ed
in
th
e
m
ass
m
edia
a
n
d
n
a
t
io
n
ally
in
accredi
t
ed
jo
u
r
n
als.
He
ca
n
be
co
nt
ac
t
ed
a
t
e
m
ail:
mu
aloard
h
ya
n
sya
h
@
g
m
ail
.
co
m
.
Arbiati
Faizah
h
as
a
m
as
t
er'
s
degree
i
n
I
n
for
m
a
t
io
n
Sys
t
e
m
s
(M.
Ko
m
.
)
fro
m
Un
iversi
t
as
Dipo
n
egoro
i
n
2017.
Her
researc
h
i
nt
eres
t
s
are
i
nt
ellige
nt
sys
t
e
m
s
a
n
d
n
a
tu
ral
lan
g
u
age
processi
n
g
.
S
h
e
is
c
u
rre
nt
ly
a
n
acade
m
ic
i
n
th
e
Depar
tm
e
nt
of
I
n
for
m
a
t
io
n
Sys
t
e
m
s,
Facu
l
t
y
of
I
n
for
m
a
t
ics,
I
n
s
t
i
tut
Tek
n
ologi
da
n
Bis
n
is
PGRI
Dewa
nt
ara,
Jo
m
ba
n
g,
I
n
do
n
esia.
Sh
e
ca
n
be
co
nt
ac
t
ed
a
t
e
m
ail:
arbia
t
i
.
faiza
h
@
i
t
ebisdewa
nt
ara
.
ac.
id
.
Wahyuni
Fithratul
Zalmi
is
a
lec
tu
rer
of
I
n
for
m
a
t
ics
E
n
gi
n
eeri
n
g
a
t
th
e
Fac
u
l
t
y
of
E
n
gi
n
eeri
n
g,
Sa
m
Ra
tu
la
n
gi
U
n
iversi
t
y.
S
h
e
received
h
er
m
as
t
er’s
d
egree
i
n
I
n
for
m
a
t
ics
En
gi
n
eeri
n
g
f
ro
m
U
n
iversi
t
as
P
ut
ra
I
n
do
n
esia
YPTK
Pada
n
g
i
n
2018.
Her
researc
h
i
nt
eres
t
is
art
ificial
i
nt
ellige
n
ce.
S
h
e
ca
n
be
co
nt
ac
t
ed
a
t
e
m
ail:
wa
h
y
un
i.
fi
th
ra
tu
l
.
zal
m
i@
un
sra
t.
ac.
id
.
Evaluation Warning : The document was created with Spire.PDF for Python.