I
AE
S In
t
er
na
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
1
5
,
No
.
4
,
A
u
g
u
s
t 2
0
2
6
,
p
p
.
3
8
1
8
~
3
8
2
6
I
SS
N:
2
2
5
2
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
15
.i
4
.
p
p
3
8
1
8
-
3
8
2
6
3818
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
G
ene
tic
a
lg
o
rith
m
-
o
ptimized
BE
RTo
pic wit
h S
H
A
P
ex
pla
ina
bility
for
institu
tiona
l resea
rch t
r
end ana
ly
sis
M
uh
a
m
m
a
d De
di I
ra
wa
n
1
,
Yus
t
ria
H
a
nd
ik
a
Sireg
a
r
2
,
H
ewa
M
a
j
ee
d Z
a
ng
a
na
3
,
Ali Ik
hwa
n
4,
5
¹D
e
p
a
r
t
me
n
t
o
f
D
a
t
a
S
c
i
e
n
c
e
,
F
a
c
u
l
t
y
o
f
S
c
i
e
n
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
U
n
i
v
e
r
si
t
a
s I
sl
a
m N
e
g
e
r
i
S
u
ma
t
e
r
a
U
t
a
r
a
,
D
e
l
i
S
e
r
d
a
n
g
,
I
n
d
o
n
e
s
i
a
²D
e
p
a
r
t
me
n
t
o
f
C
o
mp
u
t
e
r
G
r
a
p
h
i
c
s
E
n
g
i
n
e
e
r
i
n
g
T
e
c
h
n
o
l
o
g
y
,
P
o
l
i
t
e
k
n
i
k
C
e
n
d
a
n
a
,
M
e
d
a
n
,
I
n
d
o
n
e
si
a
³D
e
p
a
r
t
me
n
t
o
f
I
n
f
o
r
m
a
t
i
o
n
Te
c
h
n
o
l
o
g
y
,
D
u
h
o
k
T
e
c
h
n
i
c
a
l
C
o
l
l
e
g
e
,
D
u
h
o
k
P
o
l
y
t
e
c
h
n
i
c
U
n
i
v
e
r
si
t
y
,
D
u
h
o
k
,
I
r
a
q
4
D
e
p
a
r
t
me
n
t
o
f
I
n
f
o
r
mat
i
o
n
S
y
s
t
e
ms
,
F
a
c
u
l
t
y
o
f
S
c
i
e
n
c
e
a
n
d
Te
c
h
n
o
l
o
g
y
,
U
n
i
v
e
r
si
t
a
s
I
sl
a
m N
e
g
e
r
i
S
u
mat
e
r
a
U
t
a
r
a
,
D
e
l
i
S
e
r
d
a
n
g
,
I
n
d
o
n
e
si
a
5
U
n
i
v
e
r
s
i
t
i
M
a
l
a
y
si
a
P
e
r
l
i
s,
A
r
a
u
,
M
a
l
a
y
si
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Ap
r
1
1
,
2
0
2
5
R
ev
is
ed
J
u
l 1
0
,
2
0
2
6
Acc
ep
ted
J
u
l 2
1
,
2
0
2
6
In
stit
u
t
io
n
a
l
re
se
a
rc
h
g
ra
n
t
ti
t
les
c
o
n
stit
u
te
sh
o
rt
-
tex
t
g
re
y
li
tera
tu
re
c
h
a
ra
c
teriz
e
d
b
y
h
e
tero
g
e
n
e
o
u
s
se
m
a
n
ti
c
stru
c
tu
re
s,
m
a
k
in
g
t
o
p
i
c
id
e
n
ti
fica
ti
o
n
a
n
d
re
se
a
rc
h
tren
d
a
n
a
ly
sis
c
h
a
ll
e
n
g
in
g
.
T
h
is
st
u
d
y
p
ro
p
o
se
s
a
n
in
teg
ra
ted
b
id
irec
ti
o
n
a
l
e
n
c
o
d
e
r
re
p
re
se
n
tatio
n
fro
m
tran
sfo
rm
e
rs
-
b
a
se
d
to
p
ic
m
o
d
e
li
n
g
(
BERT
o
p
ic
)
fra
m
e
wo
rk
c
o
m
b
in
i
n
g
g
e
n
e
ti
c
a
l
g
o
rit
h
m
(G
A)
-
b
a
se
d
h
y
p
e
rp
a
ra
m
e
ter
o
p
ti
m
iza
ti
o
n
a
n
d
S
h
a
p
ley
a
d
d
it
i
v
e
e
x
p
lan
a
ti
o
n
s
(S
HA
P
)
-
b
a
se
d
i
n
terp
re
tab
i
li
ty
t
o
imp
ro
v
e
se
m
a
n
ti
c
to
p
ic
q
u
a
li
t
y
a
n
d
m
o
d
e
l
tran
sp
a
re
n
c
y
.
G
A
wa
s
a
p
p
li
e
d
to
o
p
ti
m
ize
d
ime
n
si
o
n
a
li
ty
-
re
d
u
c
ti
o
n
a
n
d
d
e
n
sity
-
b
a
se
d
c
lu
ste
ri
n
g
p
a
ra
m
e
ters
,
wh
il
e
S
HA
P
wa
s
u
se
d
t
o
e
s
ti
m
a
te
th
e
c
o
n
tri
b
u
ti
o
n
o
f
b
ig
ra
m
fe
a
tu
re
s
to
th
e
s
u
rro
g
a
te
c
las
sifier’s
p
re
d
ictio
n
s
o
f
BERT
o
p
ic
-
g
e
n
e
ra
ted
t
o
p
ic
lab
e
l
s.
Ex
p
e
rime
n
tal
re
su
l
ts
d
e
m
o
n
st
ra
ted
th
a
t
th
e
p
r
o
p
o
se
d
fra
m
e
wo
rk
im
p
r
o
v
e
d
t
o
p
ic
c
o
h
e
re
n
c
e
fro
m
0
.
3
6
7
t
o
0
.
5
4
3
wh
il
e
re
d
u
c
in
g
t
h
e
o
u
tl
ier
ra
ti
o
fro
m
2
1
.
1
2
%
t
o
1
3
.
5
5
%
.
In
a
d
d
it
io
n
,
t
h
e
n
u
m
b
e
r
o
f
to
p
ics
d
e
c
re
a
se
d
fro
m
4
6
t
o
1
0
,
re
su
lt
i
n
g
in
a
m
o
re
c
o
m
p
a
c
t
a
n
d
les
s
fra
g
m
e
n
ted
to
p
ic
stru
c
tu
re
.
Th
e
re
su
lt
in
g
to
p
ic
stru
c
t
u
re
re
v
e
a
led
d
o
m
in
a
n
t
t
h
e
m
e
s
re
late
d
to
h
ig
h
e
r
e
d
u
c
a
ti
o
n
,
re
li
g
i
o
u
s
m
o
d
e
ra
ti
o
n
,
Isla
m
ic
c
o
u
n
se
li
n
g
,
h
a
lal
to
u
rism
,
a
n
d
sh
a
ria
b
a
n
k
i
n
g
.
Ov
e
ra
ll
,
th
e
p
r
o
p
o
se
d
fra
m
e
wo
rk
c
o
n
tri
b
u
tes
to
t
h
e
d
e
v
e
lo
p
m
e
n
t
o
f
m
o
re
c
o
h
e
re
n
t
,
in
t
e
rp
re
tab
le,
a
n
d
se
m
a
n
ti
c
a
ll
y
r
o
b
u
st
t
o
p
ic
m
o
d
e
li
n
g
f
o
r
i
n
stit
u
ti
o
n
a
l
sh
o
rt
-
tex
t
g
re
y
li
tera
tu
re
a
n
a
ly
sis
.
K
ey
w
o
r
d
s
:
B
E
R
T
o
p
ic
Gen
etic
alg
o
r
ith
m
Gr
ey
liter
atu
r
e
Hy
p
er
p
ar
a
m
eter
o
p
tim
izatio
n
Sem
an
tic
to
p
ic
m
o
d
elin
g
Sh
ap
ley
ad
d
itiv
e
e
x
p
lan
atio
n
s
Sh
o
r
t te
x
t a
n
aly
s
is
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Mu
h
am
m
ad
De
d
i I
r
awa
n
Dep
ar
tm
en
t o
f
Data
Scien
ce
,
Facu
lty
o
f
Scien
ce
an
d
T
ec
h
n
o
lo
g
y
,
U
n
iv
er
s
itas
I
s
lam
Neg
er
i Su
m
ater
a
Utar
a
Deli
Ser
d
an
g
,
No
r
th
Su
m
atr
a
2
0
3
5
3
,
I
n
d
o
n
esia
E
m
ail:
m
d
.
ir
awa
n
@
u
in
s
u
.
ac
.
i
d
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
d
e
v
elo
p
m
e
n
t
o
f
d
ig
ital
t
r
an
s
f
o
r
m
atio
n
h
as
s
ig
n
if
ican
tl
y
in
cr
ea
s
ed
th
e
am
o
u
n
t
o
f
in
s
titu
tio
n
al
g
r
ey
liter
atu
r
e,
in
clu
d
in
g
r
es
ea
r
ch
g
r
a
n
t
titl
es,
in
ter
n
al
r
ep
o
r
ts
,
an
d
ac
a
d
em
ic
d
o
cu
m
en
ts
th
at
ar
e
n
o
t
p
u
b
lis
h
ed
th
r
o
u
g
h
f
o
r
m
al
s
cien
tific
ch
an
n
els
[
1
]
,
[
2
]
.
I
n
t
h
is
s
tu
d
y
,
g
r
e
y
liter
atu
r
e
is
r
ep
r
e
s
en
ted
in
th
e
f
o
r
m
o
f
in
s
titu
tio
n
al
r
esear
ch
g
r
a
n
t
titl
es,
wh
ich
ar
e
ch
ar
ac
te
r
ized
b
y
s
h
o
r
t
tex
t,
h
ete
r
o
g
en
eity
,
an
d
lac
k
o
f
s
tan
d
ar
d
izatio
n
.
T
h
e
h
ig
h
s
em
an
tic
v
ar
iatio
n
with
in
th
ese
s
h
o
r
t
tex
ts
m
ak
es
to
p
ic
i
d
en
tific
atio
n
an
d
r
esear
ch
tr
en
d
an
aly
s
is
m
o
r
e
ch
allen
g
i
n
g
.
I
n
ad
d
itio
n
to
lim
ited
lin
g
u
is
tic
co
n
tex
t,
im
b
alan
ce
d
th
e
m
atic
d
is
tr
ib
u
tio
n
s
m
ay
also
p
r
o
d
u
ce
u
n
s
tab
le
to
p
ic
s
tr
u
ctu
r
es th
at
ar
e
d
if
f
ic
u
lt to
in
ter
p
r
et
s
em
an
tically
.
T
h
e
ad
v
a
n
ce
m
en
t
o
f
tr
an
s
f
o
r
m
er
-
b
ased
n
at
u
r
al
lan
g
u
ag
e
p
r
o
ce
s
s
in
g
(
NL
P)
h
as
en
c
o
u
r
ag
ed
th
e
em
er
g
en
ce
o
f
s
em
an
tic
to
p
ic
m
o
d
elin
g
ap
p
r
o
ac
h
es c
ap
ab
le
o
f
u
n
d
e
r
s
tan
d
in
g
co
n
tex
tu
al
r
e
latio
n
s
h
ip
s
b
etwe
en
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Gen
etic
a
lg
o
r
ith
m
-
o
p
timiz
ed
B
E
R
To
p
ic
w
ith
S
HA
P
ex
p
la
in
a
b
ilit
y
fo
r
…
(
Mu
h
a
mma
d
De
d
i I
r
a
w
a
n
)
3819
wo
r
d
s
an
d
d
o
cu
m
e
n
ts
m
o
r
e
ef
f
ec
tiv
ely
th
an
tr
a
d
itio
n
al
p
r
o
b
ab
ilis
tic
ap
p
r
o
ac
h
es
[
3
]
–
[
5
]
.
O
n
e
r
ap
id
ly
g
r
o
win
g
m
eth
o
d
is
b
id
ir
ec
tio
n
al
en
co
d
er
r
ep
r
esen
tatio
n
s
f
r
o
m
tr
a
n
s
f
o
r
m
er
s
-
b
ased
to
p
ic
m
o
d
elin
g
(
B
E
R
T
o
p
ic)
,
wh
ich
co
m
b
in
es
tr
an
s
f
o
r
m
er
em
b
e
d
d
in
g
s
,
u
n
if
o
r
m
m
an
if
o
ld
ap
p
r
o
x
im
atio
n
an
d
p
r
o
jectio
n
(
UM
AP)
d
im
en
s
io
n
ality
r
ed
u
ctio
n
,
h
ier
ar
c
h
ical
d
en
s
ity
-
b
ased
s
p
atial
clu
s
ter
in
g
o
f
ap
p
licatio
n
s
with
n
o
is
e
(
HDBS
C
AN)
clu
s
ter
in
g
,
an
d
class
-
b
ased
ter
m
f
r
eq
u
e
n
cy
–
in
v
er
s
e
d
o
c
u
m
en
t
f
r
eq
u
en
cy
(c
-
TF
-
I
DF)
to
p
ic
r
e
p
r
e
s
en
tatio
n
[
6
]
.
T
h
is
ap
p
r
o
ac
h
is
co
n
s
id
er
e
d
ca
p
a
b
le
o
f
g
en
er
atin
g
m
o
r
e
co
n
tex
tu
al
to
p
ic
r
ep
r
esen
tatio
n
s
a
n
d
im
p
r
o
v
in
g
to
p
i
c
co
h
er
en
ce
ac
r
o
s
s
v
ar
io
u
s
r
ese
ar
ch
d
o
m
ain
s
[
7
]
–
[
9
]
.
Sev
er
al
s
tu
d
ies
h
av
e
ap
p
lied
B
E
R
T
o
p
ic
to
av
iatio
n
s
af
ety
an
aly
s
is
,
s
o
cial
m
ed
ia,
th
e
in
ter
n
et
o
f
th
in
g
s
,
an
d
g
en
er
ativ
e
ar
tific
ial
in
tellig
en
ce
tr
en
d
s
[
1
0
]
–
[
1
3
]
.
T
h
ese
s
tu
d
ies
d
em
o
n
s
tr
ate
th
at
B
E
R
T
o
p
ic
p
er
f
o
r
m
s
well
in
ca
p
tu
r
in
g
s
em
an
tic
s
tr
u
ctu
r
es
an
d
to
p
ic
v
ar
iatio
n
s
with
in
co
m
p
le
x
d
atasets
.
Ho
wev
er
,
B
E
R
T
o
p
ic
p
er
f
o
r
m
an
ce
is
h
i
g
h
ly
d
ep
e
n
d
en
t
o
n
th
e
co
n
f
i
g
u
r
atio
n
o
f
UM
AP
an
d
HDBS
C
AN
p
ar
am
eter
s
.
Su
b
o
p
tim
al
p
ar
am
eter
s
elec
tio
n
m
ay
r
esu
lt
in
a
h
ig
h
n
u
m
b
er
o
f
o
u
tlier
s
,
lo
w
to
p
ic
co
h
er
en
ce
,
a
n
d
th
e
f
o
r
m
atio
n
o
f
less
r
ep
r
esen
tativ
e
clu
s
ter
s
[
7
]
,
[
9
]
.
T
h
is
is
s
u
e
b
ec
o
m
es
ev
en
m
o
r
e
co
m
p
l
ex
wh
en
ap
p
lied
to
g
r
ey
liter
atu
r
e
d
atasets
ch
ar
ac
ter
ized
b
y
im
b
alan
ce
d
d
ata
d
is
tr
ib
u
tio
n
s
an
d
h
ig
h
h
eter
o
g
en
ei
ty
.
T
o
ad
d
r
ess
th
ese
is
s
u
es,
th
is
s
tu
d
y
ap
p
lies
a
g
en
etic
alg
o
r
ith
m
(
GA)
as
a
h
y
p
er
p
ar
am
eter
o
p
tim
izatio
n
a
p
p
r
o
ac
h
f
o
r
B
E
R
T
o
p
ic.
GA
p
o
s
s
ess
es
g
lo
b
al
ex
p
lo
r
atio
n
ca
p
ab
ilit
ies
f
o
r
id
en
tify
in
g
o
p
tim
al
p
ar
am
eter
c
o
m
b
in
atio
n
s
with
in
m
u
ltid
im
en
s
io
n
al
s
ea
r
ch
s
p
ac
es,
m
ak
in
g
it
ef
f
ec
tiv
e
in
i
m
p
r
o
v
i
n
g
clu
s
ter
in
g
q
u
ality
an
d
m
ac
h
in
e
lear
n
in
g
m
o
d
el
s
tab
ilit
y
[
1
4
]
–
[
1
6
]
.
Alth
o
u
g
h
n
u
m
er
o
u
s
s
tu
d
ies
h
av
e
d
is
cu
s
s
ed
h
y
p
er
p
ar
am
eter
o
p
tim
izatio
n
in
m
ac
h
in
e
lear
n
in
g
m
o
d
el
s
,
th
e
im
p
lem
en
tatio
n
o
f
G
A
-
b
ased
p
ar
am
eter
o
p
tim
izatio
n
f
o
r
B
E
R
T
o
p
ic
i
n
g
r
e
y
liter
atu
r
e
an
al
y
s
is
r
em
ain
s
r
elativ
ely
lim
ited
.
B
ey
o
n
d
m
o
d
el
q
u
ality
,
in
ter
p
r
etab
ilit
y
also
p
r
esen
ts
a
m
ajo
r
c
h
allen
g
e
in
em
b
e
d
d
in
g
-
b
ased
ap
p
r
o
ac
h
es
b
ec
au
s
e
t
h
e
to
p
ic
f
o
r
m
atio
n
p
r
o
ce
s
s
is
o
f
ten
d
if
f
ic
u
lt to
ex
p
lain
tr
an
s
p
ar
en
tly
.
T
h
e
ex
p
lain
a
b
le
ar
tific
ial
in
t
ellig
en
ce
(
XAI
)
a
p
p
r
o
ac
h
b
a
s
ed
o
n
Sh
ap
le
y
ad
d
itiv
e
ex
p
lan
atio
n
s
(
SHAP
)
ca
n
b
e
u
tili
ze
d
to
im
p
r
o
v
e
m
o
d
el
in
ter
p
r
etab
ilit
y
b
y
id
en
tify
in
g
f
ea
tu
r
e
co
n
tr
ib
u
tio
n
s
to
p
r
ed
ictio
n
o
r
to
p
ic
s
tr
u
ctu
r
e
f
o
r
m
atio
n
[
1
7
]
,
[
1
8
]
.
I
n
th
e
co
n
tex
t
o
f
to
p
i
c
m
o
d
elin
g
,
SHAP
en
ab
les
t
h
e
id
en
tific
atio
n
o
f
b
ig
r
am
f
ea
tu
r
e
co
n
t
r
ib
u
tio
n
s
in
f
o
r
m
in
g
d
o
m
in
an
t
th
em
es,
th
er
eb
y
m
ak
i
n
g
t
h
e
m
o
d
el
in
t
er
p
r
etatio
n
p
r
o
ce
s
s
m
o
r
e
tr
an
s
p
a
r
en
t
an
d
u
n
d
er
s
ta
n
d
ab
le.
T
h
e
in
teg
r
atio
n
o
f
GA
-
b
ased
p
ar
a
m
eter
o
p
tim
izatio
n
an
d
SHAP
-
b
ased
in
ter
p
r
etab
ilit
y
is
e
x
p
ec
ted
to
im
p
r
o
v
e
b
o
th
th
e
q
u
ality
an
d
tr
an
s
p
ar
e
n
cy
o
f
B
E
R
T
o
p
ic
in
a
n
aly
zin
g
g
r
e
y
liter
atu
r
e
d
atasets
.
B
ased
o
n
th
ese
is
s
u
es,
th
is
s
tu
d
y
p
r
o
p
o
s
es
a
B
E
R
T
o
p
ic
ap
p
r
o
ac
h
o
p
tim
ized
u
s
in
g
a
GA
an
d
an
aly
ze
d
u
s
in
g
SHAP
f
o
r
id
e
n
tify
in
g
r
esear
ch
tr
e
n
d
s
in
g
r
ey
liter
atu
r
e.
T
h
e
o
b
jectiv
es
o
f
th
is
s
tu
d
y
ar
e
to
:
i)
o
p
tim
ize
UM
AP
an
d
HDBS
C
AN
p
ar
am
eter
s
u
s
in
g
GA,
ii)
ev
alu
ate
im
p
r
o
v
em
en
ts
in
to
p
ic
q
u
ality
b
ased
o
n
to
p
ic
co
h
er
en
ce
,
to
p
ic
d
iv
er
s
ity
,
an
d
o
u
tlier
r
atio
,
an
d
i
ii)
an
aly
ze
th
e
co
n
t
r
ib
u
tio
n
o
f
b
ig
r
am
f
ea
tu
r
es
to
to
p
ic
f
o
r
m
atio
n
u
s
in
g
SHAP.
T
h
e
c
o
n
tr
ib
u
tio
n
o
f
th
is
r
esear
ch
lies
in
th
e
in
teg
r
atio
n
o
f
h
y
p
er
p
ar
am
ete
r
o
p
tim
izatio
n
an
d
XAI
-
b
ased
in
ter
p
r
etab
ilit
y
with
in
th
e
B
E
R
T
o
p
ic
f
r
am
ewo
r
k
to
p
r
o
d
u
ce
a
m
o
r
e
s
tab
le,
r
ep
r
esen
tativ
e,
an
d
tr
an
s
p
ar
en
t
to
p
ic
m
o
d
elin
g
ap
p
r
o
a
ch
f
o
r
h
eter
o
g
en
eo
u
s
an
d
u
n
s
tr
u
ctu
r
ed
g
r
e
y
liter
atu
r
e
d
atasets
.
2.
M
E
T
H
O
D
T
h
is
s
tu
d
y
em
p
lo
y
s
a
GA
-
o
p
tim
ized
B
E
R
T
o
p
ic
f
r
am
ewo
r
k
to
im
p
r
o
v
e
to
p
ic
q
u
ality
an
d
in
ter
p
r
etab
ilit
y
in
in
s
titu
tio
n
a
l
g
r
ey
liter
atu
r
e
an
aly
s
is
.
T
h
e
f
r
am
ewo
r
k
in
te
g
r
ates
d
ataset
co
llectio
n
,
tex
t
p
r
ep
r
o
ce
s
s
in
g
,
s
en
ten
ce
em
b
ed
d
in
g
g
en
e
r
atio
n
,
b
aselin
e
B
E
R
T
o
p
ic
m
o
d
elin
g
,
GA
-
b
a
s
ed
o
p
tim
izatio
n
o
f
UM
AP
an
d
HDB
SC
A
N
h
y
p
er
p
ar
am
eter
s
,
o
p
tim
ized
B
E
R
T
o
p
ic
m
o
d
elin
g
,
an
d
SHAP
-
b
a
s
ed
ex
p
lain
ab
ilit
y
.
T
h
e
o
v
e
r
all
r
esear
ch
wo
r
k
f
lo
w
is
p
r
esen
ted
in
Fig
u
r
e
1
.
Fig
u
r
e
1
.
Pro
p
o
s
ed
GA
-
o
p
tim
ized
B
E
R
T
o
p
ic
f
r
am
ewo
r
k
with
SHAP
-
b
ased
ex
p
lain
ab
ilit
y
2
.
1
.
Da
t
a
s
et
a
nd
prepro
ce
s
s
i
ng
T
h
e
d
ataset
co
m
p
r
is
ed
5
0
2
r
e
s
ea
r
ch
g
r
an
t
titl
es
f
u
n
d
ed
b
y
Un
iv
er
s
itas
I
s
lam
Neg
er
i
Su
m
ater
a
Utar
a
b
etwe
en
2
0
1
8
an
d
2
0
2
4
.
T
h
e
d
ata
wer
e
o
b
tain
ed
f
r
o
m
th
e
in
s
titu
tio
n
al
r
esear
ch
d
atab
ase
m
an
ag
ed
b
y
th
e
I
n
s
titu
te
f
o
r
R
esear
ch
a
n
d
C
o
m
m
u
n
ity
Ser
v
ice.
Du
p
licate,
in
co
m
p
lete,
a
n
d
a
d
m
in
is
tr
ativ
e
-
o
n
ly
e
n
tr
ies
wer
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
4
,
Au
g
u
s
t 2
0
2
6
:
3
8
1
8
-
3
8
2
6
3820
r
em
o
v
ed
b
e
f
o
r
e
an
al
y
s
is
.
T
h
e
s
h
o
r
t
-
tex
t d
ataset
was c
ateg
o
r
i
ze
d
as a
ca
d
em
ic
g
r
ey
liter
atu
r
e
an
d
ch
ar
ac
ter
ize
d
b
y
h
eter
o
g
en
e
o
u
s
ter
m
in
o
lo
g
y
an
d
lim
ited
co
n
te
x
tu
al
in
f
o
r
m
atio
n
[
1
]
,
[
2
]
.
Pre
p
r
o
ce
s
s
in
g
in
clu
d
e
d
ca
s
e
f
o
ld
in
g
,
p
u
n
ctu
atio
n
r
em
o
v
al
,
to
k
en
izatio
n
,
s
to
p
wo
r
d
r
e
m
o
v
al,
n
o
r
m
aliza
tio
n
,
an
d
b
ig
r
am
co
n
s
tr
u
ctio
n
.
Do
cu
m
en
ts
wer
e
th
en
tr
an
s
f
o
r
m
ed
in
t
o
co
n
tex
t
u
al
em
b
ed
d
in
g
s
u
s
in
g
th
e
all
-
Min
iLM
-
L6
-
v
2
Sen
ten
ce
T
r
an
s
f
o
r
m
e
r
m
o
d
el
[
4
]
,
[
5
]
,
[
1
9
]
.
2
.
2
.
B
E
RT
o
pic
f
ra
m
ewo
rk
T
h
i
s
s
t
u
d
y
e
m
p
l
o
y
e
d
B
E
R
T
o
p
i
c
t
o
g
e
n
e
r
a
t
e
s
e
m
a
n
t
i
c
t
o
p
i
c
s
t
r
u
c
t
u
r
e
s
u
s
i
n
g
t
r
a
n
s
f
o
r
m
e
r
e
m
b
e
d
d
i
n
g
s
,
UM
AP
d
im
en
s
io
n
ality
r
ed
u
ctio
n
,
HDBS
C
AN
clu
s
ter
in
g
,
an
d
c
-
TF
-
I
DF
to
p
ic
r
ep
r
esen
ta
tio
n
[
6
]
,
[
2
0
]
,
[
2
1
]
.
UM
AP
was
u
s
ed
to
r
ed
u
ce
em
b
ed
d
in
g
d
im
en
s
io
n
s
,
wh
ile
HDBS
C
AN
g
r
o
u
p
ed
d
o
cu
m
e
n
ts
b
ased
o
n
d
e
n
s
ity
d
is
tr
ib
u
tio
n
s
.
Do
c
u
m
en
ts
ass
ig
n
ed
to
t
o
p
ic
la
b
el
−1
wer
e
tr
ea
ted
as
o
u
tlier
s
.
T
o
p
ic
r
e
p
r
esen
tatio
n
s
wer
e
g
en
er
ated
u
s
in
g
c
-
TF
-
I
DF a
s
f
o
r
m
u
lated
in
(
1
)
.
-
-
,
=
,
∑
,
⋅
(
)
(
1
)
W
h
er
e
,
d
en
o
tes
th
e
f
r
eq
u
e
n
cy
o
f
ter
m
t
in
to
p
ic
,
r
ep
r
esen
ts
th
e
to
tal
n
u
m
b
er
o
f
d
o
cu
m
en
ts
,
an
d
in
d
icate
s
th
e
n
u
m
b
e
r
o
f
to
p
ics
co
n
tain
in
g
ter
m
.
2
.
3
.
G
enet
ic
a
lg
o
rit
h
m
o
pti
m
iza
t
io
n
A
GA
was
ap
p
lied
to
o
p
ti
m
ize
UM
AP
an
d
HDBS
C
A
N
p
ar
am
eter
s
,
in
cl
u
d
in
g
n
_
n
eig
h
b
o
r
s
,
m
i
n
_
cl
u
s
te
r
_
s
i
ze
,
a
n
d
m
i
n
_
s
a
m
p
les
[
1
4
]
–
[
1
6
]
.
T
h
e
s
ea
r
c
h
s
p
a
ce
co
n
s
is
te
d
o
f
n
_
n
e
ig
h
b
o
r
s
∈
{
5
,
8
,
1
0
,
1
2
,
1
5
},
m
in
_
clu
s
ter
_
s
ize
∈
{5
,
8
,
1
0
,
1
2
,
1
5
},
an
d
m
in
_
s
am
p
les
∈
{
1
,
2
,
3
,
5
}.
T
h
e
UM
AP p
ar
am
eter
s
n
_
co
m
p
o
n
e
n
ts
an
d
m
in
_
d
is
t w
er
e
f
ix
ed
at
5
a
n
d
0
.
0
,
r
esp
ec
tiv
ely
.
T
h
e
o
p
ti
m
izatio
n
p
r
o
ce
s
s
u
s
ed
a
p
o
p
u
latio
n
s
ize
o
f
5
a
n
d
3
g
en
er
atio
n
s
with
to
u
r
n
am
e
n
t selec
tio
n
,
cr
o
s
s
o
v
er
p
r
o
b
ab
ilit
y
o
f
0
.
8
,
an
d
m
u
tatio
n
p
r
o
b
ab
ili
ty
o
f
0
.
1
.
T
h
e
f
itn
ess
f
u
n
ctio
n
co
m
b
in
e
d
to
p
ic
co
h
er
en
ce
,
to
p
ic
d
i
v
er
s
ity
,
o
u
tlier
r
atio
,
an
d
to
p
ic
co
m
p
ac
tn
ess
,
as d
ef
in
ed
in
(
2
)
.
=
0
.
35
+
0
.
25
+
0
.
25
(
1
−
)
+
0
.
15
(
2
)
W
h
er
e
d
en
o
tes
th
e
t
o
p
ic
co
h
er
en
ce
,
r
ep
r
esen
ts
th
e
to
p
ic
d
iv
er
s
ity
,
in
d
icate
s
th
e
o
u
tlie
r
r
atio
,
a
n
d
d
en
o
tes
th
e
to
p
ic
co
m
p
ac
tn
ess
.
Fo
r
ea
c
h
p
ar
am
eter
co
m
b
in
atio
n
,
a
B
E
R
T
o
p
ic
m
o
d
el
was
g
en
er
ated
an
d
ev
alu
ated
iter
ativ
ely
to
id
e
n
tify
th
e
o
p
tim
al
s
em
an
tic
to
p
ic
s
tr
u
ctu
r
e
with
im
p
r
o
v
ed
cl
u
s
ter
in
g
s
tab
ilit
y
an
d
r
ed
u
ce
d
t
o
p
ic
f
r
ag
m
en
tatio
n
.
T
h
e
GA
o
p
tim
izatio
n
p
r
o
ce
s
s
co
n
s
is
ted
o
f
p
o
p
u
latio
n
in
itializatio
n
,
f
itn
ess
ev
alu
atio
n
,
s
elec
tio
n
,
c
r
o
s
s
o
v
e
r
,
m
u
tatio
n
,
an
d
p
o
p
u
latio
n
u
p
d
ates u
n
til th
e
s
to
p
p
in
g
cr
iter
i
a
wer
e
ac
h
iev
ed
.
2
.
4
.
SH
AP
-
ba
s
ed
ex
pla
ina
bil
it
y
SHAP
was
ap
p
lied
to
in
ter
p
r
et
th
e
s
u
r
r
o
g
ate
class
if
ier
b
y
esti
m
atin
g
th
e
co
n
tr
ib
u
tio
n
o
f
T
F
-
I
DF
b
ig
r
am
f
ea
tu
r
es
to
th
e
p
r
e
d
ictio
n
o
f
B
E
R
T
o
p
ic
-
g
en
er
ated
to
p
ic
lab
els
[
1
7
]
,
[
1
8
]
.
B
ec
au
s
e
B
E
R
T
o
p
ic
i
s
u
n
s
u
p
er
v
is
ed
,
t
o
p
ic
lab
els
g
e
n
er
ated
b
y
th
e
o
p
tim
ize
d
m
o
d
el
wer
e
u
s
ed
in
a
s
u
r
r
o
g
ate
lo
g
is
tic
r
eg
r
ess
io
n
class
if
ier
with
T
F
-
I
DF
b
ig
r
am
f
ea
tu
r
es.
SHAP
L
in
ea
r
E
x
p
lain
er
was
th
en
em
p
lo
y
ed
to
esti
m
ate
f
ea
tu
r
e
im
p
o
r
tan
ce
f
o
r
s
em
an
tic
to
p
ic
r
ep
r
esen
tatio
n
.
2
.
5
.
M
o
del
ev
a
lua
t
i
o
n
Mo
d
el
p
er
f
o
r
m
an
ce
was
ev
al
u
ated
u
s
in
g
to
p
ic
co
h
er
e
n
ce
,
t
o
p
ic
d
iv
er
s
ity
,
o
u
tlier
r
atio
,
a
n
d
n
u
m
b
er
o
f
to
p
ics
[
7
]
,
[
2
0
]
,
[
2
2
]
.
C
o
m
p
ar
ativ
e
an
aly
s
is
was
co
n
d
u
ct
ed
b
etwe
en
th
e
b
aselin
e
B
E
R
T
o
p
ic
m
o
d
el
a
n
d
th
e
GA
-
o
p
tim
ized
B
E
R
T
o
p
ic
m
o
d
el
.
T
h
is
c
o
m
p
ar
is
o
n
aim
ed
t
o
ass
ess
s
em
an
tic
co
h
er
en
ce
,
clu
s
ter
in
g
s
tab
ilit
y
,
an
d
to
p
ic
c
o
m
p
ac
tn
ess
.
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
3
.
1
.
E
v
a
lua
t
i
o
n o
f
g
enet
ic
a
l
g
o
rit
hm
-
o
ptim
ized
B
E
RT
o
p
ic
T
h
e
m
o
d
el
p
er
f
o
r
m
an
ce
was
ev
alu
ated
b
y
co
m
p
a
r
in
g
th
e
b
aselin
e
B
E
R
T
o
p
ic
m
o
d
el
with
d
ef
au
lt
p
ar
am
eter
s
ettin
g
s
an
d
th
e
GA
-
o
p
tim
ized
B
E
R
T
o
p
ic
m
o
d
el
u
s
in
g
to
p
ic
co
h
e
r
en
ce
,
to
p
ic
d
iv
er
s
ity
,
o
u
tlier
r
atio
,
an
d
n
u
m
b
er
o
f
to
p
ics
m
etr
ics,
as
p
r
esen
ted
in
T
ab
le
1
.
T
h
e
r
esu
lts
in
T
ab
le
1
in
d
icate
th
at
p
ar
am
eter
o
p
tim
izatio
n
u
s
in
g
GA
im
p
r
o
v
ed
to
p
ic
m
o
d
elin
g
q
u
ality
.
T
o
p
ic
co
h
er
e
n
ce
in
cr
ea
s
ed
b
y
ap
p
r
o
x
im
ately
4
8
.
0
%,
i
n
d
icatin
g
en
h
a
n
ce
d
s
em
an
tic
co
n
s
is
ten
cy
am
o
n
g
d
o
cu
m
en
ts
with
in
ea
c
h
to
p
ic.
I
n
ad
d
itio
n
,
th
e
o
u
tlier
r
atio
d
ec
r
ea
s
ed
b
y
a
p
p
r
o
x
im
ately
3
5
.
8
%,
d
em
o
n
s
t
r
atin
g
im
p
r
o
v
e
d
m
o
d
el
ca
p
a
b
ilit
y
in
clu
s
ter
in
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Gen
etic
a
lg
o
r
ith
m
-
o
p
timiz
ed
B
E
R
To
p
ic
w
ith
S
HA
P
ex
p
la
in
a
b
ilit
y
fo
r
…
(
Mu
h
a
mma
d
De
d
i I
r
a
w
a
n
)
3821
d
o
cu
m
e
n
ts
in
to
m
o
r
e
r
e
p
r
ese
n
tativ
e
g
r
o
u
p
s
.
T
h
e
r
e
d
u
ctio
n
in
th
e
n
u
m
b
er
o
f
to
p
ics
f
r
o
m
4
6
to
1
0
also
p
r
o
d
u
ce
d
a
m
o
r
e
c
o
m
p
ac
t
a
n
d
s
tab
le
t
o
p
ic
s
tr
u
ct
u
r
e.
Alth
o
u
g
h
a
s
lig
h
t
d
ec
r
ea
s
e
in
t
o
p
ic
d
i
v
er
s
ity
was
o
b
s
er
v
ed
,
th
e
v
al
u
e
r
em
ai
n
ed
with
in
a
h
ig
h
r
an
g
e,
s
u
g
g
esti
n
g
th
at
th
e
tr
ad
e
-
o
f
f
f
o
r
i
m
p
r
o
v
ed
co
h
er
en
ce
was
s
till
ac
ce
p
tab
le.
T
h
e
GA
o
p
ti
m
izatio
n
p
r
o
ce
s
s
id
en
tifie
d
a
n
o
p
tim
al
p
ar
am
eter
c
o
m
b
in
atio
n
co
n
s
is
tin
g
o
f
n
_
n
eig
h
b
o
r
s
=
1
2
,
m
i
n
_
clu
s
te
r
_
s
ize
=
1
5
,
a
n
d
m
in
_
s
am
p
les
=
1
.
T
h
is
p
ar
am
eter
co
n
f
ig
u
r
atio
n
p
r
o
d
u
ce
d
th
e
h
ig
h
est
f
itn
ess
s
co
r
e
b
y
im
p
r
o
v
in
g
s
em
an
tic
co
h
e
r
en
ce
w
h
ile
r
ed
u
ci
n
g
to
p
ic
f
r
a
g
m
en
ta
tio
n
an
d
clu
s
ter
in
g
o
u
tlier
s
in
th
e
s
h
o
r
t
-
te
x
t g
r
e
y
l
iter
atu
r
e
d
ataset.
T
ab
le
1
.
Per
f
o
r
m
an
ce
co
m
p
a
r
is
o
n
b
etwe
en
b
aselin
e
B
E
R
T
o
p
ic
an
d
GA
-
Op
tim
ized
B
E
R
T
o
p
ic
M
e
t
r
i
c
B
a
se
l
i
n
e
GA
To
p
i
c
c
o
h
e
r
e
n
c
e
0
.
3
6
7
0
.
5
4
3
To
p
i
c
d
i
v
e
r
si
t
y
~
1
.
0
0
0
~
0
.
9
8
8
O
u
t
l
i
e
r
r
a
t
i
o
2
1
.
1
2
%
1
3
.
5
5
%
N
u
mb
e
r
o
f
t
o
p
i
c
s
46
10
3
.
2
.
T
o
pic
s
t
ruct
ure
o
f
g
enet
ic
a
lg
o
rit
hm
-
o
ptim
ized
B
E
R
T
o
pic
T
h
e
s
em
an
tic
r
elatio
n
s
h
ip
s
am
o
n
g
th
e
to
p
ics
g
en
e
r
ated
b
y
th
e
GA
-
o
p
tim
ized
B
E
R
T
o
p
ic
m
o
d
el
wer
e
v
is
u
alize
d
u
s
in
g
a
n
in
te
r
to
p
ic
d
is
tan
ce
m
ap
,
as
p
r
esen
ted
i
n
Fig
u
r
e
2
.
T
h
e
m
ap
was
u
s
ed
to
e
x
am
in
e
th
e
r
elativ
e
s
em
an
tic
p
r
o
x
im
ity
a
n
d
s
ep
ar
atio
n
am
o
n
g
t
h
e
to
p
ic
s
id
en
tifie
d
b
y
th
e
o
p
tim
ized
m
o
d
el.
As
s
h
o
wn
in
Fig
u
r
e
2
,
th
e
GA
-
o
p
tim
ized
B
E
R
T
o
p
ic
m
o
d
el
g
en
er
ate
d
ten
m
ain
to
p
ics,
lab
eled
T
0
–
T
9
.
Do
c
u
m
en
ts
ass
ig
n
ed
to
to
p
ic
−1
wer
e
class
if
ied
as
o
u
tlier
s
an
d
wer
e
th
e
r
ef
o
r
e
ex
clu
d
e
d
f
r
o
m
th
e
v
is
u
aliza
tio
n
.
T
h
e
m
ap
s
h
o
ws
th
at
s
ev
er
al
to
p
ics
ar
e
p
o
s
itio
n
ed
in
clo
s
e
s
em
an
t
ic
p
r
o
x
im
ity
,
in
d
icatin
g
t
h
at
th
ey
s
h
ar
e
r
elate
d
th
em
atic
co
n
ten
t,
wh
er
ea
s
o
th
er
to
p
ics
o
cc
u
p
y
m
o
r
e
d
i
s
tan
t
p
o
s
itio
n
s
an
d
th
er
ef
o
r
e
ex
h
ib
it
r
elativ
ely
d
is
tin
ctiv
e
s
em
an
tic
ch
ar
ac
ter
is
tics
.
Fig
u
r
e
2
.
I
n
ter
to
p
ic
d
is
tan
ce
m
ap
o
f
t
h
e
GA
-
o
p
tim
ized
B
E
R
to
p
ic
m
o
d
el
T
h
e
d
is
tr
ib
u
tio
n
p
r
esen
te
d
in
Fig
u
r
e
2
s
u
g
g
ests
th
at
th
e
GA
-
b
ased
o
p
tim
izatio
n
p
r
o
d
u
ce
d
a
co
m
p
ac
t
to
p
ic
s
tr
u
ctu
r
e
with
m
an
a
g
ea
b
le
s
em
an
tic
s
ep
ar
atio
n
.
T
h
e
r
e
lativ
ely
clo
s
e
p
o
s
itio
n
in
g
o
f
s
e
v
er
al
to
p
ics
s
h
o
u
ld
n
o
t
n
ec
ess
ar
ily
b
e
in
te
r
p
r
ete
d
as
p
o
o
r
clu
s
ter
in
g
p
er
f
o
r
m
an
ce
,
b
ec
a
u
s
e
th
e
in
s
titu
tio
n
al
r
esear
ch
co
r
p
u
s
co
n
tain
s
m
u
ltid
is
cip
lin
ar
y
t
h
e
m
es
th
at
m
ay
n
at
u
r
ally
s
h
ar
e
r
elate
d
ter
m
in
o
lo
g
y
a
n
d
co
n
tex
tu
al
ch
ar
ac
ter
is
tics
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
4
,
Au
g
u
s
t 2
0
2
6
:
3
8
1
8
-
3
8
2
6
3822
At
th
e
s
am
e
tim
e,
th
e
m
o
r
e
d
is
tan
t
p
o
s
itio
n
s
o
f
s
ev
er
al
t
o
p
ics
in
d
icate
th
e
p
r
esen
ce
o
f
th
em
es
with
m
o
r
e
d
is
tin
ctiv
e
s
em
an
tic
s
tr
u
ctu
r
es.
Ov
er
all,
th
e
o
p
tim
ize
d
m
o
d
el
r
ed
u
ce
d
e
x
ce
s
s
iv
e
to
p
ic
f
r
ag
m
e
n
tatio
n
an
d
p
r
o
d
u
ce
d
a
m
o
r
e
in
ter
p
r
etab
le
s
tr
u
ctu
r
e
c
o
n
s
is
tin
g
o
f
ten
m
ai
n
to
p
ics.
T
h
e
id
e
n
tifie
d
th
em
es in
clu
d
e
h
ig
h
er
ed
u
ca
tio
n
,
r
elig
i
o
u
s
m
o
d
er
atio
n
,
s
o
cial
d
ev
elo
p
m
e
n
t,
I
s
lam
ic
co
u
n
s
elin
g
,
h
alal
t
o
u
r
is
m
,
s
h
ar
ia
b
a
n
k
in
g
,
an
d
s
ev
er
al
tech
n
ical
an
d
leg
al
d
o
m
ain
s
.
T
h
ese
f
in
d
in
g
s
in
d
icate
th
at
th
e
in
s
titu
tio
n
a
l
g
r
ey
liter
atu
r
e
co
r
p
u
s
r
ef
lects
m
u
ltid
is
cip
lin
ar
y
r
esear
ch
tr
en
d
s
with
v
a
r
y
in
g
d
eg
r
ee
s
o
f
s
em
an
tic
in
ter
r
elatio
n
s
h
ip
.
T
o
s
tr
en
g
th
en
th
e
in
ter
p
r
etatio
n
o
f
th
e
to
p
ic
s
tr
u
ctu
r
e,
t
h
e
r
ep
r
esen
tativ
e
ter
m
s
ass
o
ciate
d
with
th
e
s
elec
ted
to
p
ics
wer
e
f
u
r
th
er
ex
am
in
ed
u
s
in
g
c
-
TF
-
I
DF
s
co
r
es.
T
h
e
d
is
tr
ib
u
tio
n
o
f
th
e
r
ep
r
esen
tativ
e
to
p
ic
wo
r
d
s
is
p
r
esen
te
d
in
Fig
u
r
e
3
.
T
h
e
r
e
p
r
esen
tativ
e
wo
r
d
s
with
th
e
h
i
g
h
est
c
-
TF
-
I
DF
s
co
r
es
ac
r
o
s
s
th
e
ten
to
p
ics
g
en
er
ated
b
y
th
e
GA
-
o
p
tim
ized
B
E
R
T
o
p
ic
m
o
d
el.
T
h
ese
ter
m
s
ch
ar
ac
ter
ize
th
e
s
em
an
tic
co
n
ten
t
o
f
ea
ch
to
p
ic
an
d
f
ac
ilit
ate
to
p
ic
in
ter
p
r
etatio
n
.
T
h
e
ex
tr
ac
ted
k
ey
wo
r
d
s
in
d
icate
t
h
at
th
e
m
o
d
el
ca
p
tu
r
ed
m
ajo
r
th
em
es
r
elate
d
to
h
ig
h
e
r
ed
u
c
atio
n
,
r
elig
io
u
s
m
o
d
e
r
atio
n
,
I
s
lam
ic
co
u
n
s
elin
g
,
h
alal
to
u
r
is
m
,
s
h
ar
ia
b
a
n
k
in
g
,
m
ac
h
in
e
lear
n
in
g
,
an
d
o
th
er
in
ter
d
is
cip
lin
ar
y
r
esear
ch
d
o
m
ain
s
.
Ov
er
all,
t
h
e
c
-
TF
-
I
DF
r
ep
r
esen
tatio
n
s
d
em
o
n
s
tr
ate
th
at
t
h
e
o
p
tim
ized
m
o
d
el
p
r
o
d
u
ce
d
s
em
an
ticall
y
m
ea
n
in
g
f
u
l
an
d
d
is
tin
g
u
is
h
a
b
le
to
p
ic
s
tr
u
ctu
r
es
with
in
th
e
in
s
titu
tio
n
al
r
esear
c
h
co
r
p
u
s
.
Fig
u
r
e
3
.
R
ep
r
esen
tativ
e
to
p
ic
wo
r
d
s
co
r
es b
ased
o
n
c
-
TF
-
I
DF
r
ep
r
esen
tatio
n
3
.
3
.
Sh
a
pley
a
dd
it
iv
e
ex
pla
n
a
t
io
ns
-
ba
s
ed
t
hem
e
im
po
rt
a
nce
a
na
ly
s
is
T
o
im
p
r
o
v
e
m
o
d
el
in
ter
p
r
eta
b
ilit
y
,
th
is
s
tu
d
y
ap
p
lied
SHAP
to
esti
m
ate
th
e
co
n
tr
ib
u
tio
n
o
f
b
ig
r
a
m
f
ea
tu
r
es
to
th
e
p
r
ed
ictio
n
o
f
to
p
ic
lab
els
g
e
n
er
ated
b
y
t
h
e
GA
-
o
p
tim
ized
B
E
R
T
o
p
ic
m
o
d
el
th
r
o
u
g
h
a
s
u
r
r
o
g
ate
class
if
ier
.
Fig
u
r
e
4
s
h
o
ws
th
at
h
ig
h
er
ed
u
ca
tio
n
h
a
d
th
e
h
ig
h
est
co
n
tr
ib
u
tio
n
,
with
a
m
ea
n
ab
s
o
lu
te
SHAP
v
alu
e
o
f
ap
p
r
o
x
im
at
ely
0
.
0
4
4
,
f
o
llo
wed
b
y
r
elig
io
u
s
m
o
d
er
atio
n
an
d
I
s
lam
ic
co
u
n
s
elin
g
.
T
h
e
d
o
m
in
an
ce
o
f
th
ese
th
em
es
i
n
d
icate
s
th
at
ed
u
ca
tio
n
-
an
d
r
elig
io
n
-
r
elate
d
is
s
u
es
p
lay
ed
a
m
ajo
r
r
o
le
in
th
e
s
em
an
tic
s
tr
u
ctu
r
e
id
e
n
tifie
d
with
in
th
e
in
s
titu
tio
n
al
g
r
ey
li
ter
atu
r
e
co
r
p
u
s
.
Oth
er
in
f
lu
en
tial
b
ig
r
am
f
ea
tu
r
es
in
clu
d
ed
h
alal
to
u
r
is
m
,
ea
r
ly
ch
ild
h
o
o
d
,
s
h
ar
ia
b
an
k
in
g
,
m
ac
h
in
e
lear
n
in
g
,
co
m
m
u
n
ity
d
ev
elo
p
m
en
t,
p
u
b
lic
p
o
licy
,
ar
tific
ial
in
tellig
en
ce
,
an
d
d
ee
p
lear
n
in
g
.
T
h
e
p
r
esen
ce
o
f
th
ese
f
e
atu
r
es
r
ef
lects
th
e
in
ter
d
is
cip
lin
ar
y
n
atu
r
e
o
f
th
e
d
ataset,
en
co
m
p
ass
in
g
ed
u
ca
tio
n
al,
r
elig
io
u
s
,
s
o
cial,
ec
o
n
o
m
ic,
p
o
licy
,
an
d
tech
n
o
lo
g
ical
d
o
m
ain
s
.
T
h
e
co
m
p
a
r
ativ
ely
lo
w
c
o
n
tr
i
b
u
tio
n
o
f
co
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
(
C
NN)
alg
o
r
ith
m
s
u
g
g
ests
th
at
th
is
f
ea
tu
r
e
h
ad
o
n
ly
a
lim
ite
d
in
f
lu
e
n
ce
o
n
t
h
e
s
u
r
r
o
g
ate
m
o
d
el’
s
p
r
ed
ictio
n
o
f
B
E
R
T
o
p
ic
-
g
en
er
ate
d
to
p
ic
lab
els.
I
ts
ap
p
ea
r
an
ce
m
a
y
in
d
icate
r
esid
u
al
s
em
an
tic
n
o
is
e
o
r
a
h
ig
h
l
y
s
p
ec
if
ic
te
ch
n
ical
ex
p
r
ess
io
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Gen
etic
a
lg
o
r
ith
m
-
o
p
timiz
ed
B
E
R
To
p
ic
w
ith
S
HA
P
ex
p
la
in
a
b
ilit
y
fo
r
…
(
Mu
h
a
mma
d
De
d
i I
r
a
w
a
n
)
3823
o
cc
u
r
r
in
g
in
a
r
elativ
ely
s
m
all
s
u
b
s
et
o
f
d
o
cu
m
en
ts
.
Ov
er
all
,
th
e
SHAP
an
aly
s
is
p
r
o
v
i
d
es
a
m
o
r
e
tr
an
s
p
a
r
en
t
in
ter
p
r
etatio
n
o
f
th
e
f
ea
tu
r
es
a
s
s
o
ciate
d
with
th
e
t
o
p
ic
-
lab
el
p
r
ed
ictio
n
s
a
n
d
s
tr
en
g
th
e
n
s
th
e
ex
p
lain
a
b
ilit
y
o
f
th
e
GA
-
o
p
tim
ized
B
E
R
T
o
p
ic
f
r
am
ewo
r
k
.
Fig
u
r
e
4
.
SHAP
b
ig
r
am
t
h
em
e
im
p
o
r
tan
ce
f
o
r
t
h
e
GA
-
o
p
tim
i
ze
d
B
E
R
t
o
p
ic
m
o
d
el
3
.
4
.
Dis
cus
s
io
n
T
h
e
f
in
d
in
g
s
d
e
m
o
n
s
tr
ate
th
at
o
p
tim
izin
g
UM
AP
an
d
HDBS
C
AN
p
ar
am
eter
s
u
s
in
g
a
GA
im
p
r
o
v
ed
to
p
ic
m
o
d
elin
g
q
u
ality
in
g
r
ey
liter
atu
r
e
d
atasets
co
n
s
is
tin
g
o
f
in
s
titu
tio
n
al
r
esear
ch
g
r
an
t
titl
es.
T
o
p
ic
co
h
er
en
ce
in
cr
ea
s
ed
f
r
o
m
0
.
3
6
7
to
0
.
5
4
3
,
wh
ile
th
e
o
u
t
lier
r
atio
d
ec
r
ea
s
ed
f
r
o
m
2
1
.
1
2
%
to
1
3
.
5
5
%,
in
d
icatin
g
th
at
t
h
e
o
p
tim
izati
o
n
p
r
o
ce
s
s
s
u
cc
ess
f
u
lly
g
en
er
ated
a
m
o
r
e
s
tab
le
an
d
r
ep
r
e
s
en
tativ
e
clu
s
ter
in
g
s
tr
u
ctu
r
e
co
m
p
ar
e
d
to
th
e
b
aselin
e
B
E
R
T
o
p
ic
m
o
d
el.
T
h
e
s
e
f
in
d
in
g
s
s
u
g
g
est
th
at
p
ar
am
eter
co
n
f
ig
u
r
atio
n
s
ig
n
if
ican
tly
in
f
lu
e
n
ce
s
s
em
an
tic
r
ep
r
esen
tatio
n
q
u
ality
in
B
E
R
T
o
p
ic,
p
ar
ticu
lar
ly
f
o
r
s
h
o
r
t
-
tex
t
d
atasets
ch
ar
ac
ter
ized
b
y
h
ig
h
s
p
ar
s
ity
an
d
im
b
alan
ce
d
to
p
ic
d
is
tr
ib
u
t
io
n
s
[
2
0
]
,
[
2
1
]
.
T
h
e
r
ed
u
ctio
n
in
th
e
n
u
m
b
er
o
f
t
o
p
ics
f
r
o
m
4
6
to
1
0
in
d
icate
s
th
at
GA
-
b
ased
o
p
tim
izatio
n
s
u
cc
ess
f
u
lly
r
ed
u
ce
d
ex
ce
s
s
iv
e
to
p
ic
f
r
ag
m
en
tatio
n
wh
ile
im
p
r
o
v
i
n
g
s
em
an
tic
co
n
s
is
ten
cy
am
o
n
g
d
o
cu
m
e
n
ts
.
I
n
s
h
o
r
t
-
te
x
t
to
p
ic
m
o
d
elin
g
,
to
p
ic
f
r
ag
m
en
tatio
n
f
r
eq
u
en
tl
y
o
cc
u
r
s
d
u
e
to
lim
ited
lin
g
u
is
tic
co
n
tex
t
with
in
ea
ch
d
o
cu
m
e
n
t,
m
ak
in
g
s
em
an
tic
r
elatio
n
s
h
ip
s
b
etwe
en
tex
ts
m
o
r
e
d
if
f
icu
lt
to
p
r
eser
v
e
[
2
1
]
.
C
o
m
p
a
r
ed
co
n
ce
p
tu
ally
with
co
n
v
en
tio
n
al
p
r
o
b
ab
ilis
tic
to
p
ic
m
o
d
elin
g
ap
p
r
o
ac
h
es
s
u
ch
as
laten
t
Dir
ich
let
allo
ca
tio
n
(
L
DA
)
an
d
em
b
ed
d
in
g
-
b
ased
m
eth
o
d
s
s
u
ch
as
T
o
p
2
Vec
,
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
p
r
o
d
u
ce
d
m
o
r
e
co
m
p
ac
t
an
d
in
ter
p
r
etab
le
t
o
p
ic
r
e
p
r
esen
ta
tio
n
s
th
r
o
u
g
h
th
e
in
te
g
r
atio
n
o
f
tr
an
s
f
o
r
m
er
em
b
ed
d
i
n
g
s
an
d
ev
o
lu
tio
n
ar
y
co
m
p
u
tatio
n
-
b
ased
o
p
tim
izati
o
n
[
2
3
]
.
T
h
ese
f
in
d
in
g
s
s
u
g
g
est
th
at
tr
an
s
f
o
r
m
er
em
b
ed
d
i
n
g
s
co
m
b
in
ed
with
ev
o
lu
tio
n
ar
y
o
p
tim
izatio
n
m
ay
im
p
r
o
v
e
s
em
an
tic
co
n
s
is
ten
cy
an
d
clu
s
ter
in
g
s
tab
ilit
y
in
s
h
o
r
t
-
tex
t
g
r
ey
liter
atu
r
e
d
atasets
co
m
p
ar
ed
t
o
ap
p
r
o
ac
h
es with
o
u
t p
a
r
am
ete
r
o
p
tim
izatio
n
.
T
h
e
in
ter
to
p
ic
d
is
tan
ce
m
ap
v
is
u
aliza
tio
n
in
Fig
u
r
e
2
s
h
o
ws
th
at
m
o
s
t
to
p
ics
ex
h
ib
it
r
elati
v
ely
g
o
o
d
s
em
an
tic
s
ep
ar
atio
n
with
m
in
im
al
o
v
er
lap
b
etwe
en
c
lu
s
ter
s
.
T
h
ese
r
esu
lts
in
d
icate
th
at
GA
-
b
ased
o
p
tim
izatio
n
o
f
UM
AP
an
d
HDBS
C
AN
p
ar
am
eter
s
ef
f
ec
tiv
ely
p
r
eser
v
ed
th
e
s
em
an
tic
s
tr
u
ctu
r
e
o
f
d
o
cu
m
en
t
em
b
ed
d
in
g
s
.
T
h
ese
f
in
d
in
g
s
a
r
e
co
n
s
is
ten
t
with
p
r
ev
io
u
s
s
tu
d
ies
s
tatin
g
th
at
clu
s
ter
in
g
q
u
ality
in
B
E
R
T
o
p
ic
i
s
h
i
g
h
l
y
i
n
f
l
u
e
n
c
e
d
b
y
d
i
m
e
n
s
i
o
n
a
l
i
t
y
r
e
d
u
c
t
i
o
n
c
o
n
f
i
g
u
r
a
ti
o
n
a
n
d
d
o
c
u
m
e
n
t
e
m
b
e
d
d
i
n
g
d
i
s
t
r
i
b
u
t
i
o
n
[
2
0
]
,
[
2
2
]
.
I
n
ad
d
itio
n
to
im
p
r
o
v
in
g
clu
s
ter
in
g
q
u
ality
,
th
is
s
tu
d
y
also
d
em
o
n
s
tr
ates
th
at
th
e
SHA
P
ap
p
r
o
ac
h
en
h
an
ce
d
B
E
R
T
o
p
ic
m
o
d
el
tr
an
s
p
ar
en
cy
b
y
i
d
en
tify
in
g
th
e
co
n
tr
ib
u
tio
n
o
f
b
ig
r
a
m
f
ea
tu
r
es
to
to
p
ic
f
o
r
m
atio
n
.
T
h
e
d
o
m
in
an
ce
o
f
th
em
es
s
u
ch
as
h
ig
h
er
ed
u
ca
ti
o
n
,
r
elig
io
u
s
m
o
d
er
atio
n
,
an
d
I
s
lam
ic
co
u
n
s
elin
g
in
d
icate
s
th
at
th
e
m
o
d
el
s
u
c
ce
s
s
f
u
lly
ca
p
tu
r
ed
th
e
p
r
im
a
r
y
s
em
an
tic
s
tr
u
ctu
r
es
with
i
n
th
e
in
s
titu
tio
n
al
r
esear
ch
g
r
an
t
d
ataset.
T
h
ese
f
in
d
in
g
s
s
u
g
g
est
th
at
ed
u
ca
tio
n
al
an
d
r
elig
io
u
s
m
o
d
er
a
tio
n
is
s
u
es
r
em
ain
d
o
m
in
an
t
th
em
es
in
th
e
d
ev
elo
p
m
en
t
o
f
in
s
titu
tio
n
al
r
esear
ch
with
in
g
r
ey
liter
atu
r
e.
Fu
r
th
er
m
o
r
e,
t
h
e
em
er
g
en
ce
o
f
m
u
ltid
is
cip
lin
ar
y
th
em
es
s
u
ch
as
h
alal
to
u
r
is
m
,
s
h
ar
ia
b
an
k
in
g
,
a
n
d
m
ac
h
i
n
e
lear
n
in
g
in
d
icate
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
4
,
Au
g
u
s
t 2
0
2
6
:
3
8
1
8
-
3
8
2
6
3824
th
at
in
s
titu
tio
n
al
r
esear
ch
to
p
ics
ar
e
n
o
lo
n
g
er
c
o
n
f
i
n
ed
to
a
s
in
g
le
d
o
m
ain
,
b
u
t
h
a
v
e
e
v
o
lv
ed
t
o
war
d
th
e
in
teg
r
atio
n
o
f
e
d
u
ca
tio
n
al,
s
o
cial,
ec
o
n
o
m
ic,
an
d
tech
n
o
lo
g
ical
is
s
u
es.
T
h
ese
f
in
d
in
g
s
ar
e
co
n
s
is
ten
t
with
p
r
ev
io
u
s
s
tu
d
ies
r
ep
o
r
tin
g
th
a
t
B
E
R
T
o
p
ic
an
d
tr
an
s
f
o
r
m
er
-
b
ased
to
p
ic
m
o
d
elin
g
ap
p
r
o
ac
h
es
ar
e
ef
f
ec
tiv
e
in
ex
tr
ac
tin
g
s
em
an
tically
co
h
er
en
t
to
p
ics
f
r
o
m
u
n
s
tr
u
ct
u
r
ed
an
d
s
h
o
r
t
-
tex
t
d
atasets
ac
r
o
s
s
d
iv
er
s
e
ap
p
licatio
n
d
o
m
ain
s
,
in
clu
d
in
g
s
o
cial
m
e
d
ia,
p
u
b
lic
d
is
co
u
r
s
e,
h
ea
lth
ca
r
e,
an
d
r
elig
i
o
u
s
s
tu
d
ies [
4
]
,
[
1
9
]
,
[
2
3
]
–
[
2
6
]
.
Nev
er
th
eless
,
s
ev
er
al
lo
w
-
co
n
tr
ib
u
tio
n
b
ig
r
am
s
s
u
ch
as
C
N
N
alg
o
r
ith
m
an
d
b
est
p
alm
s
til
l
ap
p
ea
r
ed
in
th
e
SHAP
r
esu
lts
.
T
h
e
p
r
esen
ce
o
f
th
ese
f
ea
tu
r
es
in
d
ica
tes
r
esid
u
al
s
em
an
tic
n
o
is
e
in
ce
r
tain
d
o
cu
m
en
ts
ca
u
s
ed
b
y
ter
m
in
o
lo
g
y
v
ar
iat
io
n
an
d
th
e
lim
ited
c
o
n
tex
t
o
f
s
h
o
r
t
te
x
ts
.
Ho
wev
er
,
th
e
in
f
lu
en
ce
o
f
th
ese
f
ea
tu
r
es
was
r
elativ
ely
m
in
o
r
co
m
p
ar
ed
to
th
e
d
o
m
in
a
n
t
th
e
m
es
an
d
th
er
ef
o
r
e
d
id
n
o
t
s
ig
n
if
ican
tly
af
f
ec
t
th
e
o
v
er
all
to
p
ic
s
tr
u
ctu
r
e
.
T
h
is
co
n
d
itio
n
d
em
o
n
s
tr
ates
th
at
GA
-
b
ased
o
p
tim
izatio
n
im
p
r
o
v
ed
m
o
d
el
s
tab
ilit
y
with
o
u
t
elim
in
atin
g
th
em
ati
c
d
iv
er
s
ity
with
in
th
e
d
at
aset
.
Ov
er
all,
th
e
in
teg
r
atio
n
o
f
GA
-
b
ased
h
y
p
er
p
ar
am
eter
o
p
tim
izatio
n
an
d
SHAP
-
b
ased
in
ter
p
r
etab
il
ity
p
r
o
v
id
es
a
m
eth
o
d
o
lo
g
ical
co
n
tr
ib
u
tio
n
to
th
e
d
ev
elo
p
m
e
n
t
o
f
s
em
an
tic
to
p
i
c
m
o
d
elin
g
f
o
r
s
h
o
r
t
-
tex
t
g
r
e
y
liter
atu
r
e
d
atasets
.
T
h
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
n
o
t
o
n
ly
im
p
r
o
v
ed
to
p
ic
r
ep
r
ese
n
tatio
n
q
u
ality
b
u
t
also
p
r
o
d
u
ce
d
m
o
r
e
tr
a
n
s
p
ar
en
t
an
d
i
n
ter
p
r
etab
le
m
o
d
el
ex
p
lan
atio
n
s
f
o
r
in
s
titu
tio
n
al
r
esear
ch
tr
en
d
a
n
aly
s
is
.
Desp
ite
th
e
p
r
o
m
is
in
g
r
esu
lts
,
th
is
s
tu
d
y
h
as
s
ev
er
al
lim
ita
ti
o
n
s
.
First,
th
e
p
r
o
p
o
s
ed
f
r
am
e
wo
r
k
was
ev
alu
ated
o
n
ly
o
n
a
s
in
g
le
in
s
titu
tio
n
al
s
h
o
r
t
-
tex
t
g
r
ey
liter
atu
r
e
d
ataset
co
n
s
is
tin
g
o
f
r
esear
ch
g
r
an
t
titl
es.
Seco
n
d
,
d
ir
ec
t
ex
p
er
im
en
tal
b
en
ch
m
ar
k
in
g
a
g
ain
s
t
o
th
er
to
p
ic
m
o
d
elin
g
a
p
p
r
o
ac
h
es
s
u
ch
as
L
DA
an
d
T
o
p
2
Vec
was
n
o
t
co
n
d
u
cte
d
.
T
h
ir
d
,
s
tak
eh
o
ld
er
-
b
ased
q
u
a
litativ
e
v
alid
atio
n
was
n
o
t
in
clu
d
ed
b
ec
a
u
s
e
th
e
s
tu
d
y
p
r
im
ar
ily
f
o
cu
s
ed
o
n
q
u
an
titativ
e
ev
alu
atio
n
m
etr
ics
a
n
d
s
em
an
tic
in
ter
p
r
eta
b
ilit
y
.
F
u
tu
r
e
s
tu
d
ies
m
a
y
ad
d
r
ess
th
ese
lim
itatio
n
s
th
r
o
u
g
h
c
r
o
s
s
-
d
ataset
ev
alu
atio
n
,
m
u
ltil
in
g
u
al
co
r
p
o
r
a
,
f
u
ll
-
tex
t
d
o
cu
m
e
n
t
an
aly
s
is
,
an
d
in
teg
r
atio
n
with
lar
g
e
la
n
g
u
ag
e
m
o
d
els
(
L
L
Ms)
to
f
u
r
th
er
im
p
r
o
v
e
s
em
an
tic
to
p
ic
r
ep
r
esen
tatio
n
an
d
m
o
d
el
g
e
n
er
aliza
b
ilit
y
.
4.
CO
NCLU
SI
O
N
T
h
i
s
s
t
u
d
y
p
r
o
p
o
s
ed
a
B
E
R
T
o
p
i
c
ap
p
r
o
ac
h
o
p
t
im
i
z
e
d
u
s
i
n
g
a
G
A
a
n
d
in
t
e
r
p
r
e
t
ed
u
s
in
g
S
H
A
P
f
o
r
t
r
e
n
d
a
n
a
ly
s
i
s
o
n
in
s
t
i
t
u
t
i
o
n
a
l
s
h
o
r
t
-
t
ex
t
g
r
e
y
l
i
t
er
a
tu
r
e
.
T
h
e
ex
p
e
r
i
m
e
n
t
a
l
r
e
s
u
l
t
s
d
e
m
o
n
s
t
r
a
t
e
th
a
t
o
p
t
i
m
iz
i
n
g
U
M
A
P
an
d
H
D
B
S
C
A
N
p
a
r
a
m
e
t
e
r
s
u
s
i
n
g
G
A
s
i
g
n
if
i
c
a
n
tl
y
im
p
r
o
v
e
d
t
o
p
i
c
m
o
d
e
l
i
n
g
q
u
a
l
i
t
y
.
T
o
p
i
c
c
o
h
e
r
e
n
c
e
i
n
cr
e
a
s
e
d
f
r
o
m
0
.
3
6
7
t
o
0
.
5
4
3
,
wh
i
l
e
th
e
o
u
t
l
i
er
r
a
t
io
d
e
c
r
e
a
s
e
d
f
r
o
m
2
1
.
1
2
%
t
o
1
3
.
5
5
%
.
I
n
a
d
d
i
t
i
o
n
,
th
e
n
u
m
b
e
r
o
f
t
o
p
i
c
s
w
a
s
r
ed
u
c
ed
f
r
o
m
4
6
t
o
1
0
,
r
e
s
u
l
t
i
n
g
i
n
a
m
o
r
e
c
o
m
p
a
c
t
,
s
t
ab
l
e
,
a
n
d
s
e
m
a
n
t
i
c
a
l
ly
i
n
te
r
p
r
e
t
ab
l
e
t
o
p
i
c
s
tr
u
c
tu
r
e.
T
h
e
v
i
s
u
a
l
i
z
a
t
i
o
n
o
f
th
e
in
t
e
r
to
p
i
c
d
i
s
t
a
n
c
e
m
a
p
an
d
c
-
TF
-
I
D
F
-
b
a
s
ed
w
o
r
d
d
i
s
t
r
i
b
u
t
i
o
n
s
s
h
o
w
ed
t
h
a
t
th
e
m
o
d
e
l
w
a
s
c
a
p
a
b
le
o
f
c
a
p
tu
r
in
g
s
em
a
n
t
i
c
r
e
l
a
t
i
o
n
s
h
ip
s
a
m
o
n
g
to
p
i
c
s
w
i
th
g
o
o
d
s
e
p
ar
a
b
i
l
i
ty
.
T
h
e
d
o
m
in
an
t
t
h
em
es
i
d
e
n
t
i
f
ie
d
in
c
l
u
d
ed
h
ig
h
er
ed
u
c
a
t
io
n
,
r
e
l
ig
i
o
u
s
m
o
d
e
r
a
t
io
n
,
I
s
l
a
m
i
c
co
u
n
s
e
l
i
n
g
,
h
a
l
a
l
t
o
u
r
i
s
m
,
a
n
d
s
h
a
r
i
a
b
an
k
in
g
,
r
e
f
le
c
t
i
n
g
t
h
e
d
ev
e
lo
p
m
e
n
t
o
f
m
u
l
t
i
d
i
s
c
ip
l
i
n
ar
y
t
h
em
e
s
in
i
n
s
t
i
t
u
t
i
o
n
a
l
r
e
s
e
a
r
c
h
.
S
H
A
P
-
b
a
s
e
d
an
a
l
y
s
i
s
a
l
s
o
d
e
m
o
n
s
t
r
a
t
e
d
t
h
a
t
b
ig
r
am
f
e
a
t
u
r
e
s
c
o
n
t
r
ib
u
t
ed
t
o
th
e
s
u
r
r
o
g
a
t
e
m
o
d
e
l
’
s
p
r
e
d
ic
t
i
o
n
o
f
B
E
R
T
o
p
ic
-
g
e
n
er
a
t
ed
t
o
p
i
c
l
a
b
e
l
s
,
t
h
er
e
b
y
i
m
p
r
o
v
in
g
m
o
d
e
l
t
r
a
n
s
p
a
r
en
cy
an
d
i
n
te
r
p
r
e
t
ab
i
l
i
t
y
.
Ov
e
r
a
ll
,
t
h
e
in
t
e
g
r
a
t
io
n
o
f
G
A
-
b
a
s
ed
h
y
p
e
r
p
a
r
am
e
t
er
o
p
t
i
m
iz
a
t
i
o
n
a
n
d
S
H
A
P
-
b
a
s
ed
i
n
t
er
p
r
e
t
a
b
i
l
i
ty
p
r
o
v
i
d
e
s
a
m
e
t
h
o
d
o
l
o
g
i
c
a
l
c
o
n
t
r
ib
u
t
io
n
t
o
t
h
e
d
e
v
e
lo
p
m
e
n
t
o
f
s
e
m
an
t
i
c
t
o
p
i
c
m
o
d
e
l
i
n
g
f
o
r
s
h
o
r
t
-
t
e
x
t
g
r
ey
l
i
t
er
a
t
u
r
e
d
a
t
a
s
e
t
s
.
T
h
e
p
r
o
p
o
s
e
d
a
p
p
r
o
a
ch
n
o
t
o
n
l
y
i
m
p
r
o
v
e
s
to
p
i
c
r
e
p
r
e
s
e
n
t
a
t
io
n
q
u
al
i
t
y
b
u
t
a
l
s
o
p
r
o
d
u
c
e
s
m
o
r
e
tr
a
n
s
p
a
r
en
t
i
n
t
er
p
r
e
t
a
t
io
n
s
i
n
i
n
s
t
i
tu
t
i
o
n
a
l
r
e
s
e
a
r
ch
t
r
e
n
d
an
a
l
y
s
i
s
.
Fu
tu
r
e
s
tu
d
i
e
s
m
a
y
e
x
t
en
d
t
h
is
a
p
p
r
o
a
ch
to
m
u
l
t
i
l
i
n
g
u
a
l
d
a
t
a
s
e
t
s
,
f
u
l
l
-
t
e
x
t
d
o
cu
m
en
t
s
,
o
r
i
n
t
eg
r
a
t
io
n
w
i
t
h
LLMs
t
o
f
u
r
th
e
r
en
h
an
c
e
s
e
m
an
t
i
c
r
ep
r
e
s
en
t
a
t
io
n
q
u
a
l
i
ty
i
n
a
m
o
r
e
c
o
m
p
r
eh
e
n
s
iv
e
m
a
n
n
e
r
.
ACK
NO
WL
E
DG
M
E
N
T
S
T
h
e
au
th
o
r
s
g
r
atef
u
lly
ac
k
n
o
wled
g
e
th
e
I
n
s
titu
te
f
o
r
R
ese
ar
ch
an
d
C
o
m
m
u
n
it
y
Ser
v
ice
(
L
PP
M)
,
Un
iv
er
s
itas
I
s
lam
Neg
er
i Su
m
ater
a
Utar
a,
f
o
r
its
in
s
titu
tio
n
al
s
u
p
p
o
r
t.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
T
h
is
r
esear
ch
was
f
u
n
d
ed
b
y
Un
iv
er
s
itas
I
s
lam
Neg
er
i
S
u
m
ater
a
Utar
a
th
r
o
u
g
h
th
e
2
0
2
4
DI
PA
Hig
h
-
er
E
d
u
ca
tio
n
Dev
el
o
p
m
e
n
t Res
ea
r
ch
Gr
an
t (
Ag
r
ee
m
e
n
t
No
.
B
.
2
4
/Un
.
1
1
.
R
/L2
.
3
/KS/0
.
2
/0
3
/2
0
2
4
)
.
AUTHO
R
CO
NT
RI
B
UT
I
O
NS ST
A
T
E
M
E
N
T
T
h
i
s
jo
u
r
n
al
u
s
es
th
e
C
o
n
tr
ib
u
to
r
R
o
les
T
ax
o
n
o
m
y
(
C
R
ed
iT)
to
r
ec
o
g
n
ize
in
d
iv
id
u
al
au
th
o
r
co
n
tr
ib
u
tio
n
s
,
r
ed
u
ce
au
th
o
r
s
h
ip
d
is
p
u
tes,
an
d
f
ac
ilit
ate
co
llab
o
r
atio
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Gen
etic
a
lg
o
r
ith
m
-
o
p
timiz
ed
B
E
R
To
p
ic
w
ith
S
HA
P
ex
p
la
in
a
b
ilit
y
fo
r
…
(
Mu
h
a
mma
d
De
d
i I
r
a
w
a
n
)
3825
Na
m
e
o
f
Aut
ho
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
Mu
h
am
m
ad
De
d
i I
r
awa
n
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
Yu
s
tr
ia
Han
d
ik
a
Sire
g
ar
✓
✓
✓
✓
✓
✓
✓
Hew
a
Ma
jeed
Z
an
g
an
a
✓
✓
✓
✓
✓
Ali I
k
h
wan
✓
✓
✓
✓
C
:
C
o
n
c
e
p
t
u
a
l
i
z
a
t
i
o
n
M
:
M
e
t
h
o
d
o
l
o
g
y
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
d
a
t
i
o
n
Fo
:
Fo
r
mal
a
n
a
l
y
s
i
s
I
:
I
n
v
e
s
t
i
g
a
t
i
o
n
R
:
R
e
so
u
r
c
e
s
D
:
D
a
t
a
C
u
r
a
t
i
o
n
O
:
W
r
i
t
i
n
g
-
O
r
i
g
i
n
a
l
D
r
a
f
t
E
:
W
r
i
t
i
n
g
-
R
e
v
i
e
w
&
E
d
i
t
i
n
g
Vi
:
Vi
su
a
l
i
z
a
t
i
o
n
Su
:
Su
p
e
r
v
i
s
i
o
n
P
:
P
r
o
j
e
c
t
a
d
mi
n
i
st
r
a
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
si
t
i
o
n
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
T
h
e
au
t
h
o
r
s
d
ec
lar
e
th
at
th
e
y
h
av
e
n
o
k
n
o
wn
c
o
m
p
etin
g
f
in
an
cial
in
ter
ests
o
r
p
er
s
o
n
al
r
el
atio
n
s
h
ip
s
th
at
co
u
ld
h
av
e
ap
p
ea
r
ed
t
o
in
f
lu
en
ce
th
e
wo
r
k
r
e
p
o
r
te
d
in
t
h
is
p
ap
er
.
I
NF
O
RM
E
D
CO
NS
E
N
T
I
n
f
o
r
m
ed
co
n
s
en
t
is
n
o
t
ap
p
licab
le
b
ec
au
s
e
t
h
is
s
tu
d
y
d
o
es
n
o
t
in
v
o
l
v
e
h
u
m
an
p
a
r
ticip
an
ts
o
r
p
er
s
o
n
al
d
ata.
E
T
H
I
CAL AP
P
RO
V
AL
E
th
ical
ap
p
r
o
v
al
was
n
o
t
r
e
q
u
ir
ed
b
ec
a
u
s
e
th
is
s
tu
d
y
u
s
ed
i
n
s
titu
tio
n
al
r
esear
ch
g
r
an
t
titl
e
d
ata
an
d
d
id
n
o
t i
n
v
o
lv
e
h
u
m
a
n
s
u
b
ject
s
o
r
s
en
s
itiv
e
p
er
s
o
n
al
in
f
o
r
m
a
tio
n
.
DATA AV
AI
L
AB
I
L
I
T
Y
T
h
e
d
ataset
u
s
ed
in
th
is
s
tu
d
y
is
av
ailab
le
f
r
o
m
t
h
e
co
r
r
esp
o
n
d
in
g
au
th
o
r
,
[
M
DI
]
,
u
p
o
n
r
ea
s
o
n
ab
le
r
eq
u
est.
RE
F
E
R
E
NC
E
S
[
1
]
A
.
P
a
e
z
,
“
G
r
a
y
l
i
t
e
r
a
t
u
r
e
:
a
n
i
m
p
o
r
t
a
n
t
r
e
s
o
u
r
c
e
i
n
sy
s
t
e
m
a
t
i
c
r
e
v
i
e
w
s,”
J
o
u
r
n
a
l
o
f
Ev
i
d
e
n
c
e
-
B
a
se
d
M
e
d
i
c
i
n
e
,
v
o
l
.
1
0
,
n
o
.
3
,
p
p
.
2
3
3
–
2
4
0
,
A
u
g
.
2
0
1
7
,
d
o
i
:
1
0
.
1
1
1
1
/
j
e
b
m.
1
2
2
6
6
.
[
2
]
J.
A
.
W
a
n
i
a
n
d
S
.
A
.
G
a
n
a
i
e
,
“
T
h
e
s
c
i
e
n
t
i
f
i
c
o
u
t
c
o
m
e
i
n
t
h
e
d
o
m
a
i
n
o
f
g
r
e
y
l
i
t
e
r
a
t
u
r
e
:
b
i
b
l
i
o
m
e
t
r
i
c
ma
p
p
i
n
g
a
n
d
v
i
s
u
a
l
i
sat
i
o
n
u
si
n
g
t
h
e
R
-
b
i
b
l
i
o
m
e
t
r
i
x
p
a
c
k
a
g
e
a
n
d
t
h
e
V
O
S
v
i
e
w
e
r
,
”
L
i
b
ra
r
y
H
i
T
e
c
h
,
v
o
l
.
4
2
,
n
o
.
1
,
p
p
.
3
0
9
–
3
3
0
,
F
e
b
.
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
8
/
LH
T
-
01
-
2
0
2
2
-
0
0
1
2
.
[
3
]
S
.
M
i
f
r
a
h
a
n
d
E.
H
.
B
e
n
l
a
h
m
a
r
,
“
To
p
i
c
mo
d
e
l
i
n
g
w
i
t
h
t
r
a
n
sf
o
r
mers
f
o
r
s
e
n
t
e
n
c
e
-
l
e
v
e
l
u
s
i
n
g
c
o
r
o
n
a
v
i
r
u
s
c
o
r
p
u
s,”
I
n
t
e
rn
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
I
n
t
e
ra
c
t
i
v
e
M
o
b
i
l
e
T
e
c
h
n
o
l
o
g
i
e
s
,
v
o
l
.
1
6
,
n
o
.
1
7
,
p
p
.
5
0
–
5
9
,
S
e
p
.
2
0
2
2
,
d
o
i
:
1
0
.
3
9
9
1
/
i
j
i
m.
v
1
6
i
1
7
.
3
3
2
8
1
.
[
4
]
A
.
Za
f
a
r
,
M
.
W
a
si
m
,
S
.
Z
u
l
f
i
q
a
r
,
T.
W
a
h
e
e
d
,
a
n
d
A
.
S
i
d
d
i
q
u
e
,
“
Tr
a
n
sf
o
r
mer
-
b
a
s
e
d
t
o
p
i
c
m
o
d
e
l
i
n
g
f
o
r
u
r
d
u
t
r
a
n
s
l
a
t
i
o
n
s
o
f
t
h
e
h
o
l
y
q
u
r
a
n
,
”
A
C
M
T
r
a
n
sa
c
t
i
o
n
s
o
n
Asi
a
n
a
n
d
L
o
w
-
R
e
so
u
rc
e
L
a
n
g
u
a
g
e
I
n
f
o
rm
a
t
i
o
n
Pro
c
e
ss
i
n
g
,
v
o
l
.
2
3
,
n
o
.
1
0
,
p
p
.
1
–
2
1
,
O
c
t
.
2
0
2
4
,
d
o
i
:
1
0
.
1
1
4
5
/
3
6
9
4
9
6
7
.
[
5
]
M
.
A
.
K
h
a
d
i
j
a
a
n
d
W
.
N
u
r
h
a
r
j
a
d
mo
,
“
E
n
h
a
n
c
i
n
g
I
n
d
o
n
e
si
a
n
c
u
st
o
mer
c
o
mp
l
a
i
n
t
a
n
a
l
y
si
s
:
LD
A
t
o
p
i
c
m
o
d
e
l
l
i
n
g
w
i
t
h
B
ER
T
e
mb
e
d
d
i
n
g
s,
”
S
I
N
ERG
I
,
v
o
l
.
2
8
,
n
o
.
1
,
p
p
.
1
5
3
–
1
6
2
,
D
e
c
.
2
0
2
3
,
d
o
i
:
1
0
.
2
2
4
4
1
/
si
n
e
r
g
i
.
2
0
2
4
.
1
.
0
1
5
.
[
6
]
M
.
G
r
o
o
t
e
n
d
o
r
s
t
,
“
B
E
R
T
o
p
i
c
:
n
e
u
r
a
l
t
o
p
i
c
m
o
d
e
l
i
n
g
w
i
t
h
a
c
l
a
ss
-
b
a
s
e
d
TF
-
I
D
F
p
r
o
c
e
d
u
r
e
,
”
M
a
r
.
2
0
2
2
,
a
rX
i
v
:
2
2
0
3
.
0
5
7
9
4
.
[
7
]
C
.
Y
a
n
g
a
n
d
Y
.
K
i
m,
“
En
h
a
n
c
i
n
g
t
o
p
i
c
c
o
h
e
r
e
n
c
e
a
n
d
d
i
v
e
r
s
i
t
y
i
n
d
o
c
u
men
t
e
m
b
e
d
d
i
n
g
s
u
s
i
n
g
L
LM
s
:
a
f
o
c
u
s
o
n
B
ER
To
p
i
c
,
”
Ex
p
e
rt
S
y
st
e
m
s w
i
t
h
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
2
8
1
,
J
u
l
.
2
0
2
5
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
e
sw
a
.
2
0
2
5
.
1
2
7
5
1
7
.
[
8
]
D
.
M
e
d
v
e
c
k
i
,
B
.
B
a
š
a
r
a
g
i
n
,
A
.
L
j
a
j
i
ć
,
a
n
d
N
.
M
i
l
o
še
v
i
ć
,
“
M
u
l
t
i
l
i
n
g
u
a
l
t
r
a
n
sf
o
r
m
e
r
a
n
d
B
ER
T
o
p
i
c
f
o
r
s
h
o
r
t
t
e
x
t
t
o
p
i
c
mo
d
e
l
i
n
g
:
t
h
e
c
a
s
e
o
f
S
e
r
b
i
a
n
,
”
i
n
C
o
n
f
e
re
n
c
e
o
n
I
n
f
o
rm
a
t
i
o
n
T
e
c
h
n
o
l
o
g
y
a
n
d
i
t
s
A
p
p
l
i
c
a
t
i
o
n
s
,
2
0
2
4
,
p
p
.
1
6
1
–
173
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
031
-
5
0
7
5
5
-
7
_
1
6
.
[
9
]
D
.
P
.
G
o
t
t
u
m
u
k
k
a
l
a
,
P
.
R
.
P
.
V
.
G
.
D
,
a
n
d
S
.
K
.
R
a
o
,
“
To
p
i
c
m
o
d
e
l
i
n
g
-
b
a
s
e
d
p
r
e
d
i
c
t
i
o
n
o
f
s
o
f
t
w
a
r
e
d
e
f
e
c
t
s
a
n
d
r
o
o
t
c
a
u
se
u
si
n
g
B
ER
To
p
i
c
,
a
n
d
mu
l
t
i
o
u
t
p
u
t
c
l
a
ss
i
f
i
e
r
,
”
S
c
i
e
n
t
i
f
i
c
R
e
p
o
r
t
s
,
v
o
l
.
1
5
,
n
o
.
1
,
J
u
l
.
2
0
2
5
,
d
o
i
:
1
0
.
1
0
3
8
/
s
4
1
5
9
8
-
025
-
1
1
4
5
8
-
0.
[
1
0
]
A
.
N
a
n
y
o
n
g
a
,
K
.
J
o
i
n
e
r
,
U
.
T
u
r
h
a
n
,
a
n
d
G
.
W
i
l
d
,
“
S
e
m
a
n
t
i
c
t
o
p
i
c
mo
d
e
l
i
n
g
o
f
a
v
i
a
t
i
o
n
saf
e
t
y
r
e
p
o
r
t
s:
a
c
o
m
p
a
r
a
t
i
v
e
a
n
a
l
y
s
i
s
u
si
n
g
B
ER
T
o
p
i
c
a
n
d
P
LSA
,
”
A
e
r
o
sp
a
c
e
,
v
o
l
.
1
2
,
n
o
.
6
,
J
u
n
.
2
0
2
5
,
d
o
i
:
1
0
.
3
3
9
0
/
a
e
r
o
s
p
a
c
e
1
2
0
6
0
5
5
1
.
[
1
1
]
H
.
S
o
n
a
n
d
Y
.
E.
P
a
r
k
,
“
A
g
e
n
d
a
-
set
t
i
n
g
e
f
f
e
c
t
s
f
o
r
C
O
V
I
D
-
1
9
v
a
c
c
i
n
a
t
i
o
n
:
i
n
s
i
g
h
t
s
f
r
o
m
1
0
mi
l
l
i
o
n
t
e
x
t
u
a
l
d
a
t
a
f
r
o
m
so
c
i
a
l
med
i
a
a
n
d
n
e
w
s
a
r
t
i
c
l
e
s
u
si
n
g
B
ER
T
o
p
i
c
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
I
n
f
o
rm
a
t
i
o
n
M
a
n
a
g
e
m
e
n
t
,
v
o
l
.
8
3
,
A
u
g
.
2
0
2
5
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
j
i
n
f
o
m
g
t
.
2
0
2
5
.
1
0
2
9
0
7
.
[
1
2
]
Y.
-
S
.
S
u
,
J.
W
a
n
g
,
S
.
-
H
.
T
u
,
K
.
-
T
.
L
i
a
o
,
a
n
d
C
.
-
L
.
Li
n
,
“
D
e
t
e
c
t
i
n
g
l
a
t
e
n
t
t
o
p
i
c
s
a
n
d
t
r
e
n
d
s
i
n
I
o
T
a
n
d
e
-
c
o
mm
e
r
c
e
u
si
n
g
B
ER
To
p
i
c
m
o
d
e
l
i
n
g
,
”
I
n
t
e
rn
e
t
o
f
T
h
i
n
g
s
,
v
o
l
.
3
2
,
Ju
l
.
2
0
2
5
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
o
t
.
2
0
2
5
.
1
0
1
6
0
4
.
[
1
3
]
M
.
K
i
l
i
n
c
,
F
.
G
u
r
c
a
n
,
a
n
d
A
.
S
o
y
l
u
,
“
LLM
-
b
a
s
e
d
g
e
n
e
r
a
t
i
v
e
A
I
i
n
me
d
i
c
i
n
e
:
a
n
a
l
y
si
s o
f
c
u
r
r
e
n
t
r
e
se
a
r
c
h
t
r
e
n
d
s wit
h
B
E
R
T
o
p
i
c
,
”
I
EEE
A
c
c
e
ss
,
v
o
l
.
1
3
,
p
p
.
1
5
7
0
2
6
–
1
5
7
0
4
3
,
2
0
2
5
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
5
.
3
6
0
6
3
3
5
.
[
1
4
]
H
.
R
.
-
B
e
r
u
m
e
n
,
A
.
Za
f
r
a
,
a
n
d
S
.
V
e
n
t
u
r
a
,
“
A
su
r
v
e
y
o
f
g
e
n
e
t
i
c
a
l
g
o
r
i
t
h
ms
f
o
r
c
l
u
st
e
r
i
n
g
:
t
a
x
o
n
o
m
y
a
n
d
e
m
p
i
r
i
c
a
l
a
n
a
l
y
s
i
s,
”
S
w
a
rm
a
n
d
Ev
o
l
u
t
i
o
n
a
ry
C
o
m
p
u
t
a
t
i
o
n
,
v
o
l
.
9
1
,
D
e
c
.
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
s
w
e
v
o
.
2
0
2
4
.
1
0
1
7
2
0
.
[
1
5
]
B
.
B
i
sc
h
l
e
t
a
l
.
,
“
H
y
p
e
r
p
a
r
a
me
t
e
r
o
p
t
i
mi
z
a
t
i
o
n
:
f
o
u
n
d
a
t
i
o
n
s,
a
l
g
o
r
i
t
h
ms,
b
e
s
t
p
r
a
c
t
i
c
e
s,
a
n
d
o
p
e
n
c
h
a
l
l
e
n
g
e
s,”
WIR
Es
D
a
t
a
Mi
n
i
n
g
a
n
d
K
n
o
w
l
e
d
g
e
D
i
sc
o
v
e
ry
,
v
o
l
.
1
3
,
n
o
.
2
,
M
a
r
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
2
/
w
i
d
m.
1
4
8
4
.
[
1
6
]
L.
Y
a
n
g
a
n
d
A
.
S
h
a
mi
,
“
O
n
h
y
p
e
r
p
a
r
a
met
e
r
o
p
t
i
mi
z
a
t
i
o
n
o
f
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
ms
:
t
h
e
o
r
y
a
n
d
p
r
a
c
t
i
c
e
,
”
N
e
u
ro
c
o
m
p
u
t
i
n
g
,
v
o
l
.
4
1
5
,
p
p
.
2
9
5
–
3
1
6
,
N
o
v
.
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
n
e
u
c
o
m
.
2
0
2
0
.
0
7
.
0
6
1
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
4
,
Au
g
u
s
t 2
0
2
6
:
3
8
1
8
-
3
8
2
6
3826
[
1
7
]
S
.
M
.
L
u
n
d
b
e
r
g
a
n
d
S
.
-
I
.
Le
e
,
“
A
u
n
i
f
i
e
d
a
p
p
r
o
a
c
h
t
o
i
n
t
e
r
p
r
e
t
i
n
g
mo
d
e
l
p
r
e
d
i
c
t
i
o
n
s
,
”
A
d
v
a
n
c
e
s
i
n
N
e
u
r
a
l
I
n
f
o
rm
a
t
i
o
n
Pro
c
e
ssi
n
g
S
y
s
t
e
m
s
,
p
p
.
4
7
6
8
–
4
7
7
7
,
2
0
1
7
.
[
1
8
]
R
.
M
u
t
h
u
sam
i
,
K
.
S
a
r
i
t
h
a
,
K
.
S
.
R
a
o
,
P
.
S
u
g
a
p
r
i
y
a
,
a
n
d
G
.
S
a
v
e
e
t
h
a
,
“
I
n
t
e
r
p
r
e
t
a
b
l
e
s
t
a
n
c
e
d
e
t
e
c
t
i
o
n
i
n
so
c
i
a
l
med
i
a
v
i
a
t
o
p
i
c
-
g
u
i
d
e
d
t
r
a
n
sf
o
r
mers
,
”
D
i
s
c
o
v
e
r
Art
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
,
v
o
l
.
5
,
n
o
.
1
,
N
o
v
.
2
0
2
5
,
d
o
i
:
1
0
.
1
0
0
7
/
s
4
4
1
6
3
-
0
2
5
-
0
0
6
3
5
-
9.
[
1
9
]
L.
M
a
e
t
a
l
.
,
“
A
I
-
p
o
w
e
r
e
d
t
o
p
i
c
mo
d
e
l
i
n
g
:
c
o
m
p
a
r
i
n
g
LD
A
a
n
d
B
ER
T
o
p
i
c
i
n
a
n
a
l
y
z
i
n
g
o
p
i
o
i
d
-
r
e
l
a
t
e
d
c
a
r
d
i
o
v
a
sc
u
l
a
r
r
i
s
k
s
i
n
w
o
me
n
,
”
Ex
p
e
ri
m
e
n
t
a
l
B
i
o
l
o
g
y
a
n
d
Me
d
i
c
i
n
e
,
v
o
l
.
2
5
0
,
F
e
b
.
2
0
2
5
,
d
o
i
:
1
0
.
3
3
8
9
/
e
b
m.
2
0
2
5
.
1
0
3
8
9
.
[
2
0
]
H
.
O
k
a
z
a
k
i
a
n
d
H
.
Ta
k
a
h
a
s
h
i
,
“
A
n
a
l
y
si
s
o
f
e
x
t
r
a
c
t
i
n
g
c
o
r
p
o
r
a
t
e
t
r
e
n
d
s
f
r
o
m
n
e
w
s
d
a
t
a
u
s
i
n
g
B
ER
To
p
i
c
,
”
N
e
w
G
e
n
e
r
a
t
i
o
n
C
o
m
p
u
t
i
n
g
,
v
o
l
.
4
3
,
n
o
.
1
,
F
e
b
.
2
0
2
5
,
d
o
i
:
1
0
.
1
0
0
7
/
s
0
0
3
5
4
-
025
-
0
0
2
9
0
-
9.
[
2
1
]
K.
-
K
.
La
i
,
C
.
-
W
.
H
s
i
a
o
,
a
n
d
Y
.
-
J
.
H
su
,
“
S
t
r
a
t
e
g
i
c
ma
n
a
g
e
me
n
t
k
n
o
w
l
e
d
g
e
ma
p
v
i
a
B
ER
T
o
p
i
c
(
1
9
8
0
–
2
0
2
5
)
:
e
v
o
l
u
t
i
o
n
,
i
n
t
e
g
r
a
t
i
o
n
,
a
n
d
a
p
p
l
i
c
a
t
i
o
n
,
”
A
p
p
l
i
e
d
S
y
s
t
e
m
I
n
n
o
v
a
t
i
o
n
,
v
o
l
.
8
,
n
o
.
5
,
S
e
p
.
2
0
2
5
,
d
o
i
:
1
0
.
3
3
9
0
/
a
s
i
8
0
5
0
1
4
2
.
[
2
2
]
A
.
N
a
n
y
o
n
g
a
,
K
.
Jo
i
n
e
r
,
U
.
T
u
r
h
a
n
,
a
n
d
G
.
W
i
l
d
,
“
D
o
e
s
t
h
e
c
h
o
i
c
e
o
f
t
o
p
i
c
m
o
d
e
l
i
n
g
t
e
c
h
n
i
q
u
e
i
m
p
a
c
t
t
h
e
i
n
t
e
r
p
r
e
t
a
t
i
o
n
o
f
a
v
i
a
t
i
o
n
i
n
c
i
d
e
n
t
r
e
p
o
r
t
s?
a
met
h
o
d
o
l
o
g
i
c
a
l
a
sse
ssm
e
n
t
,
”
T
e
c
h
n
o
l
o
g
i
e
s
,
v
o
l
.
1
3
,
n
o
.
5
,
M
a
y
2
0
2
5
,
d
o
i
:
1
0
.
3
3
9
0
/
t
e
c
h
n
o
l
o
g
i
e
s
1
3
0
5
0
2
0
9
.
[
2
3
]
A
.
K
a
u
r
a
n
d
J.
W
a
l
l
a
c
e
,
“
F
r
o
m
t
r
a
n
sf
o
r
mers
c
o
me
t
h
e
mes:
e
v
a
l
u
a
t
i
n
g
B
ER
To
p
i
c
f
o
r
q
u
a
l
i
t
a
t
i
v
e
a
n
a
l
y
s
i
s
o
f
s
o
c
i
a
l
m
e
d
i
a
d
a
t
a
,
”
i
n
2
0
2
5
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
H
u
m
a
n
-
E
n
g
a
g
e
d
C
o
m
p
u
t
i
n
g
,
N
o
v
.
2
0
2
5
,
p
p
.
1
–
14
,
d
o
i
:
1
0
.
1
1
4
5
/
3
7
8
6
9
9
5
.
3
7
8
6
9
9
7
.
[
2
4
]
R
.
Es
p
i
n
o
z
a
e
t
a
l
.
,
“
Le
v
e
r
a
g
i
n
g
s
o
c
i
a
l
me
d
i
a
f
o
r
p
u
b
l
i
c
h
e
a
l
t
h
:
N
LP
i
mp
l
e
men
t
a
t
i
o
n
s
f
o
r
b
l
o
o
d
d
o
n
a
t
i
o
n
d
a
t
a
a
n
a
l
y
s
i
s
i
n
Ja
p
a
n
,
”
S
o
c
i
a
l
N
e
t
w
o
r
k
A
n
a
l
y
si
s
a
n
d
M
i
n
i
n
g
,
v
o
l
.
1
5
,
n
o
.
1
,
A
p
r
.
2
0
2
5
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
3
2
7
8
-
0
2
5
-
0
1
4
3
7
-
8.
[
2
5
]
A
.
A
l
a
ms
y
a
h
a
n
d
R
.
F
.
A
s
t
u
t
i
,
“
A
n
a
l
y
z
i
n
g
p
u
b
l
i
c
d
i
sc
o
u
r
s
e
o
n
D
e
F
i
a
n
d
C
B
D
C
u
si
n
g
a
d
v
a
n
c
e
d
N
LP
t
e
c
h
n
i
q
u
e
s
:
i
n
s
i
g
h
t
s
f
o
r
f
i
n
a
n
c
i
a
l
p
o
l
i
c
y
a
n
d
i
n
n
o
v
a
t
i
o
n
,
”
D
i
g
i
t
a
l
P
o
l
i
c
y
,
Re
g
u
l
a
t
i
o
n
a
n
d
G
o
v
e
r
n
a
n
c
e
,
v
o
l
.
2
7
,
n
o
.
6
,
p
p
.
7
0
3
–
7
2
3
,
O
c
t
.
2
0
2
5
,
d
o
i
:
1
0
.
1
1
0
8
/
D
P
R
G
-
09
-
2
0
2
4
-
0
2
4
0
.
[
2
6
]
R
.
E
g
g
e
r
a
n
d
J.
Y
u
,
“
A
t
o
p
i
c
m
o
d
e
l
i
n
g
c
o
mp
a
r
i
s
o
n
b
e
t
w
e
e
n
LD
A
,
N
M
F
,
To
p
2
V
e
c
,
a
n
d
B
E
R
T
o
p
i
c
t
o
d
e
my
s
t
i
f
y
t
w
i
t
t
e
r
p
o
s
t
s,”
Fro
n
t
i
e
rs
i
n
S
o
c
i
o
l
o
g
y
,
v
o
l
.
7
,
M
a
y
2
0
2
2
,
d
o
i
:
1
0
.
3
3
8
9
/
f
s
o
c
.
2
0
2
2
.
8
8
6
4
9
8
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Mu
h
a
m
m
a
d
De
d
i
Ir
a
wa
n
is
a
lec
tu
re
r
a
t
th
e
De
p
a
rtme
n
t
o
f
Da
ta
S
c
ien
c
e
,
F
a
c
u
lt
y
o
f
S
c
ien
c
e
a
n
d
Tec
h
n
o
lo
g
y
,
Un
i
v
e
rsitas
Isla
m
Ne
g
e
ri
S
u
m
a
tera
Uta
ra
,
In
d
o
n
e
sia
.
He
re
c
e
iv
e
d
b
a
c
h
e
lo
r’s
a
n
d
m
a
ste
r’s
d
e
g
re
e
s
in
I
n
fo
rm
a
ti
c
s
fr
o
m
Un
i
v
e
rsitas
As
a
h
a
n
a
n
d
Un
i
v
e
rsitas
P
u
tra
In
d
o
n
e
sia
YPT
K
P
a
d
a
n
g
,
re
sp
e
c
ti
v
e
ly
.
His
re
se
a
rc
h
in
tere
sts
i
n
c
lu
d
e
a
rti
ficia
l
i
n
telli
g
e
n
c
e
,
d
e
c
isio
n
su
p
p
o
r
t
s
y
ste
m
s,
in
f
o
r
m
a
ti
o
n
tec
h
n
o
lo
g
y
,
a
n
d
c
o
m
p
u
t
e
r
sc
ien
c
e
.
He
h
a
s
p
u
b
li
sh
e
d
v
a
rio
u
s
sc
ien
ti
fic
p
a
p
e
rs
in
c
o
m
p
u
ti
n
g
a
n
d
i
n
fo
rm
a
ti
o
n
sy
ste
m
s
a
n
d
a
c
ti
v
e
ly
p
a
rti
c
i
p
a
tes
i
n
a
c
a
d
e
m
ic
re
se
a
rc
h
a
n
d
c
o
m
m
u
n
it
y
se
rv
ice
p
r
o
jec
ts.
He
h
a
s
a
lso
b
e
e
n
i
n
v
o
lv
e
d
in
n
a
ti
o
n
a
ll
y
fu
n
d
e
d
re
se
a
rc
h
a
c
ti
v
it
ies
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
m
d
.
iraw
a
n
@u
in
su
.
a
c
.
id
.
Yustr
ia
H
a
n
d
ik
a
S
ire
g
a
r
is
a
lec
tu
re
r
i
n
t
h
e
C
o
m
p
u
ter
G
ra
p
h
ic
En
g
i
n
e
e
rin
g
P
ro
g
ra
m
a
t
P
o
li
tek
n
ik
Ce
n
d
a
n
a
,
In
d
o
n
e
sia
.
S
h
e
re
c
e
iv
e
d
b
a
c
h
e
l
o
r’s
a
n
d
m
a
ste
r’s
d
e
g
re
e
s
i
n
In
fo
rm
a
ti
c
s
fro
m
Un
i
v
e
rsitas
As
a
h
a
n
a
n
d
Un
i
v
e
rsitas
P
u
tra
In
d
o
n
e
sia
YPT
K
P
a
d
a
n
g
,
re
sp
e
c
ti
v
e
ly
.
He
r
re
se
a
rc
h
in
tere
sts in
c
lu
d
e
a
rti
ficia
l
i
n
telli
g
e
n
c
e
,
d
e
c
isio
n
su
p
p
o
rt
sy
ste
m
s,
ima
g
e
p
ro
c
e
ss
in
g
,
a
n
d
c
o
m
p
u
t
in
g
tec
h
n
o
lo
g
y
.
S
h
e
h
a
s
c
o
n
tri
b
u
te
d
to
se
v
e
ra
l
sc
ien
ti
fic
p
u
b
li
c
a
ti
o
n
s
in
in
fo
rm
a
ti
o
n
tec
h
n
o
lo
g
y
a
n
d
h
a
s
b
e
e
n
a
c
ti
v
e
ly
in
v
o
lv
e
d
in
re
se
a
rc
h
a
n
d
c
o
m
m
u
n
i
ty
se
rv
ice
a
c
ti
v
it
ies
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
y
u
striah
a
n
d
i
k
a
sire
g
a
r@g
m
a
il
.
c
o
m
.
H
e
wa
Ma
je
e
d
Za
n
g
a
n
a
is
a
n
a
ss
istan
t
p
ro
fe
ss
o
r
a
t
De
p
a
rtme
n
t
o
f
In
fo
rm
a
ti
o
n
Tec
h
n
o
l
o
g
y
,
Du
h
o
k
Tec
h
n
ica
l
Co
ll
e
g
e
,
D
u
h
o
k
P
o
ly
tec
h
n
ic
U
n
iv
e
rsit
y
,
Ira
q
,
a
n
d
c
u
rre
n
tl
y
p
u
rsu
i
n
g
a
P
h
.
D
.
i
n
De
p
a
rtme
n
t
o
f
I
n
fo
rm
a
ti
o
n
Tec
h
n
o
lo
g
y
M
a
n
a
g
e
m
e
n
t
a
t
Du
h
o
k
P
o
l
y
tec
h
n
ic
Un
iv
e
rsity
.
His
re
se
a
rc
h
in
tere
sts
in
c
lu
d
e
n
e
two
r
k
sy
ste
m
s,
in
fo
rm
a
ti
o
n
se
c
u
rit
y
,
m
o
b
il
e
c
o
m
m
u
n
ica
ti
o
n
,
i
n
telli
g
e
n
t
sy
ste
m
s,
a
n
d
d
e
e
p
lea
rn
in
g
a
p
p
li
c
a
ti
o
n
s.
He
h
a
s h
e
ld
se
v
e
ra
l
a
c
a
d
e
m
ic
a
n
d
a
d
m
i
n
istrativ
e
p
o
siti
o
n
s
a
t
Du
h
o
k
P
o
l
y
tec
h
n
ic
Un
i
v
e
rsity
a
n
d
Na
wro
z
Un
i
v
e
rsity
.
He
h
a
s
c
o
n
tri
b
u
ted
t
o
sc
ien
t
ifi
c
p
u
b
li
c
a
ti
o
n
s
i
n
c
y
b
e
rse
c
u
rit
y
,
ima
g
e
p
ro
c
e
ss
in
g
,
a
n
d
a
rti
ficia
l
in
telli
g
e
n
c
e
re
se
a
rc
h
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
h
e
wa
.
z
a
n
g
a
n
a
@d
p
u
.
e
d
u
.
k
r
d
.
Ali
Ikh
w
a
n
He
e
a
rn
e
d
h
is
b
a
c
h
e
lo
r’s
d
e
g
re
e
i
n
I
n
fo
rm
a
ti
o
n
S
y
ste
m
s
fro
m
S
TM
IK
Tri
g
u
n
a
Dh
a
rm
a
,
M
e
d
a
n
,
In
d
o
n
e
sia
,
in
2
0
1
3
.
He
re
c
e
iv
e
d
h
is
m
a
ste
r’s
d
e
g
re
e
in
C
o
m
p
u
ter
S
c
ien
c
e
fro
m
Un
iv
e
rsitas
P
u
tra
In
d
o
n
e
sia
YPT
K,
P
a
d
a
n
g
,
I
n
d
o
n
e
sia
,
in
2
0
1
5
.
He
is
c
u
rre
n
tl
y
p
u
rsu
i
n
g
a
d
o
c
to
ra
l
d
e
g
re
e
a
t
Un
i
v
e
rsiti
M
a
lay
sia
P
e
rli
s,
M
a
lay
sia
,
with
a
re
se
a
rc
h
fo
c
u
s
o
n
d
a
ta
m
in
in
g
a
n
d
c
ry
p
to
g
ra
p
h
y
.
He
is
a
lec
tu
re
r
a
t
th
e
De
p
a
rtm
e
n
t
o
f
In
fo
rm
a
ti
o
n
S
y
ste
m
s,
F
a
c
u
lt
y
o
f
S
c
ien
c
e
a
n
d
Tec
h
n
o
l
o
g
y
,
Un
iv
e
rsitas
Isla
m
Ne
g
e
ri
S
u
m
a
tera
U
tara
,
In
d
o
n
e
sia
.
His
re
se
a
rc
h
in
tere
sts
in
c
lu
d
e
d
a
ta
a
n
a
ly
sis,
d
a
ta
m
in
in
g
,
a
n
d
c
ry
p
t
o
g
ra
p
h
y
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
a
li
_
ik
h
wa
n
@u
i
n
su
.
a
c
.
i
d
.
Evaluation Warning : The document was created with Spire.PDF for Python.