I
n
d
on
e
s
ian
Jou
r
n
al
o
f
E
lec
t
r
ica
l
E
n
gin
e
e
r
in
g
a
n
d
Com
p
u
t
e
r
S
c
ience
Vo
l
.
25
,
N
o
.
2
,
F
e
b
r
ua
r
y
20
22
,
pp.
1131
~
1139
I
S
S
N:
2502
-
4752,
DO
I
:
10
.
11591/i
j
e
e
c
s
.
v
25
.i
2
.
pp
1131
-
1139
1131
Jou
r
n
al
h
o
m
e
page
:
ht
tp:
//
ij
e
e
c
s
.
iaes
c
or
e
.
c
om
C
r
im
e
p
r
e
d
ic
t
io
n
u
s
in
g a
h
y
b
r
id
se
n
t
i
m
e
n
t
an
al
y
si
s ap
p
r
oac
h
b
ase
d
on
t
h
e
b
i
d
ir
e
c
t
io
n
al
e
n
c
od
e
r
r
e
p
r
e
se
n
t
a
t
io
n
s f
r
om
t
r
an
sf
or
m
e
r
s
M
oh
am
m
e
d
B
ou
k
ab
ou
s
,
M
os
t
af
a
Az
i
z
i
M
A
T
S
I
R
e
s
e
a
r
c
h
L
a
b,
E
c
ol
e
S
upé
r
ie
u
r
e
de
T
e
c
hn
o
l
o
gi
e
O
uj
da
(
E
S
T
O
)
, M
o
ha
mm
e
d F
i
r
s
t
U
ni
ve
r
s
it
y
, O
uj
da
, M
or
oc
c
o
Ar
t
ic
l
e
I
n
f
o
AB
S
T
RA
CT
A
r
ti
c
le
h
is
tor
y
:
R
e
c
e
i
ve
d
J
u
l
28
,
2021
R
e
vi
s
e
d
No
v
12
,
2021
A
c
c
e
pt
e
d
De
c
1
,
2021
Sen
t
i
m
en
t
a
n
al
y
s
i
s
(SA
)
is
w
i
d
e
l
y
u
s
e
d
t
o
d
a
y
in
m
an
y
are
as
s
u
ch
as
c
ri
me
d
e
t
ec
t
i
o
n
(s
e
cu
ri
t
y
i
n
t
el
l
i
g
en
c
e
)
to
d
e
t
ec
t
p
o
t
e
n
t
i
a
l
s
ecu
ri
t
y
t
h
r
e
at
s
in
r
e
a
l
-
t
i
me
u
s
i
n
g
s
o
c
i
a
l
me
d
i
a
p
l
at
fo
r
m
s
s
u
c
h
as
T
w
i
t
t
e
r
.
T
h
e
mo
s
t
p
ro
m
i
s
i
n
g
t
e
ch
n
i
q
u
e
s
in
s
e
n
t
i
men
t
an
a
l
y
s
i
s
ar
e
t
h
o
s
e
of
d
ee
p
l
e
arn
i
n
g
(
D
L
)
,
p
art
i
cu
l
ar
l
y
b
i
d
i
r
ec
t
i
o
n
al
e
n
co
d
e
r
r
e
p
re
s
e
n
t
at
i
o
n
s
fro
m
t
ran
s
fo
r
me
rs
(BE
RT
)
in
the
fi
el
d
of
n
at
u
ral
l
an
g
u
a
g
e
p
ro
ce
s
s
i
n
g
(
N
L
P).
H
o
w
ev
e
r
,
em
p
l
o
y
i
n
g
t
h
e
B
E
RT
al
g
o
r
i
t
h
m
to
d
e
t
ec
t
c
r
i
me
s
re
q
u
i
r
e
s
a
c
r
i
me
d
at
as
e
t
l
ab
e
l
ed
by
t
h
e
l
e
x
i
co
n
-
b
as
e
d
ap
p
ro
ac
h
.
In
t
h
i
s
p
ap
e
r,
we
u
s
e
d
a
h
y
b
r
i
d
ap
p
ro
ach
t
h
at
c
o
m
b
i
n
e
s
b
o
t
h
l
e
x
i
co
n
-
b
as
ed
an
d
d
ee
p
l
e
arn
i
n
g
,
w
i
t
h
BE
RT
as
the
DL
mo
d
el
.
We
em
p
l
o
y
ed
the
l
e
x
i
c
o
n
-
b
as
e
d
ap
p
ro
ach
to
l
ab
e
l
our
T
w
i
t
t
e
r
d
at
as
e
t
w
i
t
h
a
s
e
t
of
n
o
r
m
al
an
d
c
r
i
me
-
r
e
l
at
e
d
l
e
x
i
c
o
n
s
;
t
h
en
,
we
u
s
ed
t
h
e
o
b
t
ai
n
e
d
l
ab
e
l
ed
d
at
as
e
t
to
t
rai
n
our
BE
RT
m
o
d
el
.
T
h
e
e
x
p
e
ri
men
t
al
re
s
u
l
t
s
s
h
o
w
t
h
at
our
h
y
b
ri
d
t
ech
n
i
q
u
e
o
u
t
p
e
rf
o
r
m
s
e
x
i
s
t
i
n
g
ap
p
ro
a
c
h
e
s
in
s
e
v
e
ra
l
me
t
ri
c
s
,
w
i
t
h
9
4
.
9
1
%
an
d
9
4
.
9
2
%
in
a
ccu
ra
cy
an
d
F1
-
s
c
o
r
e
r
e
s
p
ec
t
i
v
el
y
.
K
e
y
w
o
r
d
s
:
B
E
R
T
C
r
i
m
e
t
e
x
t
-
de
t
e
c
t
i
o
n
De
e
p
l
e
a
r
ni
ng
Na
t
ur
a
l
l
a
n
gua
ge
pr
o
c
e
s
s
i
n
g
S
e
c
ur
i
t
y
i
n
t
e
l
li
ge
n
c
e
S
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
Th
i
s
i
s
a
n
o
p
en
a
c
ces
s
a
r
t
i
c
l
e
u
n
d
e
r
t
h
e
CC
B
Y
-
SA
l
i
cen
s
e.
C
or
r
e
s
pon
din
g
A
u
th
or
:
M
o
h
a
m
m
e
d
B
o
uka
b
o
us
M
A
T
S
I
R
e
s
e
a
r
c
h
L
a
b
,
E
c
o
l
e
S
upé
r
i
e
ur
e
de
T
e
c
hno
l
o
g
i
e
Ou
j
d
a
(
E
S
T
O
)
,
M
o
h
a
m
m
e
d
F
i
r
s
t
Uni
ve
r
s
i
t
y
BP
473
c
o
m
p
l
e
xe
Uni
ve
r
s
i
t
a
i
r
e
Al
Qo
ds
,
Ou
j
da
60000,
M
o
r
o
c
c
o
E
m
a
i
l
:
m
.
b
o
uka
b
o
us
@
u
m
p.
a
c
.
m
a
1.
I
NT
RODU
C
T
I
ON
S
e
c
u
r
i
t
y
i
n
te
l
l
i
ge
n
c
e
i
s
n
o
wa
da
y
s
o
n
e
o
f
t
h
e
pr
i
m
a
r
y
c
o
n
c
e
r
n
s
o
f
a
ny
c
oun
tr
y
.
I
t
i
s
a
n
or
i
e
n
t
a
t
i
o
n
o
f
s
e
c
u
r
i
t
y
t
e
c
h
ni
que
s
b
a
s
e
d
o
n
a
r
t
i
f
i
c
i
a
l
i
n
t
e
l
l
i
ge
n
c
e
th
a
t
a
i
m
s
to
c
o
l
l
e
c
t
a
n
d
or
g
a
ni
z
e
a
l
l
da
ta
r
e
l
a
ted
to
th
r
e
a
ts
i
n
c
y
b
e
r
s
pa
c
e
[
1]
to
d
e
te
c
t
p
os
s
i
bl
e
a
tt
a
c
ks
i
n
r
e
a
l
-
t
i
m
e
a
n
d
f
o
c
us
on
i
de
n
t
i
f
y
i
n
g
c
r
i
m
i
n
a
l
c
h
a
r
a
c
ter
i
s
t
i
c
s
[
2]
.
T
h
i
s
r
e
qui
r
e
s
d
e
e
pl
y
i
de
n
t
i
f
y
i
n
g
a
n
d
a
n
a
l
y
z
i
n
g
b
ot
h
p
a
tt
e
r
n
s
a
n
d
tr
e
n
ds
o
f
c
r
i
m
e
[
3]
.
S
uc
h
p
r
o
f
i
l
i
n
g
a
l
l
o
ws
r
e
l
e
v
a
n
t
p
r
ote
c
t
i
o
n
a
n
d
a
n
t
i
c
i
pa
t
i
n
g
t
h
e
v
a
r
i
ous
i
n
c
i
de
n
t
s
by
e
a
r
l
i
e
r
de
tec
t
i
o
n
o
f
m
a
j
or
a
tt
a
c
ks
.
I
n
tod
a
y
's
w
or
l
d,
i
n
d
i
vi
dua
l
s
,
i
n
c
l
ud
i
n
g
c
y
b
e
r
c
r
i
m
i
n
a
l
s
,
h
a
v
e
b
e
c
o
m
e
te
c
h
ni
c
a
l
l
y
s
o
phi
s
t
i
c
a
t
e
d
a
n
d
r
e
pe
a
tedl
y
e
x
p
r
e
s
s
i
n
g
t
h
e
i
r
e
m
ot
i
o
n
s
o
n
t
h
e
we
b
,
e
s
p
e
c
i
a
l
l
y
o
n
p
op
u
l
a
r
s
oc
i
a
l
m
e
d
i
a
we
b
s
i
t
e
s
[
4]
.
T
h
e
i
n
t
e
r
n
e
t
ph
e
n
o
m
e
n
a
l
gr
o
w
th
h
a
s
l
e
d
m
or
e
us
e
r
s
to
e
x
p
r
e
s
s
th
e
i
r
op
i
ni
o
n
s
o
nl
i
n
e
wh
e
t
h
e
r
i
n
da
i
l
y
c
h
a
t
(
s
ta
tus
m
e
s
s
a
g
e
s
a
b
ou
t
wh
a
t
th
e
us
e
r
i
s
d
o
i
n
g
)
,
c
o
n
v
e
r
s
a
t
i
o
n
s
(
tw
e
e
t
i
n
g
to
a
us
e
r
or
g
r
oup
o
f
us
e
r
s
wi
t
h
i
n
a
c
o
m
m
uni
t
y
)
,
i
nf
or
m
a
t
i
o
n
s
h
a
r
i
n
g
(
p
os
t
i
n
g
l
i
n
ks
to
w
e
b
pa
ge
s
)
,
or
n
e
ws
r
e
p
or
t
i
n
g
(
up
da
tes
o
n
t
h
e
c
u
r
r
e
n
t
c
a
s
e
)
.
T
r
a
n
s
f
e
r
l
e
a
r
ni
ng
ha
s
b
e
c
o
m
e
o
n
e
o
f
t
h
e
m
o
s
t
us
e
d
i
m
a
ge
c
l
a
s
s
if
i
c
a
t
i
o
n
a
ppr
o
a
c
h
e
s
f
o
r
r
e
us
i
n
g
a
r
c
hi
t
e
c
t
ur
e
s
a
n
d
w
e
i
g
h
t
s
l
e
a
r
n
e
d
o
n
h
uge
da
t
a
s
e
t
s
to
e
nh
a
n
c
e
s
m
a
ll
a
n
d
pa
r
t
i
c
u
l
a
r
c
l
a
s
s
if
i
c
a
t
i
o
n
t
a
s
k
s
[
5]
.
A
s
i
mi
l
a
r
o
ut
c
o
m
e
m
a
y
b
e
a
c
hi
e
v
e
d
i
n
n
a
t
ur
a
l
l
a
n
gua
ge
pr
o
c
e
s
s
i
ng
(
N
L
P
)
by
r
e
us
i
ng
a
n
d
t
r
a
n
s
f
e
r
r
i
n
g
a
l
a
n
gua
ge
m
o
de
l
[
6]
.
T
h
e
bi
d
i
r
e
c
t
i
o
n
a
l
e
n
c
o
de
r
r
e
pr
e
s
e
n
t
a
t
i
o
ns
f
r
o
m
t
r
a
n
s
f
o
r
m
e
r
s
(
B
E
R
T
)
a
l
go
r
i
t
hm
,
i
n
pa
r
t
i
c
u
l
a
r
,
h
a
s
b
e
e
n
de
m
o
ns
t
r
a
t
e
d
to
pe
r
f
o
r
m
we
ll
o
n
a
v
a
r
i
e
t
y
o
f
E
n
g
l
i
s
h
t
e
x
t
c
a
t
e
g
or
i
z
a
t
i
o
n
t
a
s
ks
,
s
uc
h
a
s
s
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
[
7]
.
S
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
us
i
ng
NL
P
a
n
d
d
e
e
p
l
e
a
r
ni
ng
(
DL
)
m
e
t
h
o
ds
h
a
s
r
e
c
e
n
t
l
y
be
e
n
a
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2502
-
4752
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
,
Vo
l
.
25
,
N
o
.
2
,
F
e
b
r
ua
r
y
20
22
:
1131
-
1139
1132
h
o
t
s
t
udy
s
u
bj
e
c
t
.
E
m
o
t
i
o
n
s
a
r
e
o
n
e
o
f
t
h
e
m
o
s
t
i
m
po
r
t
a
n
t
c
o
m
po
ne
n
t
s
o
f
h
u
m
a
n
e
xi
s
t
e
n
c
e
t
h
a
t
m
a
y
a
i
d
i
n
t
h
e
i
d
e
n
t
i
f
i
c
a
t
i
o
n
o
f
t
r
e
n
ds
a
n
d
us
e
r
n
e
c
e
s
s
i
t
i
e
s
.
DL
i
s
a
s
u
bf
i
e
l
d
o
f
m
a
c
hi
ne
l
e
a
r
ni
ng
t
h
a
t
a
l
l
o
ws
m
a
c
hi
ne
s
t
o
a
u
to
m
a
t
i
c
a
ll
y
li
nk
pr
o
c
e
s
s
e
s
to
ge
t
h
e
r
,
by
a
ll
o
w
i
ng
s
e
v
e
r
a
l
a
l
go
r
i
t
hm
s
t
o
b
e
us
e
d
pr
o
gr
e
s
s
i
ve
ly
w
hil
e
p
a
s
s
i
n
g
f
r
o
m
s
t
e
p
t
o
s
t
e
p,
i
t
c
a
n
s
o
l
v
e
c
o
m
p
l
e
x
pr
o
bl
e
m
s
i
n
a
lm
o
s
t
t
h
e
s
a
m
e
w
a
y
t
h
e
h
u
m
a
n
b
r
a
i
n
do
e
s
.
B
E
R
T
i
s
a
m
o
de
l
f
o
r
l
e
a
r
ni
ng
li
ngu
i
s
t
i
c
r
e
pr
e
s
e
n
t
a
t
i
o
ns
t
h
a
t
us
e
t
h
e
a
t
t
e
n
t
i
o
n
t
r
a
n
s
f
o
r
m
e
r
s
m
e
t
h
o
d.
A
D
L
m
o
de
l
de
s
i
g
n
e
d
to
h
a
n
d
l
e
s
e
que
n
t
i
a
l
da
t
a
,
to
l
e
a
r
n
t
h
e
w
o
r
d
-
to
-
w
o
r
d
c
o
n
t
e
x
t
ua
l
r
e
l
a
t
i
o
n
i
n
a
t
e
x
t
[
8]
.
T
h
e
r
e
s
t
o
f
t
hi
s
pa
pe
r
i
s
s
t
r
uc
t
ur
e
d
a
s
f
o
l
l
o
w
s
.
I
n
s
e
c
t
i
o
n
2,
we
d
i
s
c
us
s
r
e
l
e
v
a
n
t
r
e
l
a
t
e
d
wo
r
ks
.
Our
m
e
t
h
o
do
l
o
g
y
i
s
pr
e
s
e
n
t
e
d
i
n
s
e
c
t
i
o
n
3.
E
x
pe
r
i
m
e
nt
s
a
n
d
r
e
s
u
l
t
s
a
r
e
de
s
c
r
i
be
d
i
n
s
e
c
t
i
o
n
4.
F
i
na
ll
y
,
s
e
c
t
i
o
n
5
c
o
n
c
l
ude
s
t
h
e
s
t
udy
a
n
d
s
ugg
e
s
t
s
s
o
m
e
a
r
e
a
s
f
o
r
f
u
t
ur
e
r
e
s
e
a
r
c
h
.
2.
RE
L
AT
E
D
WORK
Ve
n
t
i
r
o
z
o
s
e
t
al.
[
9]
a
ppl
i
e
d
a
l
e
xi
c
o
n
-
ba
s
e
d
a
ppr
o
a
c
h
to
de
t
e
c
t
a
ggr
e
s
s
i
v
e
,
a
n
t
i
s
o
c
i
a
l
,
o
r
i
na
ppr
o
pr
i
a
t
e
be
h
a
vi
o
r
,
w
i
t
hi
n
t
h
e
c
o
n
t
e
x
t
o
f
t
h
e
d
i
s
c
us
s
i
o
n
.
T
h
e
a
ut
h
o
r
s
e
m
p
l
o
y
S
A
a
t
t
h
e
m
e
s
s
a
ge
l
e
v
e
l
,
b
ut
s
t
udy
t
h
e
w
h
o
l
e
c
o
m
m
u
ni
c
a
t
i
o
n
t
h
r
e
a
d
us
i
ng
a
s
e
t
o
f
n
-
gr
a
m
,
w
hi
c
h
i
s
us
e
d
f
o
r
c
l
a
s
s
if
y
i
ng
t
h
e
w
h
o
l
e
t
h
r
e
a
d
a
s
a
ggr
e
s
s
i
ve
o
r
n
e
ut
r
a
l
.
T
h
e
l
e
xi
c
o
n
-
b
a
s
e
d
a
ppr
o
a
c
h
h
a
s
t
h
e
a
dv
a
n
t
a
ge
to
n
ot
r
e
qui
r
e
pr
i
o
r
t
r
a
i
nin
g
(
o
n
a
da
t
a
s
e
t)
to
m
i
ne
t
h
e
da
t
a
b
ut
t
h
e
f
i
na
l
r
e
s
u
l
t
c
a
n
d
i
f
f
e
r
a
c
c
o
r
d
i
n
g
t
o
t
h
e
c
o
n
t
e
x
t
i
n
w
hi
c
h
t
h
e
l
e
xi
c
o
n
s
we
r
e
c
r
e
a
t
e
d.
S
i
r
i
a
r
a
y
a
e
t
al.
[
10]
a
pp
l
i
e
d
t
h
e
m
a
c
hi
ne
l
e
a
r
ni
ng
(
M
L
)
a
ppr
o
a
c
h
to
c
r
e
a
t
e
a
c
r
i
m
e
-
s
o
l
vi
ng
too
l
t
h
a
t
g
i
v
e
s
c
o
n
t
e
x
t
f
o
r
c
r
i
mi
na
l
e
v
e
n
t
s
.
T
h
e
a
ut
h
o
r
s
ut
i
li
z
e
d
a
t
a
f
r
o
m
T
w
i
t
t
e
r
by
pr
o
vi
d
i
ng
c
o
n
t
e
x
t
ua
l
i
n
f
o
r
m
a
t
i
o
n
a
b
o
ut
c
r
i
m
e
i
nc
i
de
n
t
s
o
c
c
ur
r
i
n
g
i
n
a
s
p
e
c
i
f
i
c
a
r
e
a
(
S
a
n
F
r
a
n
c
i
s
c
o
)
us
i
ng
m
a
c
hi
ne
l
e
a
r
ni
ng
c
l
a
s
s
if
i
c
a
t
i
o
n
m
o
de
l
s
(
l
o
g
i
s
t
i
c
r
e
gr
e
s
s
i
o
n
a
n
d
s
uppo
r
t
v
e
c
to
r
m
a
c
hi
ne
)
.
T
h
e
m
a
c
hi
ne
l
e
a
r
ni
ng
a
ppr
o
a
c
h
c
a
n
a
da
pt
a
n
d
c
r
e
a
t
e
tr
a
i
n
e
d
m
o
de
l
s
f
o
r
s
pe
c
i
f
i
c
pur
po
s
e
s
a
n
d
c
o
n
t
e
x
t
s
,
b
ut
i
t
h
a
s
l
o
w
a
pp
l
i
c
a
bil
i
t
y
o
n
n
e
w
da
t
a
b
e
c
a
us
e
i
t
n
e
c
e
s
s
i
t
a
t
e
s
t
h
e
a
v
a
il
a
bil
i
t
y
o
f
l
a
b
e
l
e
d
da
t
a
.
P
e
r
e
i
r
a
-
K
o
ha
t
s
u
e
t
al.
[
11
]
a
ppl
i
e
d
a
de
e
p
l
e
a
r
ni
ng
a
ppr
o
a
c
h
to
de
v
e
l
o
p
a
T
wi
t
t
e
r
-
b
a
s
e
d
i
n
t
e
l
li
ge
nc
e
s
y
s
t
e
m
f
o
r
d
e
t
e
c
t
i
n
g
a
n
d
a
n
a
ly
z
i
ng
h
a
t
e
s
pe
e
c
h
c
a
ll
e
d
H
a
t
e
r
Ne
t
.
T
h
e
a
ut
h
o
r
s
u
t
i
li
z
e
d
a
m
u
l
t
il
a
y
e
r
pe
r
c
e
pt
r
o
n
n
e
ur
a
l
n
e
t
wor
k
t
h
a
t
a
c
c
e
p
t
s
a
s
i
n
put
t
h
e
t
we
e
t
'
s
w
o
r
d
,
e
m
o
j
i
,
a
n
d
e
x
pr
e
s
s
i
o
n
e
m
be
dd
i
n
g
s
to
ke
n
s
e
nha
n
c
e
d
by
t
h
e
t
f
-
i
d
f
a
n
d
o
u
t
pu
t
t
h
e
a
r
e
a
un
de
r
t
h
e
c
ur
v
e
(
A
U
C
)
.
B
E
R
T
i
s
o
n
e
o
f
t
h
e
b
e
s
t
DL
a
l
go
r
i
t
hm
s
i
n
S
A
,
a
s
s
h
o
wn
i
n
[
12]
.
Ya
da
v
e
t
al.
[
13]
us
e
d
t
h
e
B
E
R
T
a
l
go
r
i
t
hm
t
o
i
de
n
t
i
f
y
c
y
be
r
b
u
l
lyi
ng
o
n
s
o
c
i
a
l
m
e
d
ia
p
l
a
t
f
o
r
m
s
,
ut
i
li
z
i
ng
t
h
e
B
E
R
T
m
o
de
l
a
s
a
c
l
a
s
s
i
f
ier
wi
t
h
a
s
i
n
g
l
e
li
ne
a
r
n
e
ur
a
l
n
e
t
wor
k
l
a
y
e
r
,
t
r
a
i
n
e
d
a
n
d
e
v
a
l
ua
t
e
d
o
n
t
w
o
s
o
c
i
a
l
m
e
d
i
a
da
t
a
s
e
t
s
,
o
n
e
s
m
a
ll
a
n
d
o
n
e
f
a
i
r
ly
bi
g.
W
e
i
r
e
t
al.
[
14]
pr
o
p
o
s
e
d
a
s
y
s
t
e
m
t
h
a
t
c
o
m
bi
ne
s
b
o
t
h
m
a
c
hi
ne
l
e
a
r
ni
n
g
a
n
d
l
e
xi
c
o
n
-
ba
s
e
d
a
ppr
o
a
c
h
e
s
t
o
de
t
e
c
t
t
e
r
r
or
i
s
t
we
b
pa
ge
s
,
ga
uge
t
he
s
t
r
e
n
gt
h
o
f
t
h
e
i
r
c
o
n
t
e
n
t
,
a
n
d
c
a
t
e
g
o
r
i
z
e
da
t
a
c
o
l
lec
t
e
d
o
n
t
e
r
r
o
r
i
s
m
a
n
d
e
x
t
r
e
m
i
s
m
n
e
t
wo
r
ks
.
T
h
e
a
ut
h
o
r
s
d
e
v
e
l
o
pe
d
a
n
i
n
-
de
pt
h
f
r
e
qu
e
n
c
y
s
t
ud
y
o
f
t
h
e
s
y
n
t
a
x
u
s
i
ng
t
h
e
P
o
s
i
t
t
e
x
t
ua
l
a
n
a
ly
s
i
s
t
oo
l
k
i
t
,
whi
c
h
i
nc
l
ude
d
m
u
l
t
i
-
wo
r
d
uni
t
s
a
n
d
t
h
e
i
r
r
e
l
a
t
e
d
pa
r
t
s
o
f
s
pe
e
c
h
.
Af
t
e
r
t
h
a
t,
ut
i
l
i
z
i
ng
k
n
o
w
l
e
dge
e
x
t
r
a
c
t
i
o
n
t
e
c
hni
que
s
,
t
h
e
f
i
nd
i
ngs
a
r
e
us
e
d
i
n
a
k
n
o
w
l
e
dg
e
e
x
t
r
a
c
t
i
o
n
pr
o
c
e
s
s
(
de
c
i
s
i
o
n
t
r
e
e
,
r
a
n
do
m
f
o
r
e
s
t
).
T
h
e
a
b
o
v
e
-
m
e
n
t
i
o
n
e
d
r
e
l
a
t
e
d
wo
r
ks
c
a
n
g
i
ve
a
de
c
e
n
t
s
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
pr
e
d
i
c
t
i
o
n
r
a
t
e
i
n
s
e
c
ur
i
t
y
i
n
t
e
ll
i
ge
n
c
e
c
o
n
t
e
x
t
.
Ho
we
v
e
r
,
wor
ks
t
h
a
t
us
e
a
hy
br
i
d
a
ppr
o
a
c
h
o
f
b
o
t
h
l
e
xi
c
o
n
-
b
a
s
e
d
a
n
d
de
e
p
l
e
a
r
ni
ng
a
ppr
o
a
c
h
e
s
a
r
e
r
a
r
e
,
e
s
pe
c
i
a
ll
y
i
n
t
hi
s
c
o
n
t
e
x
t
.
A
s
a
r
e
s
u
l
t
,
o
ur
r
e
s
e
a
r
c
h
c
o
n
t
r
i
b
ut
e
s
t
o
t
h
e
a
b
o
v
e
-
m
e
n
t
i
o
n
e
d
wo
r
ks
a
s
a
n
e
s
s
e
n
t
i
a
l
e
x
pe
r
i
m
e
n
t
a
l
e
x
pa
ns
i
o
n
us
i
n
g
a
hy
b
r
i
d
a
ppr
o
a
c
h
t
h
a
t
c
o
m
bi
ne
s
b
o
t
h
l
e
xi
c
o
n
-
ba
s
e
d
a
nd
de
e
p
l
e
a
r
ni
ng
a
ppr
o
a
c
h
e
s
.
3.
M
E
T
HO
D
S
e
n
t
i
m
e
n
t
a
na
ly
s
i
s
h
a
s
b
e
e
n
pr
a
c
t
i
c
e
d
o
n
a
v
a
r
i
e
t
y
o
f
t
o
pi
c
s
l
i
k
e
m
o
vi
e
r
e
vi
e
ws
,
s
e
r
vi
c
e
a
n
d
pr
o
duc
t
r
e
vi
e
ws
,
n
e
w
s
,
bl
o
gs
,
c
h
a
t
b
o
t,
a
n
d
s
e
c
ur
i
t
y
i
n
t
e
l
li
ge
n
c
e
[
7]
.
Dur
i
n
g
a
n
e
x
t
r
e
m
e
i
nc
i
de
n
t
,
m
o
s
t
o
f
t
h
e
m
e
s
s
a
ge
s
s
h
a
r
e
d
o
n
T
w
i
t
t
e
r
a
r
e
c
o
m
m
e
n
t
a
r
y
-
r
e
l
a
t
e
d,
s
uc
h
a
s
e
x
pr
e
s
s
i
ng
t
h
o
ugh
t
s
,
r
e
po
r
t
i
n
g
t
he
m
,
a
n
d
a
tt
e
m
pt
i
n
g
t
o
un
de
r
s
t
a
n
d
wh
a
t
ha
ppe
n
e
d
a
n
d
w
hy
.
S
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
r
e
s
e
a
r
c
h
ha
s
m
a
i
n
ly
f
o
c
us
e
d
e
i
t
h
e
r
o
n
i
de
n
t
i
f
y
i
ng
w
h
e
t
h
e
r
a
g
i
v
e
n
t
e
x
t
ua
l
e
n
t
i
t
y
i
s
o
bj
e
c
t
i
v
e
o
r
s
ubj
e
c
t
i
ve
o
r
o
n
i
de
n
t
i
f
yi
ng
t
h
e
po
l
a
r
i
t
y
o
f
s
u
bj
e
c
t
i
v
e
t
e
x
t
s
[
15
]
.
Our
c
a
s
e
s
t
udi
e
s
a
r
e
e
x
a
m
p
l
e
s
o
f
s
e
ve
r
e
o
c
c
ur
r
e
n
c
e
s
b
r
o
ugh
t
o
n
by
i
nd
i
vi
du
a
l
c
h
o
i
c
e
s
t
h
a
t
h
a
d
a
phy
s
i
c
a
l
im
pa
c
t
o
n
i
n
d
i
v
i
dua
l
s
,
a
n
d
t
h
e
r
e
f
o
r
e
t
h
e
s
e
e
ve
n
t
s
wo
ul
d
h
a
v
e
m
u
l
t
i
p
l
e
s
o
c
i
a
l
m
e
d
i
a
c
o
m
m
u
ni
c
a
t
i
o
n
s
.
S
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
(
o
r
o
pi
ni
o
n
mi
n
i
ng)
i
s
e
m
p
l
o
y
e
d
o
n
T
w
i
t
t
e
r
by
t
h
e
m
e
a
ns
o
f
o
n
e
o
f
t
h
e
f
o
l
l
o
w
i
n
g
t
e
c
hni
que
s
:
−
L
e
xi
c
o
n
-
ba
s
e
d
(
r
u
l
e
-
b
a
s
e
d)
a
ppr
o
a
c
h
:
i
t
i
s
us
e
d
to
c
h
a
r
a
c
t
e
r
i
z
e
t
h
e
po
l
a
r
i
t
y
(
n
e
ga
t
i
v
e
,
po
s
i
t
i
ve
,
a
nd
n
e
ut
r
a
l
)
o
f
t
e
x
t
ua
l
m
a
t
e
r
i
a
l
(
wo
r
ds
o
r
ph
r
a
s
e
s
t
h
a
t
c
o
n
v
e
y
t
h
e
s
e
n
t
i
m
e
n
t
o
f
t
h
e
e
n
t
i
r
e
t
e
x
t
)
.
T
hi
s
m
e
t
h
o
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
I
S
S
N:
2502
-
4752
C
r
ime
pr
e
diction
us
ing
a
hy
br
id
s
e
nti
me
nt
analys
is
appr
oac
h
bas
e
d
on
the
…
(
M
ohamm
e
d
B
ouk
abo
us
)
1133
c
a
n
b
e
d
i
v
i
de
d
i
n
t
o
i
)
c
o
r
pus
-
b
a
s
e
d
(
us
i
ng
c
o
r
pus
da
t
a
i
n
e
i
t
h
e
r
a
s
t
a
t
i
s
t
i
c
a
l
o
r
s
e
m
a
n
t
i
c
m
a
nn
e
r
)
a
n
d
ii
)
d
i
c
t
i
o
n
a
r
y
-
ba
s
e
d
(
us
i
ng
d
i
c
t
i
o
n
a
r
y
d
a
t
a
)
.
−
M
a
c
hi
ne
l
e
a
r
ni
ng
a
ppr
o
a
c
h
:
t
hi
s
t
e
c
hni
que
a
ll
o
w
s
m
a
c
hi
ne
s
t
o
l
e
a
r
n
t
a
s
ks
w
i
t
h
o
ut
b
e
i
n
g
p
a
r
t
i
c
u
l
a
r
ly
pr
o
g
r
a
m
m
e
d
t
o
pe
r
f
o
r
m
t
h
e
m
.
B
y
t
r
a
i
ni
ng
M
L
a
lgo
r
i
t
hm
s
w
i
t
h
e
x
a
m
p
l
e
s
o
f
l
a
b
e
l
e
d
e
m
o
t
i
o
ns
i
n
t
e
x
t
s
(
da
t
a
s
e
t
)
,
m
a
c
hi
ne
s
a
ut
o
m
a
t
i
c
a
ll
y
l
e
a
r
n
h
o
w
to
pr
e
d
i
c
t
s
e
n
t
i
m
e
n
t
w
i
t
h
o
ut
h
u
m
a
n
i
n
t
e
r
v
e
n
t
i
o
n
.
M
L
m
o
de
l
s
c
a
n
be
t
r
a
i
n
e
d
i
n
S
A
t
a
s
ks
t
o
r
e
a
d
b
e
y
o
n
d
s
im
p
l
e
de
f
i
ni
t
i
o
ns
,
a
n
d
u
n
de
r
s
t
a
n
d
m
o
r
e
c
o
m
p
l
e
x
t
hi
n
gs
li
ke
c
o
n
t
e
x
t
,
m
e
t
a
ph
o
r
,
s
a
r
c
a
s
m
,
a
n
d
m
i
s
a
pp
l
i
e
d
wo
r
ds
.
T
h
e
m
o
s
t
pr
o
m
i
s
i
ng
a
l
go
r
i
t
hm
s
a
r
e
DL
o
n
e
s
.
−
H
y
b
r
i
d
a
ppr
o
a
c
h
:
i
t
i
s
t
h
e
c
o
m
bi
na
t
i
o
n
o
f
M
L
m
e
t
h
o
ds
a
n
d
l
e
xi
c
o
n
-
b
a
s
e
d
a
ppr
o
a
c
h
e
s
i
n
t
o
o
n
e
s
y
s
t
e
m
t
o
e
nh
a
nc
e
t
h
e
m
o
de
l
's
pe
r
f
o
r
m
a
n
c
e
a
s
we
ll
a
s
s
e
n
t
i
m
e
n
t
s
c
o
r
i
n
g.
T
hi
s
g
i
ve
s
us
a
m
a
c
hi
ne
l
e
a
r
ni
n
g
mo
de
l
t
h
a
t
h
a
s
b
e
e
n
t
r
a
i
n
e
d
o
n
a
l
a
b
e
l
e
d
c
o
r
pus
.
Our
pr
o
p
o
s
e
d
m
e
t
h
o
d
i
s
t
o
us
e
t
h
e
hy
b
r
i
d
a
ppr
o
a
c
h
t
o
b
u
i
l
d
a
B
E
R
T
m
o
de
l
o
n
a
t
we
e
t
’
s
da
t
a
s
e
t
,
l
a
b
e
l
e
d
us
i
ng
t
h
e
l
e
xi
c
o
n
-
b
a
s
e
d
a
ppr
o
a
c
h
by
f
o
l
l
o
w
i
n
g
t
h
e
n
e
x
t
s
t
e
ps
a
s
s
h
o
wn
i
n
F
i
gur
e
1.
i)
F
i
r
s
t
S
t
e
p:
Da
t
a
c
o
l
l
e
c
t
i
o
n
.
I
n
t
hi
s
s
t
a
ge
,
da
t
a
f
o
r
o
ur
a
n
a
ly
s
i
s
i
s
c
r
a
w
l
e
d
f
r
o
m
T
w
i
t
t
e
r
us
i
n
g
T
w
i
t
t
e
r
a
pp
l
i
c
a
t
i
o
n
pr
o
gr
a
m
mi
ng
i
n
t
e
r
f
a
c
e
(
A
P
I
)
.
W
e
r
e
mi
nd
t
h
a
t
t
h
e
da
t
a
f
r
o
m
t
hi
s
s
o
c
i
a
l
m
e
d
i
a
p
l
a
t
f
o
r
m
is
a
v
a
l
ua
bl
e
r
e
c
o
r
d
o
f
t
h
e
i
n
t
e
r
s
u
bj
e
c
t
i
v
e
u
n
d
e
r
s
t
a
n
d
in
g
o
f
a
n
e
x
t
r
e
m
e
e
v
e
n
t
by
i
nd
i
vi
du
a
l
s
a
n
d
gr
o
ups
due
to
t
h
e
n
a
vi
ga
t
i
o
n
a
l
e
a
s
e
,
a
n
o
nymi
t
y
(
t
h
e
a
bil
i
t
y
t
o
po
s
t
a
ny
m
a
t
e
r
i
a
l
w
i
t
h
o
u
t
di
s
c
l
o
s
i
ng
n
a
m
e
s
)
,
a
n
d
th
e
publi
c
a
t
i
o
n
s
y
s
t
e
m
's
f
l
a
w
s
(
us
e
r
s
o
n
l
y
ne
e
d
to
h
a
ve
e
a
c
h
a
va
l
i
d
a
c
c
o
un
t
)
.
ii
)
S
e
c
o
n
d
S
t
e
p:
E
n
r
i
c
hm
e
n
t
.
W
e
u
s
e
d
t
h
e
S
pa
r
k
na
t
ur
a
l
l
a
n
gua
ge
t
oo
l
k
i
t
(
NL
T
K
)
li
b
r
a
r
y
t
o
b
r
e
a
k
t
h
e
s
e
n
t
e
n
c
e
s
i
n
t
o
wo
r
ds
a
n
d
t
a
gg
i
n
g
t
h
e
m
.
Af
t
e
r
do
i
ng
t
h
e
pa
r
t
s
o
f
s
p
e
e
c
h
(
P
OS
)
t
a
gg
i
n
g
w
hi
c
h
s
t
a
n
ds
f
o
r
P
a
r
t
o
f
S
pe
e
c
h
t
a
gg
i
n
g
o
n
t
h
e
wo
r
ds
o
f
t
we
e
t
s
,
we
a
l
s
o
t
a
gge
d
t
h
e
d
i
f
f
e
r
e
n
t
wo
r
ds
b
a
s
e
d
o
n
f
i
ve
d
i
c
t
i
o
n
a
r
i
e
s
t
h
a
t
we
f
e
d
i
n
t
o
o
ur
s
y
s
t
e
m
(
po
s
i
t
i
v
e
,
ne
ga
t
i
v
e
,
i
n
c
r
e
m
e
n
t
a
l
,
de
c
r
e
m
e
n
t
a
l
,
a
n
d
i
nv
e
r
s
e
wo
r
ds
)
e
a
c
h
o
n
e
w
i
t
h
a
d
i
f
f
e
r
e
n
t
w
e
i
g
h
t
.
ii
i
)
T
hi
r
d
S
t
e
p:
Da
t
a
c
l
a
s
s
if
i
c
a
t
i
o
n
.
Us
i
ng
t
h
e
a
b
o
v
e
d
i
c
t
i
o
n
a
r
y
(
t
a
gs
)
,
t
h
e
m
o
de
l
c
a
l
c
u
l
a
t
e
s
t
h
e
s
e
n
t
i
men
t
s
c
o
r
e
o
f
e
a
c
h
s
e
n
t
e
n
c
e
a
n
d
g
i
ve
s
t
h
e
o
v
e
r
a
ll
s
e
n
t
i
m
e
n
t
s
c
o
r
e
.
A
po
s
i
t
i
v
e
s
c
o
r
e
i
s
t
a
ke
n
a
s
a
n
o
r
m
a
l
t
we
e
t
a
n
d
a
ne
ga
t
i
v
e
o
n
e
i
s
t
a
ke
n
a
s
a
c
r
i
m
e
-
r
e
l
a
t
e
d
t
we
e
t.
i
v)
F
o
ur
t
h
S
t
e
p:
P
r
e
-
pr
o
c
e
s
s
i
n
g
.
B
e
f
o
r
e
we
s
t
a
r
t
b
u
i
l
d
i
ng
o
ur
B
E
R
T
m
o
de
l
,
t
h
e
a
c
qu
i
r
e
d
da
t
a
i
s
c
l
e
a
n
e
d
to
m
a
ke
i
t
r
e
a
d
y
f
o
r
f
e
e
d
i
ng
i
t
i
n
t
o
t
h
e
c
l
a
s
s
if
i
e
r
.
v)
F
i
f
t
h
S
t
e
p:
B
u
i
l
d
i
ng
t
h
e
m
o
de
l
.
W
e
l
o
a
d
t
h
e
pr
e
-
tr
a
i
ne
d
B
E
R
T
S
e
que
n
c
e
C
l
a
s
s
i
f
i
e
r
a
n
d
T
o
ke
n
i
z
e
r
.
T
h
e
n
,
we
us
e
d
t
h
e
S
e
que
n
c
e
C
l
a
s
s
if
i
e
r
a
n
d
B
E
R
T
'
s
T
o
ke
ni
z
e
r
,
to
b
u
i
l
d
o
ur
m
o
de
l
a
n
d
t
o
ke
ni
z
e
r
.
F
i
na
ll
y
,
we
us
e
d
t
hi
s
m
o
de
l
o
n
t
h
e
T
wi
t
t
e
r
da
t
a
s
e
t
t
h
a
t
we
a
l
r
e
a
d
y
l
a
b
e
l
e
d
t
o
f
i
ne
-
t
un
e
t
h
e
c
l
a
s
s
i
f
i
e
r
.
vi
)
S
i
x
t
h
S
t
e
p:
E
v
a
l
ua
t
i
n
g
t
h
e
m
o
de
l
.
W
e
a
s
s
e
s
s
o
ur
m
o
de
l
us
i
ng
t
h
e
t
e
s
t
da
t
a
s
e
t
by
pr
e
d
i
c
t
i
n
g
s
e
n
t
i
m
e
n
t
s
a
f
t
e
r
i
t
h
a
s
b
e
e
n
b
u
il
t
a
n
d
ge
ne
r
a
t
e
d.
F
i
gur
e
1.
P
r
o
p
o
s
e
d
m
e
t
h
o
d
3.
1.
L
ab
e
l
in
g
t
h
e
d
at
a
T
h
e
T
w
i
t
t
e
r
da
t
a
c
o
l
l
e
c
t
e
d
f
o
r
t
e
x
t
a
n
a
ly
s
i
s
c
o
n
t
a
i
n
70
,
000
t
we
e
t
s
(
27
,
000
c
r
i
m
e
t
we
e
t
s
,
43
,
000
n
o
r
m
a
l
t
e
x
t
s
)
.
T
h
e
c
o
r
pus
f
o
r
C
h
r
i
s
t
c
h
ur
c
h
M
o
s
qu
e
a
tt
a
c
ks
c
o
n
t
a
i
ns
37
,
000
t
we
e
t
s
,
E
l
P
a
s
o
s
h
o
ot
i
n
g
c
o
n
t
a
i
ns
15
,
000
a
n
d
Ha
n
a
u
a
t
t
a
c
ks
c
o
n
t
a
i
n
18
,
000
.
T
h
e
s
t
r
e
a
m
i
ng
T
w
i
t
t
e
r
A
P
I
wa
s
us
e
d
to
ga
t
h
e
r
t
h
e
s
e
t
we
e
t
s
.
W
e
we
r
e
un
a
bl
e
t
o
a
c
qui
r
e
hi
s
t
o
r
i
c
a
l
da
t
a
us
i
n
g
T
w
i
t
t
e
r
s
i
n
c
e
i
t
o
nly
pe
r
m
i
t
s
us
t
o
c
o
l
l
e
c
t
da
t
a
f
r
o
m
t
h
e
l
a
s
t
s
e
v
e
n
da
y
s
,
a
n
d
s
o
we
b
e
g
i
n
t
o
c
o
l
l
e
c
t
t
we
e
t
s
da
t
a
t
h
e
s
e
l
a
s
t
t
h
r
e
e
y
e
a
r
s
.
T
h
e
da
t
a
we
us
e
d
wa
s
t
a
ke
n
o
v
e
r
t
h
r
e
e
D
a
t
a
L
o
a
d
in
g
T
r
a
n
s
fo
r
m
t
o
D
a
t
a
F
r
a
m
e
Po
s
i
t
i
v
e
w
o
r
d
s
N
e
g
a
t
i
v
e
w
o
r
d
s
I
n
c
r
e
m
e
n
t
a
l
w
o
r
d
s
D
e
c
r
e
m
e
n
t
a
l
w
o
r
d
s
I
n
v
e
r
s
e
w
o
r
d
s
A
s
s
i
g
n
p
o
s
i
t
i
v
e
t
a
g
s
A
s
s
i
g
n
n
e
g
a
t
i
v
e
t
a
g
s
A
s
s
i
g
n
i
n
c
r
e
m
e
n
t
a
l
t
a
g
s
A
s
s
i
g
n
de
c
r
e
m
e
n
t
a
l
t
a
g
s
A
s
s
i
g
n
i
n
v
e
r
s
e
t
a
g
s
P
O
S
T
a
g
g
e
r
T
a
g
f
i
lt
e
r
T
a
g
s
e
n
t
e
n
c
e
C
a
l
c
ul
a
t
e
s
e
n
t
i
m
e
n
t
s
c
o
r
e
P
r
e
d
ic
t
s
e
n
t
ime
n
t
L
a
b
e
le
d
D
a
t
a
B
E
R
T
R
e
m
o
v
e
t
w
e
e
t
s
f
e
a
t
u
r
e
s
R
e
m
o
v
e
r
e
p
e
a
t
e
d
t
w
e
e
t
s
R
e
m
o
v
e
r
e
p
e
a
t
e
d
l
e
t
t
e
r
s
R
e
m
o
v
e
s
t
o
p
w
o
r
ds
Lo
w
e
r
c
a
s
i
n
g
Ta
g
f
i
l
t
e
r
T
o
k
e
n
iz
e
P
a
r
t
it
io
n
in
g
T
e
s
t
in
g
s
e
t
T
r
a
in
in
g
S
e
t
P
a
r
a
m
e
t
e
r
s
a
d
j
us
t
e
m
e
n
t
T
r
a
i
n
i
n
g
t
h
e
M
o
d
e
l
P
r
e
d
i
c
t
i
n
g
t
e
s
t
d
a
t
a
S
a
v
in
g
t
h
e
mo
d
e
l
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2502
-
4752
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
,
Vo
l
.
25
,
N
o
.
2
,
F
e
b
r
ua
r
y
20
22
:
1131
-
1139
1134
da
y
s
f
o
r
e
a
c
h
o
f
t
h
e
i
n
c
i
d
e
n
t
s
,
o
n
e
da
y
b
e
f
o
r
e
,
o
n
e
da
y
a
f
t
e
r
,
a
n
d
t
h
e
da
y
i
n
q
ue
s
t
i
o
n
,
a
l
l
u
s
i
n
g
t
h
e
c
i
t
y
's
n
a
me
a
s
a
ke
y
wo
r
d.
−
C
h
r
i
s
t
c
h
ur
c
h
M
o
s
que
a
tt
a
c
ks
a
r
e
a
s
e
r
i
e
s
o
f
f
a
r
-
r
i
gh
t
t
e
r
r
or
i
s
t
a
tt
a
c
ks
c
o
m
m
i
t
t
e
d
o
n
M
a
r
c
h
15,
2019
,
b
y
B
r
e
n
t
o
n
T
a
r
r
a
n
t
a
ga
i
ns
t
t
w
o
m
o
s
que
s
i
n
t
h
e
c
i
t
y
o
f
C
h
r
i
s
t
c
h
ur
c
h
,
Ne
w
Z
e
a
l
a
n
d,
w
hi
c
h
l
e
f
t
51
de
a
d
a
nd
49
i
nj
ur
e
d
[
16]
.
I
t
i
s
t
h
e
d
e
a
d
li
e
s
t
m
a
s
s
a
c
r
e
t
o
h
a
ve
o
c
c
ur
r
e
d
i
n
p
e
a
c
e
t
i
m
e
i
n
Ne
w
Z
e
a
l
a
n
d
s
i
nc
e
t
h
a
t
o
f
B
o
y
d
i
n
1809
(
66
to
70
de
a
d)
.
−
E
l
P
a
s
o
s
h
o
ot
i
n
g
t
oo
k
pl
a
c
e
o
n
A
ugus
t
3,
2019,
i
n
E
l
P
a
s
o
,
T
e
x
a
s
[
17]
.
T
we
n
t
y
-
t
h
r
e
e
pe
o
p
l
e
a
r
e
k
il
led,
a
t
l
e
a
s
t
t
we
n
t
y
-
t
h
r
e
e
ot
h
e
r
s
a
r
e
i
nj
ur
e
d.
T
h
e
k
i
ll
e
r
wa
s
a
r
r
e
s
t
e
d
a
f
t
e
r
s
ur
r
e
n
de
r
i
n
g
t
o
t
h
e
po
l
i
c
e
,
s
a
ys
h
e
wa
n
t
e
d
to
ki
ll
a
s
m
a
ny
M
e
xi
c
a
ns
a
s
po
s
s
i
bl
e
.
R
a
c
i
s
t
h
a
t
e
c
r
i
m
e
i
s
t
h
e
hy
po
t
h
e
s
i
s
f
a
v
o
r
e
d
by
a
s
ur
v
e
y
.
T
h
e
M
e
xi
c
a
n
go
v
e
r
nm
e
n
t
c
o
n
s
i
de
r
s
i
t
a
n
a
n
t
i
-
M
e
xi
c
a
n
a
t
t
a
c
k,
a
n
d
t
h
e
UN
s
pe
a
ks
o
f
a
t
e
r
r
o
r
i
s
t
a
c
t
a
ga
i
ns
t
t
h
e
L
a
t
i
n
Am
e
r
i
c
a
n
c
o
m
m
u
ni
t
y
i
n
t
h
e
U
ni
t
e
d
S
t
a
t
e
s
.
−
Ha
n
a
u
a
tt
a
c
ks
too
k
pl
a
c
e
o
n
F
e
b
r
ua
r
y
19,
2020,
43
-
y
e
a
r
-
o
l
d
T
o
bi
a
s
R
a
t
hj
e
n
f
r
o
m
Ha
n
a
u,
h
e
r
e
i
na
f
t
e
r
r
e
f
e
r
r
e
d
to
a
s
"
R
.
,
”
s
h
o
t
a
n
d
k
i
ll
e
d
ni
ne
Ha
n
a
u
c
i
t
ize
ns
w
i
t
h
a
mi
gr
a
t
i
o
n
ba
c
kgr
o
un
d
i
n
f
r
o
n
t
o
f
a
s
his
ha
b
a
r
[
18]
.
He
t
h
e
n
s
h
o
t
hi
m
s
e
lf
a
n
d
hi
s
m
o
t
h
e
r
i
n
his
pa
r
e
n
t
s
'
a
pa
r
t
m
e
n
t
.
Af
t
e
r
c
o
l
l
e
c
t
i
n
g
t
he
da
t
a
s
e
t
,
we
to
ke
ni
z
e
i
t
a
n
d
pe
r
f
o
r
m
pa
r
t
o
f
s
pe
e
c
h
t
a
gg
i
n
g
(
P
OS)
f
o
r
e
a
c
h
wo
r
d
us
i
n
g
t
h
e
S
pa
r
k
N
L
T
K
li
br
a
r
y
t
o
b
r
e
a
k
t
h
e
s
e
n
t
e
nc
e
s
i
n
t
o
wor
ds
a
n
d
t
a
g
t
h
e
m
.
T
h
e
n
,
we
e
n
r
i
c
h
i
t
u
s
i
n
g
f
i
ve
d
i
c
t
i
o
n
a
r
i
e
s
t
h
a
t
we
f
e
d
i
n
t
o
o
ur
s
y
s
t
e
m
(
po
s
i
t
i
ve
,
n
e
g
a
t
i
ve
,
i
nc
r
e
m
e
n
t
a
l
,
de
c
r
e
m
e
n
t
a
l
,
a
n
d
i
nve
r
s
e
wo
r
ds
)
e
a
c
h
o
n
e
w
i
t
h
a
d
i
f
f
e
r
e
n
t
we
i
g
h
t
(
+
1,
-
1,
*2,
/2
,
*
-
1,
r
e
s
pe
c
t
i
v
e
ly
)
.
T
h
e
po
s
i
t
i
v
e
a
n
d
n
e
ga
t
i
v
e
d
i
c
t
i
o
na
r
i
e
s
a
r
e
b
a
s
e
d
o
n
L
i
u
a
n
d
Hu
o
pi
ni
o
n
l
e
xi
c
o
n
[
19]
t
h
a
t
w
e
h
a
v
e
c
o
m
p
l
e
t
e
d
w
i
t
h
ot
h
e
r
s
e
c
ur
i
t
y
i
n
t
e
l
li
ge
n
c
e
l
e
xi
c
o
n
f
r
o
m
d
i
f
f
e
r
e
n
t
publi
c
s
e
c
ur
i
t
y
s
o
ur
c
e
s
li
ke
t
h
e
F
B
I
NI
B
R
S
da
t
a
[
20]
.
W
e
h
a
v
e
a
l
s
o
i
nc
l
ude
d
i
n
o
ur
n
e
ga
t
i
v
e
l
e
xi
c
o
n
t
h
e
I
s
l
a
m
o
p
h
o
bi
a
(
t
e
r
r
or
i
s
m
a
n
d
e
x
t
r
e
m
i
s
m
a
ga
i
ns
t
M
us
li
m
s
)
m
o
s
t
us
e
d
wo
r
ds
,
r
e
s
u
l
t
i
n
g
f
r
o
m
i
nc
i
de
n
t
s
a
n
d
e
v
e
n
t
s
t
h
a
t
a
f
f
e
c
t
M
us
l
im
c
o
m
m
u
ni
t
i
e
s
i
n
ge
n
e
r
a
l
[
21]
a
s
s
h
o
wn
i
n
F
i
gur
e
2.
S
ubs
e
que
n
t
l
y
,
w
e
ge
t
l
a
b
e
l
e
d
wo
r
ds
whi
c
h
we
us
e
d
t
o
c
a
l
c
u
l
a
t
e
s
e
nt
e
n
c
e
s
s
e
n
t
i
m
e
n
t
s
i
n
g
e
n
e
r
a
l
,
w
h
e
t
h
e
r
n
o
r
m
a
l
o
r
c
r
i
m
i
na
l
.
An
d
s
o
o
n
,
we
we
r
e
a
bl
e
t
o
l
a
b
e
l
o
ur
da
t
a
s
e
t.
F
i
gur
e
2
.
W
o
r
d
c
l
o
ud
o
f
I
s
l
a
m
o
p
h
o
bi
a
3.
2.
P
r
e
p
r
oc
e
s
s
in
g
Af
t
e
r
t
h
e
T
w
i
t
t
e
r
da
t
a
h
a
s
b
e
e
n
ga
t
h
e
r
e
d
a
n
d
c
o
nv
e
r
t
e
d
to
t
e
x
t
f
o
r
m
a
t
,
i
t
m
us
t
b
e
c
l
e
a
n
e
d
a
n
d
pr
e
-
pr
o
c
e
s
s
e
d
b
e
f
o
r
e
b
e
i
ng
ut
i
li
z
e
d
i
n
s
e
n
t
i
m
e
n
t
a
n
a
lys
i
s
.
T
w
i
t
t
e
r
da
t
a
i
s
n
oto
r
i
o
us
f
o
r
b
e
i
ng
hi
g
hly
l
o
u
d,
w
i
t
h
a
l
o
t
o
f
b
a
na
l
c
h
a
t
t
e
r
a
n
d
li
ngu
i
s
t
i
c
i
nc
o
ns
i
s
t
e
n
c
i
e
s
[
22]
.
T
o
m
a
ke
t
h
e
T
w
i
t
t
e
r
da
t
a
a
s
c
l
e
a
n
a
n
d
e
a
s
y
a
s
f
e
a
s
i
b
l
e
f
o
r
o
u
r
s
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
a
l
go
r
i
t
hm
,
t
h
e
f
o
l
l
o
w
i
ng
pr
e
pa
r
a
t
i
o
n
pr
o
c
e
dur
e
s
we
r
e
d
o
n
e
.
−
R
e
m
o
va
l
o
f
t
we
e
t
s
f
e
a
t
ur
e
s
:
T
we
e
t
s
a
r
e
f
r
e
que
n
t
l
y
c
l
o
gge
d
w
i
t
h
d
a
t
a
t
h
a
t
i
s
n'
t
i
mm
e
d
i
a
t
e
l
y
r
e
l
e
v
a
nt
o
r
us
e
f
u
l
f
o
r
s
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
.
T
h
e
h
a
s
h
t
a
g
s
i
g
n
(
#)
,
UR
L
s
/
hy
pe
r
l
i
nk
s
,
n
u
m
be
r
s
,
r
e
f
e
r
e
n
c
e
s
t
o
o
th
e
r
us
e
r
s
,
r
e
t
we
e
t
s
y
m
b
o
l
(
R
T
)
,
a
n
d
e
m
o
t
i
c
o
ns
a
r
e
a
l
l
i
nc
l
ude
d
i
n
t
hi
s
da
t
a
.
W
hil
e
n
o
t
n
e
c
e
s
s
a
r
i
ly
l
a
c
k
i
n
g
in
i
n
t
r
i
ns
i
c
wo
r
t
h
,
s
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
a
l
go
r
i
t
hm
s
d
o
n
ot
i
de
n
t
i
f
y
t
hi
s
da
t
a
c
o
r
r
e
c
t
l
y
,
t
h
u
s
i
t
m
u
s
t
b
e
e
l
im
i
na
t
e
d
f
i
r
s
t
.
−
R
e
m
o
va
l
o
f
r
e
pe
a
t
e
d
l
e
tt
e
r
s
a
n
d
t
we
e
t
s
:
I
n
t
we
e
t
s
,
wor
ds
a
r
e
f
r
e
que
n
t
l
y
wr
i
t
t
e
n
i
n
e
x
t
e
n
de
d
o
r
e
x
a
gg
e
r
a
t
e
d
f
o
r
m
s
(
f
o
r
e
x
a
m
p
l
e
,
“
n
o
”
a
n
d
“
n
o
oo!
!
”
)
.
T
o
a
c
c
o
m
m
o
da
t
e
f
o
r
t
hi
s
,
t
h
e
c
o
r
r
e
c
t
l
y
s
pe
ll
e
d
v
a
r
i
a
n
t
o
f
e
a
c
h
l
e
n
gt
hy
v
e
r
s
i
o
n
o
f
a
wo
r
d
h
a
s
b
e
e
n
s
ubs
t
i
t
u
t
e
d.
A
dd
i
t
i
o
n
a
ll
y
,
w
hil
e
e
x
t
r
a
c
t
i
n
g
t
we
e
t
s
,
t
h
e
A
P
I
o
c
c
a
s
i
o
n
a
ll
y
pr
o
vi
d
e
s
i
de
n
t
i
c
a
l
t
we
e
t
s
,
wh
i
c
h
we
r
e
m
o
v
e
d
to
pr
e
v
e
n
t
g
i
vi
ng
a
s
i
ng
l
e
twe
e
t
e
x
c
e
s
s
i
ve
we
i
g
h
t
.
−
R
e
m
o
va
l
o
f
s
t
o
p
w
o
r
ds
:
we
f
i
l
t
e
r
e
d
t
h
e
s
t
o
p
wo
r
ds
us
i
n
g
P
y
t
h
o
n
N
L
T
K
s
t
o
p
w
or
ds
a
s
t
h
e
y
a
r
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
I
S
S
N:
2502
-
4752
C
r
ime
pr
e
diction
us
ing
a
hy
br
id
s
e
nti
me
nt
analys
is
appr
oac
h
bas
e
d
on
the
…
(
M
ohamm
e
d
B
ouk
abo
us
)
1135
c
o
n
s
i
de
r
e
d
n
e
ut
r
a
l
po
l
a
r
i
t
y
a
n
d
a
r
e
n
o
t
us
e
f
u
l
f
o
r
po
l
a
r
i
t
y
de
c
i
s
i
o
n
s
.
−
L
o
we
r
c
a
s
i
n
g:
a
s
s
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
a
l
go
r
i
t
hm
s
i
n
c
l
ude
c
a
s
e
s
e
n
s
i
t
i
v
i
t
y
,
a
ll
c
a
p
i
t
a
l
c
h
a
r
a
c
t
e
r
s
m
us
t
be
c
o
n
v
e
r
t
e
d
to
l
o
we
r
c
a
s
e
.
W
o
r
ds
w
i
t
h
t
h
e
d
i
s
t
i
nc
t
c
a
s
e
wo
u
l
d
ot
h
e
r
w
i
s
e
b
e
c
o
n
s
i
de
r
e
d
a
s
i
nde
p
e
n
d
e
n
t
wo
r
ds
a
n
d
pr
o
c
e
s
s
e
d
a
c
c
o
r
di
n
g
ly
.
F
i
na
ll
y
,
w
e
f
il
t
e
r
o
ur
t
a
gs
to
h
a
v
e
o
nly
a
d
j
e
c
t
i
v
e
s
,
a
dve
r
bs
,
a
n
d
n
o
un
s
a
s
t
h
e
y
a
r
e
t
h
e
m
o
s
t
r
e
l
e
v
a
n
t
P
OS
i
n
S
A
f
o
r
a
ny
l
a
n
gua
ge
[
23]
.
3.
3.
B
u
il
d
in
g
ou
r
m
od
e
l
No
w
t
h
a
t
we
h
a
ve
c
l
e
a
n
e
d
a
n
d
pr
e
pa
r
e
d
o
u
r
da
t
a
,
we
di
vi
de
i
t
i
n
t
o
da
t
a
X
(
t
e
x
t
s
)
a
n
d
l
a
b
e
l
Y
(
s
e
n
t
i
m
e
n
t
s
)
,
a
n
d
t
h
e
n
i
n
t
o
a
r
a
n
do
m
t
r
a
i
ni
ng
s
u
b
s
e
t
(
80
%
)
a
n
d
t
e
s
t
i
n
g
s
u
b
s
e
t
(
20%
)
.
W
e
b
u
i
l
t
o
ur
S
A
m
o
de
l
us
i
n
g
a
hy
br
i
d
a
ppr
o
a
c
h
w
i
t
h
f
i
r
s
t
t
h
e
l
e
xi
c
o
n
-
b
a
s
e
d
to
ge
n
e
r
a
t
e
l
a
b
e
l
e
d
o
ut
pu
t
.
T
h
e
n
,
t
h
e
o
ut
pu
t
wa
s
f
e
d
i
n
t
o
o
ur
B
E
R
T
m
o
de
l
f
o
r
t
r
a
i
ni
ng.
T
hi
s
m
o
d
e
l
us
e
d
t
h
e
ba
s
i
c
B
E
R
T
m
o
de
l
(
B
E
R
T
-
ba
s
e
t
h
a
t
c
o
n
s
i
s
t
s
o
f
1
2
t
r
a
n
s
f
o
r
m
e
r
l
a
y
e
r
s
)
a
n
d
b
u
il
t
o
ur
s
e
n
t
i
m
e
n
t
c
l
a
s
s
if
ier
o
n
to
p
o
f
i
t
.
F
o
r
t
r
a
i
ni
n
g
t
h
e
B
E
R
T
m
o
de
l
,
we
n
e
e
d
to
d
o
s
o
m
e
a
dd
i
t
i
o
n
a
l
pr
e
pr
o
c
e
s
s
i
ng:
−
A
dd
s
pe
c
i
a
l
t
o
ke
n
s
to
s
e
pa
r
a
t
e
s
e
n
t
e
n
c
e
s
a
n
d
do
c
l
a
s
s
if
i
c
a
t
i
o
n
:
[
S
E
P
]
:
e
n
d
i
n
g
o
f
a
s
e
n
t
e
n
c
e
.
[
C
L
S
]
:
s
t
a
r
t
o
f
e
a
c
h
s
e
n
t
e
n
c
e
.
[
P
A
D]
:
s
pe
c
i
a
l
t
o
ke
n
f
o
r
pa
dd
i
n
g.
[
UN
K
]
:
e
v
e
r
y
t
hi
ng
e
l
s
e
(
un
k
n
o
wn
t
o
ke
n
)
.
−
P
a
s
s
t
h
e
s
e
que
n
c
e
s
w
i
t
h
a
c
o
n
s
t
a
n
t
l
e
n
gt
h
(
pa
dd
i
ng
)
,
t
h
a
t
we
l
im
i
t
t
o
512
to
ke
n
s
.
−
C
r
e
a
t
e
a
n
a
r
r
a
y
o
f
pa
dde
d
t
o
ke
n
s
(
0
s
)
a
n
d
r
e
a
l
o
n
e
s
(
1
s
)
c
a
l
l
e
d
a
t
t
e
n
t
i
o
n
m
a
s
k.
W
e
t
r
a
i
n
e
d
o
ur
m
o
de
l
o
n
t
h
r
e
e
e
po
c
hs
(
B
E
R
T
m
o
de
l
s
a
r
e
a
l
r
e
a
d
y
pr
e
-
t
r
a
i
n
e
d,
a
n
d
a
de
li
c
a
t
e
f
i
ne
-
t
uni
n
g
us
ua
ll
y
g
i
ve
s
b
e
t
t
e
r
r
e
s
ul
t
s
)
w
i
t
h
a
ba
t
c
h
s
ize
o
f
s
i
x
(
t
h
e
t
ot
a
l
n
u
m
be
r
o
f
t
r
a
i
ni
ng
s
a
m
p
l
e
s
i
n
a
b
a
t
c
h
)
a
n
d
a
l
e
a
r
ni
ng
r
a
t
e
o
f
2e
-
5.
T
o
ge
t
t
h
e
pr
e
d
i
c
t
e
d
pr
o
b
a
bil
i
t
i
e
s
f
r
o
m
o
ur
t
r
a
i
n
e
d
m
o
de
l
,
we
a
pp
li
e
d
t
h
e
s
o
f
t
m
a
x
f
u
n
c
t
i
o
n
t
o
t
h
e
o
u
tpu
t
s
.
On
t
h
e
pr
e
d
i
c
t
i
o
n
s
t
e
p,
we
us
e
d
a
f
o
r
wa
r
d
pa
s
s
t
o
c
o
m
put
e
l
o
g
i
t
s
a
n
d
s
o
f
t
m
a
x
t
o
c
a
l
c
u
l
a
t
e
pr
o
b
a
bi
li
t
i
e
s
.
4.
E
XP
E
R
I
M
E
NT
S
AN
D
RE
S
UL
T
S
4.
1.
T
e
c
h
n
ical
r
e
s
ou
r
c
e
s
T
h
e
f
o
l
l
o
w
i
n
g
h
a
r
dwa
r
e
s
pe
c
s
w
e
r
e
us
e
d
i
n
o
ur
s
t
udi
e
s
o
n
t
h
e
M
A
R
W
AN
hi
g
h
-
pe
r
f
o
r
m
a
n
c
e
c
o
m
put
i
n
g
(
HPC
)
i
nf
r
a
s
t
r
uc
t
ur
e
:
−
C
P
U:
2
I
n
t
e
l
Xe
o
n
Go
l
d
6148,
−
R
AM
:
192
GB
,
−
GPU:
2x
NV
I
DI
A
T
e
s
l
a
P
100
(
12
GB
)
.
W
e
ut
i
li
z
e
d
K
e
r
a
s
(
2.
4.
0)
[
24]
,
a
n
o
pe
n
-
s
o
ur
c
e
p
y
t
h
o
n
D
L
f
r
a
m
e
wo
r
k
t
h
a
t
o
pe
r
a
t
e
s
o
n
to
p
o
f
Go
o
gl
e
'
s
o
pe
n
-
s
o
ur
c
e
da
t
a
f
l
o
w
s
o
f
t
wa
r
e
,
a
n
d
T
e
n
s
o
r
F
l
o
w
-
GPU
[
25]
a
s
t
h
e
b
a
c
ke
n
d
e
n
g
i
ne
i
n
o
ur
e
x
pe
r
i
m
e
n
t
.
4.
2.
E
val
u
at
ion
m
e
t
r
ics
T
o
e
v
a
l
ua
t
e
o
ur
m
o
de
l
,
we
us
e
d
t
h
e
f
o
l
l
o
w
i
ng
m
e
t
r
i
c
s
:
a
cc
ur
a
c
y
,
l
o
s
s
,
pr
e
c
i
s
i
o
n
,
r
e
c
a
l
l
,
F
1
-
s
c
o
r
e
,
a
n
d
t
h
e
c
o
nf
u
s
i
o
n
m
a
t
r
i
x
.
−
A
c
c
ur
a
c
y
:
i
s
t
h
e
pe
r
c
e
n
t
a
ge
o
f
c
o
r
r
e
c
t
p
r
e
d
i
c
t
i
o
ns
a
m
o
n
g
a
ll
pr
e
d
i
c
t
i
o
n
s
a
s
s
h
o
wn
i
n
(
1
)
.
−
L
o
s
s
:
i
s
t
h
e
d
i
f
f
e
r
e
n
c
e
b
e
t
we
e
n
t
h
e
m
o
de
l
's
pr
e
d
i
c
t
e
d
v
a
l
ue
a
n
d
t
h
e
a
c
t
ua
l
v
a
l
ue
.
C
r
o
s
s
-
e
n
t
r
o
py
i
s
t
h
e
m
o
s
t
w
i
de
ly
us
e
d
l
o
s
s
f
u
n
c
t
i
o
n
i
n
D
L
a
s
s
h
o
wn
i
n
(
2
)
,
wh
e
r
e
p(
x
)
i
s
t
h
e
r
e
a
l
d
i
s
t
r
i
b
ut
i
o
n
a
n
d
q(
x
)
i
s
t
h
e
e
s
t
i
m
a
t
e
d
d
i
s
t
r
i
b
ut
i
o
n
,
b
o
t
h
s
pe
c
i
f
i
e
d
o
v
e
r
t
h
e
d
i
s
c
r
e
t
e
v
a
r
i
a
bl
e
x
[
26]
.
=
+
+
+
+
(
1
)
(
,
)
=
−
∑
(
)
log
(
(
)
)
∀
(
2
)
W
h
e
r
e
:
T
r
ue
p
o
s
i
t
i
v
e
(
)
:
i
s
t
h
e
n
u
m
be
r
o
f
s
uc
c
e
s
s
f
u
l
ly
c
l
a
s
s
if
i
e
d
po
s
i
t
i
ve
c
l
a
s
s
r
e
c
o
r
ds
.
T
r
ue
n
e
ga
t
i
v
e
(
)
:
i
s
t
h
e
n
u
m
be
r
o
f
s
uc
c
e
s
s
f
u
l
ly
c
l
a
s
s
if
i
e
d
n
e
g
a
t
i
ve
c
l
a
s
s
r
e
c
o
r
ds
.
F
a
l
s
e
po
s
i
t
i
v
e
(
)
:
i
s
t
h
e
n
u
m
be
r
o
f
i
n
c
o
r
r
e
c
t
l
y
c
l
a
s
s
if
i
e
d
n
e
ga
t
i
v
e
c
l
a
s
s
r
e
c
o
r
ds
.
F
a
l
s
e
n
e
ga
t
i
v
e
(
)
:
i
s
t
h
e
n
u
m
be
r
o
f
i
n
c
o
r
r
e
c
t
l
y
c
l
a
s
s
if
i
e
d
po
s
i
t
i
v
e
c
l
a
s
s
r
e
c
o
r
ds
.
−
P
r
e
c
i
s
i
o
n
:
i
s
t
h
e
pe
r
c
e
n
t
a
ge
o
f
a
l
l
po
s
i
t
i
v
e
r
e
s
u
l
t
s
th
a
t
we
r
e
a
c
c
ur
a
t
e
l
y
i
de
n
t
i
f
i
e
d
a
s
s
h
o
wn
i
n
(
3
)
.
−
R
e
c
a
ll
:
i
s
t
h
e
pr
o
por
t
i
o
n
o
f
a
c
c
ur
a
t
e
l
y
i
de
n
t
i
f
i
e
d
po
s
i
t
i
ve
r
e
s
u
l
t
s
a
m
o
n
g
t
h
e
tot
a
l
n
u
m
be
r
o
f
e
xi
s
t
i
n
g
po
s
i
t
i
v
e
c
l
a
s
s
a
s
s
h
o
wn
i
n
(
4
)
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2502
-
4752
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
,
Vo
l
.
25
,
N
o
.
2
,
F
e
b
r
ua
r
y
20
22
:
1131
-
1139
1136
=
+
(
3
)
=
+
(
4
)
−
F1
-
s
c
o
r
e
:
c
o
m
m
o
nly
k
n
o
wn
a
s
t
h
e
F
-
s
c
o
r
e
,
i
s
t
he
h
a
r
m
o
ni
c
m
e
a
n
o
f
pr
e
c
i
s
i
o
n
a
n
d
r
e
c
a
ll
[
27]
,
w
i
t
h
a
v
a
l
ue
t
h
a
t
e
m
p
ha
s
i
z
e
o
n
e
or
t
h
e
ot
h
e
r
a
s
s
h
o
wn
i
n
(
5
)
.
F
-
s
c
o
r
e
h
a
s
a
m
a
xim
u
m
v
a
l
ue
o
f
1
a
n
d
a
mi
n
im
u
m
v
a
l
ue
o
f
0,
a
n
d
i
s
a
l
wa
y
s
b
e
t
we
e
n
pr
e
c
i
s
i
o
n
a
n
d
r
e
c
a
l
,
F
β
=
(
1+
β
2
)
∙
p
r
e
c
i
s
i
o
n
∙
r
e
c
a
l
l
p
r
e
c
i
s
i
o
n
+
r
e
c
a
l
l
(
5
)
w
h
e
r
e
=
1
,
we
h
a
v
e
t
h
e
s
t
a
n
da
r
d
F
-
m
e
a
s
ur
e
o
r
b
a
l
a
n
c
e
d
1
-
s
c
or
e
a
s
s
h
o
wn
i
n
(
6
)
.
F
1
-
s
co
r
e
=
2
∙
p
r
e
c
i
s
i
o
n
∙
r
e
c
a
l
l
p
r
e
c
i
s
i
o
n
+
r
e
c
a
l
l
(
6
)
−
C
o
nf
us
i
o
n
m
a
t
r
i
x
:
i
s
a
t
a
bl
e
t
h
a
t
vi
s
ua
li
z
e
s
t
h
e
m
o
de
l
's
pe
r
f
o
r
m
a
n
c
e
by
put
t
i
n
g
pr
e
d
i
c
t
e
d
c
l
a
s
s
i
ns
t
a
nc
e
s
i
n
t
h
e
r
o
ws
a
n
d
r
e
a
l
c
l
a
s
s
i
ns
t
a
n
c
e
s
i
n
t
h
e
c
o
l
u
m
ns
.
4.
3.
E
val
u
at
in
g
ou
r
m
od
e
l
T
hi
s
s
t
ud
y
c
o
ns
i
s
t
s
o
f
pr
e
d
i
c
t
i
n
g
c
r
im
e
s
us
i
ng
da
t
a
f
r
o
m
s
o
c
i
a
l
m
e
d
i
a
(
T
w
i
t
t
e
r
)
wi
t
h
a
hy
b
r
i
d
a
ppr
o
a
c
h
(
l
e
xi
c
o
n
-
b
a
s
e
d
a
n
d
B
E
R
T
)
a
n
d
h
a
s
pr
o
vi
de
d
ve
r
y
pr
o
m
i
s
i
ng
r
e
s
u
l
t
s
.
I
n
t
h
e
f
o
ll
o
w
i
ng
f
i
gur
e
s
(
F
i
gur
e
s
3
to
5)
ge
n
e
r
a
t
e
d
by
T
e
n
s
o
r
b
o
a
r
d,
we
p
r
e
s
e
n
t
t
h
e
tr
a
i
ni
ng
a
c
c
ur
a
c
y
,
va
l
i
d
a
t
i
o
n
a
c
c
ur
a
c
y
,
t
r
a
i
ni
ng
l
o
s
s
,
a
n
d
v
a
li
da
t
i
o
n
l
o
s
s
a
n
d
f
o
r
e
a
c
h
c
l
a
s
s
,
we
pr
e
s
e
n
t
t
h
e
R
e
c
a
l
l
a
n
d
P
r
e
c
i
s
i
o
n
,
a
l
o
n
g
w
i
t
h
t
he
F
1
s
c
o
r
e
whi
c
h
wa
s
t
r
a
i
ne
d
o
v
e
r
t
h
r
e
e
e
po
c
hs
.
A
s
s
h
o
wn
i
n
F
i
gur
e
3,
o
ur
m
o
de
l
i
n
t
h
e
s
e
t
h
r
e
e
e
po
c
h
s
(
7
h
o
ur
s
10
mi
n
ut
e
s
o
f
t
r
a
i
ni
ng)
t
h
e
a
c
c
ur
a
c
y
r
e
a
c
h
e
d
97
.
24%
a
n
d
t
h
e
l
o
s
s
a
t
t
a
i
n
e
d
8
.
41%
i
n
t
r
a
i
ni
ng,
a
n
d
f
o
r
t
h
e
v
a
li
da
t
i
o
n
,
i
t
r
e
a
c
h
e
d
94
.
91%
i
n
t
h
e
a
c
c
ur
a
c
y
a
n
d
16
.
26%
i
n
t
h
e
l
o
s
s
a
s
s
h
o
wn
i
n
F
i
gur
e
4
.
R
e
ga
r
d
i
n
g
t
h
e
R
e
c
a
ll
,
P
r
e
c
i
s
i
o
n
,
a
n
d
F
1
-
s
c
o
r
e
,
we
o
b
t
a
i
n
e
d
m
o
r
e
pr
e
c
i
s
e
r
e
s
u
l
t
s
f
o
r
e
a
c
h
o
f
o
ur
t
w
o
c
l
a
s
s
e
s
(
c
r
i
m
e
o
r
n
ot)
.
W
e
s
e
e
t
h
a
t
f
o
r
t
h
e
c
r
i
m
e
-
r
e
l
a
t
e
d
c
l
a
s
s
,
we
got
a
pr
e
c
i
s
i
o
n
o
f
92%
,
a
r
e
c
a
ll
o
f
94%
,
a
n
d
a
n
F
1
-
s
c
o
r
e
o
f
93%
,
a
n
d
f
o
r
t
h
e
n
o
r
m
a
l
c
l
a
s
s
(
n
o
r
m
a
l
t
e
x
t
c
h
a
t
)
,
we
o
b
ta
i
n
e
d
a
pr
e
c
i
s
i
o
n
o
f
96%
,
a
r
e
c
a
ll
o
f
95%
,
a
n
d
a
n
F
1
-
s
c
o
r
e
o
f
95%
a
s
s
h
o
wn
in
F
i
gur
e
5.
F
i
gur
e
3
.
T
r
a
i
ni
ng
a
n
d
va
l
i
d
a
t
i
o
n
a
c
c
ur
a
c
y
F
i
gur
e
4
.
T
r
a
i
ni
ng
a
n
d
va
l
i
d
a
t
i
o
n
l
o
s
s
T
o
v
e
r
i
f
y
t
h
e
pe
r
f
o
r
m
a
n
c
e
o
f
o
ur
m
o
de
l
,
we
c
o
m
p
a
r
e
d
t
h
e
r
e
s
u
l
t
s
o
b
t
a
i
n
e
d
(
a
c
c
ur
a
c
y
a
n
d
F
1
-
s
c
o
r
e
)
w
i
t
h
o
t
h
e
r
pa
pe
r
s
'
r
e
s
u
l
t
s
t
h
a
t
a
l
s
o
us
e
d
T
w
i
t
t
e
r
a
s
t
h
e
da
t
a
s
e
t
a
n
d
de
a
l
t
w
i
t
h
t
h
e
s
e
c
ur
i
t
y
i
n
t
e
ll
i
ge
n
c
e
f
i
e
l
d
a
s
s
h
o
w
n
i
n
T
a
bl
e
1.
W
e
c
a
n
s
e
e
t
h
a
t
t
h
e
hy
br
i
d
a
p
pr
o
a
c
h
us
i
n
g
t
h
e
N
L
P
’
s
m
o
de
l
B
E
R
T
g
i
v
e
s
b
e
t
t
e
r
r
e
s
u
l
t
s
t
h
a
n
t
h
e
l
e
xi
c
o
n
-
b
a
s
e
d
a
n
d
m
a
c
hi
ne
l
e
a
r
ni
ng
a
pp
r
o
a
c
h
e
s
i
n
b
o
t
h
a
c
c
ur
a
c
y
a
n
d
F
1
-
s
c
o
r
e
m
e
t
r
i
c
s
,
a
s
B
E
R
T
o
u
t
pe
r
f
o
r
m
t
h
e
o
t
h
e
r
NL
P
’
s
m
o
de
l
i
n
c
o
n
t
e
x
t
un
d
e
r
s
t
a
n
d
i
n
g
pa
r
t
,
e
s
pe
c
i
a
l
ly
i
n
t
h
e
“
c
o
n
t
e
x
t
h
e
a
vy
”
o
n
e
,
f
o
r
whi
c
h
i
t
i
s
s
o
i
m
po
r
t
a
n
t
f
o
r
a
n
a
ly
t
i
c
s
pur
po
s
e
s
.
0.82
0.84
0.86
0.88
0.9
0.92
0.94
0.96
0.98
0
1
2
3
4
A
c
c
ur
a
c
y
E
po
c
hs
T
r
a
in
in
g
V
a
li
da
ti
on
0
0.05
0.1
0.15
0.2
0.25
0.3
0
1
2
3
4
L
o
s
s
E
po
c
hs
T
r
a
in
in
g
V
a
li
da
ti
on
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
I
S
S
N:
2502
-
4752
C
r
ime
pr
e
diction
us
ing
a
hy
br
id
s
e
nti
me
nt
analys
is
appr
oac
h
bas
e
d
on
the
…
(
M
ohamm
e
d
B
ouk
abo
us
)
1137
F
i
gur
e
5
.
M
o
de
l
c
o
nf
us
i
o
n
m
a
t
r
i
x
T
a
bl
e
1.
C
o
m
pa
r
i
s
o
n
o
f
o
ur
r
e
s
u
l
t
s
w
i
t
h
t
h
o
s
e
o
f
oth
e
r
pa
pe
r
s
P
a
pe
r
A
ppr
o
a
c
h
A
c
c
u
r
a
c
y
F1
-
s
c
o
r
e
O
ur
a
ppr
o
a
c
h
H
y
br
id
(
l
e
x
ic
o
n
-
ba
s
e
d +
B
E
R
T
)
94.91%
94.92%
[
28]
M
a
c
hi
ne
L
e
a
r
ni
ng
-
S
V
M
91.55%
-
M
a
c
hi
ne
L
e
a
r
ni
ng
-
C
N
B
88.17%
-
M
a
c
hi
ne
L
e
a
r
ni
ng
-
DT
82.46%
-
M
a
c
hi
ne
L
e
a
r
ni
ng
-
K
N
N
78.06%
-
[
29]
L
e
x
i
c
o
n
-
ba
s
e
d
-
55%
[
30]
M
a
c
hi
ne
L
e
a
r
ni
ng
-
M
N
B
-
83%
M
a
c
hi
ne
L
e
a
r
ni
ng
-
S
V
M
-
89%
M
a
c
hi
ne
L
e
a
r
ni
ng
-
RF
-
87%
[
31]
L
e
x
i
c
o
n
-
ba
s
e
d
93%
-
T
h
e
i
s
s
ue
o
f
a
n
a
ly
z
i
ng
a
n
d
pr
o
c
e
s
s
i
n
g
E
n
g
li
s
h
m
e
t
a
ph
o
r
i
c
a
l
,
s
a
r
c
a
s
t
i
c
,
a
n
d
e
n
c
r
y
pt
e
d
e
x
pr
e
s
s
i
o
ns
i
n
t
h
e
c
o
n
t
e
x
t
o
f
s
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
pr
o
duc
e
s
a
de
gr
a
da
t
i
o
n
i
n
t
h
e
pe
r
f
o
r
m
a
n
c
e
o
f
t
h
e
m
o
de
l
w
h
e
n
i
t
c
o
m
e
s
t
o
t
h
e
s
e
t
y
p
e
s
o
f
s
e
n
t
e
n
c
e
s
.
Al
t
h
o
ugh
s
u
b
s
t
a
n
t
i
a
l
pr
o
gr
e
s
s
h
a
s
b
e
e
n
a
c
hi
e
ve
d
i
n
t
h
e
s
e
a
r
e
a
s
o
f
s
t
ud
y
,
pr
e
s
e
n
t
c
o
m
put
a
t
i
o
n
a
l
m
o
de
l
s
a
n
d
m
e
t
h
o
ds
a
r
e
s
t
i
ll
u
n
a
bl
e
to
h
a
n
d
l
e
t
h
e
s
e
s
o
r
t
s
o
f
e
x
pr
e
s
s
i
o
n
s
.
A
m
e
t
a
ph
o
r
i
s
a
f
i
gur
e
o
f
s
pe
e
c
h
t
h
a
t
pr
o
vi
de
s
a
l
a
n
gua
ge
t
e
c
hni
que
f
o
r
c
o
n
v
e
yi
ng
t
h
o
ugh
t
s
a
n
d
n
o
t
i
o
n
s
t
h
a
t
a
r
e
n
ot
t
h
e
s
a
m
e
a
s
t
h
e
y
a
ppe
a
r
o
n
t
h
e
s
ur
f
a
c
e
,
i
t
di
r
e
c
t
l
y
r
e
f
e
r
s
t
o
o
n
e
t
hi
n
g
by
m
e
n
t
i
o
ni
ng
a
n
o
t
h
e
r
,
a
n
d
to
e
f
f
i
c
i
e
n
t
l
y
i
d
e
n
t
i
f
y
i
t
s
hi
dde
n
s
e
n
ti
m
e
n
t
,
t
h
e
m
o
de
l
m
us
t
b
e
c
a
pa
bl
e
o
f
r
e
a
d
i
ng
b
e
t
we
e
n
t
he
li
ne
s
.
I
n
a
s
a
r
c
a
s
t
i
c
m
e
s
s
a
ge
,
pe
o
pl
e
us
e
o
p
t
i
mi
s
t
i
c
p
h
r
a
s
e
s
t
o
e
x
pr
e
s
s
t
h
e
i
r
b
a
d
f
e
e
li
ng
s
.
B
e
c
a
u
s
e
o
f
t
hi
s
,
s
a
r
c
a
s
m
m
a
y
r
e
a
d
il
y
m
i
s
l
e
a
d
S
A
m
o
de
l
s
unl
e
s
s
t
h
e
y
a
r
e
e
x
pr
e
s
s
ly
c
o
n
s
t
r
uc
t
e
d
to
a
c
c
o
un
t
f
o
r
i
t
.
I
t
c
a
n
be
d
i
f
f
i
c
u
l
t
t
o
de
t
e
r
m
i
ne
i
t
s
s
e
n
t
i
m
e
n
t
w
i
t
h
o
ut
a
t
h
o
r
o
ugh
a
wa
r
e
n
e
s
s
o
f
t
h
e
s
i
t
ua
t
i
o
n
,
t
h
e
s
u
bj
e
c
t
,
a
n
d
t
h
e
s
ur
r
o
un
d
i
n
g
e
nvi
r
o
nm
e
n
t
.
C
r
y
pt
o
g
r
a
phy
f
o
c
us
o
n
pr
ot
e
c
t
i
n
g
m
e
s
s
a
g
e
s
(
e
ns
ur
i
n
g
c
o
nf
i
d
e
n
t
i
a
l
i
t
y
,
a
ut
h
e
n
t
i
c
i
t
y
,
a
n
d
i
n
t
e
gr
i
t
y
)
by
o
f
t
e
n
us
i
ng
s
e
c
r
e
t
ke
y
s
.
I
t
m
a
k
e
s
a
m
e
s
s
a
ge
s
uppo
s
e
d
l
y
u
ni
n
t
e
ll
i
g
i
b
l
e
t
o
a
ny
o
n
e
o
th
e
r
t
h
a
n
t
h
o
s
e
wh
o
a
r
e
e
n
t
i
t
l
e
d
t
o
i
t
,
a
n
d
s
o
i
t
i
s
n
o
l
o
n
ge
r
po
s
s
i
bl
e
to
un
de
r
s
t
a
n
d
i
t
s
c
o
n
t
e
n
t
wi
t
h
o
ut
m
o
r
e
kn
o
w
l
e
dge
.
5.
CONC
L
USI
ON
T
hi
s
p
a
pe
r
pr
e
s
e
n
t
s
a
c
r
i
m
e
de
t
e
c
t
i
o
n
a
n
d
c
l
a
s
s
i
f
ica
t
i
o
n
m
e
t
h
o
d
b
a
s
e
d
o
n
a
l
e
xi
c
o
n
-
ba
s
e
d
a
ppr
o
a
c
h
c
o
m
bi
ne
d
w
i
t
h
t
h
e
B
E
R
T
de
e
p
l
e
a
r
ni
ng
a
l
go
r
i
t
hm
.
T
h
e
c
l
a
s
s
if
i
c
a
t
i
o
n
a
c
c
ur
a
c
y
o
f
t
h
e
m
e
t
h
o
d
us
e
d
i
n
t
hi
s
a
r
t
i
c
l
e
o
n
t
h
e
T
w
i
t
t
e
r
da
t
a
s
e
t
r
e
a
c
h
e
d
94.
91%
,
a
l
o
n
g
w
i
t
h
16.
26%
l
o
s
s
,
94.
94%
pr
e
c
i
s
i
o
n
,
94.
91%
r
e
c
a
l
l
,
a
n
d
94.
92
F
1
-
s
c
o
r
e
.
I
t
c
a
n
b
e
s
e
e
n
f
r
o
m
t
h
e
e
x
pe
r
i
m
e
n
t
a
l
r
e
s
u
l
t
s
t
h
a
t
t
h
e
m
e
t
h
o
d
us
e
d
i
n
t
hi
s
a
r
t
i
c
l
e
c
a
n
b
e
e
f
f
e
c
t
i
v
e
f
o
r
c
r
i
m
e
de
t
e
c
t
i
o
n
.
F
ur
t
h
e
r
m
o
r
e
,
i
t
c
a
n
b
e
o
b
s
e
r
ve
d
t
h
a
t
t
h
e
hy
br
i
d
m
e
t
h
o
d
b
a
s
e
d
on
B
E
R
T
o
u
t
pe
r
f
o
r
m
s
e
xi
s
t
i
n
g
wo
r
ks
i
n
c
r
i
m
e
d
e
t
e
c
t
i
o
n
.
A
s
f
ut
ur
e
wo
r
ks
,
we
wi
ll
de
t
e
c
t
c
r
i
m
e
us
i
ng
ot
h
e
r
m
e
d
i
a
s
uc
h
a
s
a
ud
i
o
,
i
m
a
ge
s
,
a
n
d
vi
d
e
o
s
b
a
s
e
d
o
n
t
hi
s
m
o
de
l
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2502
-
4752
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
,
Vo
l
.
25
,
N
o
.
2
,
F
e
b
r
ua
r
y
20
22
:
1131
-
1139
1138
AC
K
NOWL
E
DGE
M
E
NT
S
T
hi
s
r
e
s
e
a
r
c
h
w
a
s
s
uppo
r
t
e
d
t
h
r
o
ugh
c
o
m
put
a
t
i
o
n
a
l
r
e
s
o
ur
c
e
s
o
f
HPC
-
M
A
R
W
AN
(
ww
w.
m
a
r
wa
n
.
m
a
/
h
pc
)
pr
o
vi
de
d
by
t
h
e
Na
t
i
o
n
a
l
C
e
n
t
e
r
f
o
r
S
c
i
e
n
t
i
f
i
c
a
n
d
T
e
c
hni
c
a
l
R
e
s
e
a
r
c
h
(
C
NR
S
T
)
,
R
a
b
a
t
,
M
o
r
o
c
c
o
.
RE
F
E
R
E
NC
E
S
[
1]
I
.
I
dr
is
s
i,
M
.
B
o
uka
b
o
us
,
M
.
A
z
iz
i,
O
.
M
o
us
s
a
o
ui
,
a
nd
H
.
E
l
F
a
di
li
,
“
T
o
w
a
r
d
a
de
e
p
l
e
a
r
ni
ng
-
ba
s
e
d
in
t
r
us
io
n
d
e
t
e
c
ti
o
n
s
y
s
t
e
m
f
or
I
oT
a
ga
in
s
t
bo
tn
e
t
a
tt
a
c
ks
,”
I
A
E
S
I
nt
e
r
nat
io
nal
J
our
nal
of
A
r
ti
f
ic
ia
l
I
nt
e
ll
ig
e
nc
e
(
I
J
-
A
I
)
,
vol
.
10,
n
o
.
1,
pp.
110
-
120,
M
a
r
.
2021, do
i:
10.11591/i
ja
i.
v
10.
i1
.pp110
-
120
.
[
2]
M
.
B
e
r
r
a
ha
l
a
nd
M
.
A
z
iz
i,
“
A
ugme
nt
e
d
bi
na
r
y
mul
ti
-
la
be
l
e
d
C
N
N
f
or
pr
a
c
ti
c
a
l
f
a
c
ia
l
a
tt
r
ib
ut
e
c
la
s
s
if
i
c
a
ti
o
n
,”
I
ndone
s
ia
n
J
our
nal
o
f
E
le
c
tr
i
c
al
E
ngi
ne
e
r
in
g
and
C
om
put
e
r
Sc
i
e
nc
e
(
I
J
E
E
C
S)
,
vo
l.
23,
no
.
2,
pp.
973
-
979,
2021
,
do
i
:
10.11591
/i
je
e
c
s
.v
23.i
2.pp973
-
979
.
[
3]
E
.
A
.
K
ir
il
l
ov
a
,
R
.
A
.
K
ur
ba
n
ov
,
N
.
V
.
S
ve
c
hni
k
ov
a
,
T
.
E
.
Z
ul
’
f
uga
r
z
a
d
e
,
a
nd
S
.
S
.
Z
e
ni
n,
“
P
r
o
bl
e
ms
of
f
ig
ht
in
g
c
r
im
e
s
o
n
t
he
in
te
r
n
e
t,
”
J
our
nal
of
A
dv
anc
e
d R
e
s
e
a
r
c
h i
n L
aw
and E
c
onomic
s
, vo
l.
8, n
o
. 3, pp. 849
–
856, J
un. 2017.
[
4]
M
. S
a
lt
e
r
,
C
r
ime
, j
us
ti
c
e
and s
oc
ia
l
m
e
di
a
, E
ngl
a
nd, U.K
.:
R
o
u
tl
e
dg
e
, J
a
n. 2016
.
[
5]
L
.
T
or
r
e
y
a
nd
J
.
S
ha
v
li
k,
“
T
r
a
ns
f
e
r
le
a
r
n
in
g,”
in
H
andbook
of
r
e
s
e
ar
c
h
on
m
ac
hi
ne
le
a
r
ni
ng
appl
ic
at
io
ns
and
tr
e
n
ds
:
al
gor
it
hm
s
,
m
e
th
ods
,
and
te
c
hni
que
s
,
e
di
te
d
b
y
E
.
S
or
ia
,
J
.
M
a
r
ti
n,
R
.
M
a
gda
le
na
,
M
.
M
a
r
ti
ne
z
,
a
nd
A
.
S
e
r
r
a
no
,
I
G
I
gl
oba
l,
2009
, pp. 242
–
264.
[
6]
I
.
I
dr
is
s
i,
M
.
A
z
iz
i,
a
nd
O
.
M
o
us
s
a
o
ui
,
“
A
c
c
e
l
e
r
a
ti
ng
th
e
u
p
da
te
of
a
D
L
-
b
a
s
e
d
I
D
S
f
or
I
oT
us
in
g
de
e
p
tr
a
ns
f
e
r
l
e
a
r
ni
n
g
,”
I
ndone
s
ia
n
J
our
nal
o
f
E
le
c
tr
i
c
al
E
ngi
n
e
e
r
in
g
and
C
o
m
put
e
r
Sc
ie
nc
e
(
I
J
E
E
C
S)
,
vo
l.
23,
n
o
.
2,
pp.
1059
-
1067,
2021
,
do
i
:
10.11591/i
je
e
c
s
.v
23.i
2.pp1059
-
1067
.
[
7]
M
. B
o
uka
bo
us
a
nd M
. A
z
iz
i,
“
R
e
v
i
e
w
of
L
e
a
r
n
in
g
-
B
a
s
e
d
T
e
c
h
ni
que
s
of
S
e
nt
im
e
nt
A
na
l
y
s
is
f
or
S
e
c
ur
it
y
P
ur
p
o
s
e
s
,”
I
nnov
at
io
ns
in
Sm
ar
t
C
it
ie
s
A
ppl
ic
at
io
ns
V
ol
um
e
4.
SC
A
2020.
L
e
c
tu
r
e
N
ot
e
s
in
N
e
tw
or
k
s
and
S
y
s
te
m
s
,
v
o
l.
183,
pp.
96
–
109
,
d
o
i
:
10.1007/978
-
3
-
030
-
66840
-
2
_
8.
[
8]
A
. V
a
s
w
a
ni
e
t
al
.
, “
A
tt
e
nt
i
o
n i
s
a
ll
y
o
u n
e
e
d,”
ar
X
iv
. 2017.
[
9]
F
.
K
.
V
e
nt
ir
oz
o
s
,
I
.
V
a
r
la
mi
s
,
a
nd
G
.
T
s
a
ts
a
r
o
ni
s
,
“
D
e
t
e
c
ti
ng
a
ggr
e
s
s
iv
e
be
ha
v
i
or
in
di
s
c
us
s
io
n
th
r
e
a
ds
us
in
g
te
x
t
mi
ni
ng,”
I
n:
G
e
lb
ukh A. (
e
ds
)
C
om
put
at
io
nal
L
in
gui
s
ti
c
s
and I
nt
e
ll
ig
e
nt
T
e
x
t
P
r
oc
e
s
s
in
g. C
I
C
L
in
g 2017. L
e
c
tu
r
e
N
ot
e
s
i
n C
om
put
e
r
Sc
ie
n
c
e
,
vo
l.
10762,
2018
,
d
o
i:
10.1007/978
-
3
-
319
-
77116
-
8_31.
[
10]
P
.
S
ir
ia
r
a
y
a
e
t
al
.
,
“
W
it
n
e
s
s
in
g
c
r
im
e
th
r
o
ugh
tw
e
e
ts
:
A
c
r
i
me
in
ve
s
ti
ga
ti
o
n
t
oo
l
ba
s
e
d
o
n
s
oc
ia
l
m
e
di
a
,”
SI
G
S
P
A
T
I
A
L
'
19:
P
r
oc
e
e
di
ngs
of
th
e
27t
h
A
C
M
SI
G
SP
A
T
I
A
L
I
nt
e
r
nat
io
nal
C
onf
e
r
e
nc
e
on
A
dv
an
c
e
s
in
G
e
ogr
aphi
c
I
n
f
or
m
at
io
n
Sy
s
te
m
s
,
20
19
,
pp. 568
–
571,
do
i:
10.1145/3347146.3
359082.
[
11]
J
.
C
.
P
e
r
e
ir
a
-
K
o
ha
ts
u,
L
.
Q
ui
ja
no
-
S
á
n
c
he
z
,
F
.
L
ib
e
r
a
t
or
e
,
a
nd
M
.
C
a
ma
c
ho
-
C
o
l
la
do
s
,
“
D
e
t
e
c
ti
ng
a
nd
mo
ni
t
o
r
in
g
ha
te
s
pe
e
c
h
in
twi
tt
e
r
,”
Se
ns
or
s
(
Sw
it
z
e
r
la
nd)
,
vo
l.
19, n
o
. 21,
2019, d
o
i
:
10.3
390/
s
19214654.
[
12]
M
.
B
o
uka
b
o
us
a
nd
M
.
A
z
iz
i,
“
A
c
o
mpa
r
a
ti
ve
s
tu
d
y
of
D
L
-
ba
s
e
d
la
ngua
ge
r
e
pr
e
s
e
nt
a
ti
o
n
l
e
a
r
ni
ng
m
o
d
e
ls
,”
I
ndone
s
ia
n
J
our
nal
of
E
le
c
tr
i
c
al
E
ngi
ne
e
r
in
g
and
C
om
put
e
r
Sc
i
e
nc
e
(
I
J
E
E
C
S)
,
vo
l.
22,
n
o
.
2,
pp.
1032
–
1040,
2021,
d
o
i
:
10.11591/i
je
e
c
s
.v
22.i
2.pp1032
-
1040.
[
13]
J
.
Y
a
da
v
,
D
.
K
uma
r
,
a
nd
D
.
C
ha
uh
a
n,
“
C
y
b
e
r
bul
l
y
in
g
D
e
t
e
c
t
i
o
n
us
in
g
P
r
e
-
T
r
a
in
e
d
B
E
R
T
M
o
d
e
l,
”
2020
I
nt
e
r
nat
io
nal
C
onf
e
r
e
nc
e
on
E
le
c
tr
oni
c
s
and
Su
s
ta
in
abl
e
C
om
m
u
ni
c
at
io
n
Sy
s
te
m
s
(
I
C
E
SC
)
,
2020,
pp.
1096
-
1100,
do
i:
10.1109/I
C
E
S
C
48915.2020.9155700
.
[
14]
G
.
R
.
S
.
W
e
ir
,
E
.
D
o
s
S
a
nt
o
s
,
B
.
C
a
r
twr
ig
ht
,
a
nd
R
.
F
r
a
nk,
“
P
o
s
it
in
g
th
e
p
r
o
bl
e
m:
E
nha
n
c
in
g
c
la
s
s
if
i
c
a
ti
o
n
of
e
x
t
r
e
mi
s
t
w
e
b
c
o
nt
e
nt
th
r
o
ugh
t
e
x
tu
a
l
a
na
l
y
s
is
,”
2016
I
E
E
E
I
nt
e
r
nat
io
nal
C
o
nf
e
r
e
nc
e
on
C
y
be
r
c
r
im
e
and
C
om
put
e
r
F
or
e
ns
ic
(
I
C
C
C
F
)
,
20
16,
pp. 1
-
3, do
i
:
10.1109/I
C
C
C
F
.2016.7740431
.
[
15]
B
. P
a
ng
a
nd
L
.
L
e
e
, “
O
pi
ni
o
n m
in
in
g a
nd s
e
nt
im
e
nt
a
na
l
y
s
is
,”
F
ounda
ti
ons
and T
r
e
nds
i
n
I
nf
or
m
at
io
n R
e
tr
ie
v
al
, vo
l.
2,
n
o
. 1
–
2,
pp. 1
–
135, 2008, do
i:
10.1561/150000000
1.
[
16]
K
.
G
e
l
in
e
a
u
a
nd
J
.
G
a
mbr
e
ll
,
“
N
e
w
Z
e
a
la
nd
M
o
s
que
s
h
oo
t
e
r
i
s
a
w
hi
te
na
ti
o
na
li
s
t
w
ho
ha
te
s
im
mi
g
r
a
nt
s
,
doc
um
e
nt
s
a
nd
v
id
e
o
r
e
ve
a
l,
”
C
hi
c
ago T
r
ib
une
, 2019.
A
c
c
e
s
s
e
d:
A
p
r
. 09, 2011.
[
O
nl
in
e
]
. A
v
a
il
a
bl
e
:
ht
tp
s
:/
/ww
w
.c
hi
c
a
g
o
tr
ib
un
e
.
c
o
m/
na
ti
o
n
-
w
o
r
ld
/
c
t
-
mo
s
que
-
ki
ll
e
r
-
w
hi
te
-
s
upr
e
ma
c
y
-
20190315
-
s
to
r
y
.ht
ml
[1
7]
T
.
L
a
w
a
nd
J
.
B
a
te
s
,
“
E
l
P
a
s
o
S
h
oo
ti
ng
S
us
p
e
c
t
T
o
ld
P
o
li
c
e
H
e
W
a
s
T
a
r
ge
ti
ng
‘
M
e
x
i
c
a
ns
.’
H
e
r
e
’
s
W
ha
t
to
K
n
o
w
A
bo
u
t
th
e
C
a
s
e
,”
T
ime
, 2019.
A
c
c
e
s
s
e
d:
A
pr
. 09, 2011.
[
O
nl
in
e
]
. A
v
a
il
a
bl
e
:
ht
tp
s
:/
/t
im
e
.
c
o
m/
5643110/
e
l
-
pa
s
o
-
t
e
x
a
s
-
ma
ll
-
s
h
oo
ti
ng/
[
18]
F
.
G
a
r
dn
e
r
,
“
G
e
r
ma
n
y
s
h
oo
ti
ng:
‘
F
a
r
-
r
ig
ht
e
x
tr
e
mi
s
t’
c
a
r
r
ie
d
out
s
hi
s
ha
ba
r
s
a
tt
a
c
ks
-
B
B
C
N
e
w
s
,”
B
B
C
N
e
w
s
,
2020.
A
c
c
e
s
s
e
d:
A
pr
. 09, 2011.
[
O
nl
in
e
]
.
A
v
a
il
a
bl
e
:
ht
tp
s
:/
/ww
w
.bbc
.
c
o
m/
n
e
w
s
/wo
r
ld
-
e
u
r
o
pe
-
51567971
[
19]
M
.
H
u
a
nd
B
.
L
iu
,
“
M
in
in
g
a
nd
s
umm
a
r
i
z
in
g
c
us
t
o
m
e
r
r
e
v
i
e
w
s
,”
K
D
D
'
04:
P
r
oc
e
e
di
ngs
of
th
e
te
nt
h
A
C
M
SI
G
K
D
D
in
te
r
nat
io
nal
c
onf
e
r
e
nc
e
on K
now
le
dge
di
s
c
ov
e
r
y
and data min
in
g
,
2004
, pp. 168
–
177,
do
i:
10.1145/1014052.1
014073.
[
20]
F
B
I
,
“
F
e
d
e
r
a
l
B
ur
e
a
u
of
in
ve
s
ti
ga
ti
o
n,
c
r
im
e
da
ta
e
x
pl
o
r
e
r
,
”
F
B
I
U
ni
f
or
m
C
r
im
e
R
e
por
ti
ng
P
r
ogr
am
,
2021.
A
c
c
e
s
s
e
d:
A
pr
.
08,
2021. [
O
nl
in
e
]
. A
v
a
il
a
bl
e
:
ht
tp
s
:/
/
c
r
i
me
-
da
ta
-
e
x
pl
or
e
r
.
f
r
.
c
l
o
ud.
gov
/d
o
w
nl
o
a
ds
-
a
nd
-
d
oc
s
[
21]
C
a
le
b
E
l
f
e
nb
e
in
,
“
D
a
ta
–
M
a
ppi
ng
I
s
la
mo
pho
bi
a
,
”
M
appi
ng
I
s
la
m
ophobia
,
20
21.
A
c
c
e
s
s
e
d:
J
un.
10,
2021.
[
O
nl
in
e
]
.
A
v
a
il
a
b
le
:
ht
tp
s
:/
/m
a
ppi
ngi
s
la
mo
ph
o
bi
a
.
o
r
g/
da
ta
/
[
22]
P
.
B
ur
na
p
a
nd
M
.
L
.
W
il
li
a
ms
,
“
C
y
be
r
ha
te
s
p
e
e
c
h
o
n
twi
tt
e
r
:
A
n
a
ppl
ic
a
ti
o
n
of
ma
c
h
in
e
c
la
s
s
if
ic
a
ti
o
n
a
nd
s
ta
ti
s
ti
c
a
l
m
o
d
e
li
n
g
f
or
p
o
li
c
y
a
nd de
c
is
i
o
n ma
ki
ng,”
P
ol
ic
y
and I
nt
e
r
ne
t
,
vo
l.
7, n
o. 2, pp. 223
-
242,
2015, do
i
:
10.1002/p
o
i3
.85.
[
23]
M
.
S
o
k
o
l
ov
a
a
nd
G
.
L
a
pa
lm
e
,
“
C
la
s
s
if
i
c
a
ti
o
n
of
o
pi
ni
o
ns
w
it
h
no
n
-
a
f
f
e
c
ti
v
e
a
dve
r
bs
a
nd
a
dj
e
c
ti
ve
s
,”
I
nt
e
r
nat
io
nal
C
onf
e
r
e
n
c
e
R
A
N
L
P
2009
-
B
or
ov
e
ts
, B
ul
gar
ia
,
2009
, pp. 421
-
427
.
[
24]
“
K
e
r
a
s
:
th
e
P
y
th
o
n d
e
e
p l
e
a
r
ni
ng A
P
I
,
”
K
e
r
as
. A
c
c
e
s
s
e
d:
A
ug.
18, 2020. [
O
nl
in
e
]
. A
v
a
il
a
bl
e
:
ht
tp
s
:/
/k
e
r
a
s
.i
o
/
[
25]
“
T
e
ns
o
r
F
l
o
w
.”
A
c
c
e
s
s
e
d A
ug. 18, 2020. [
O
nl
in
e
]
. A
v
a
il
a
bl
e
:
ht
tp
s
:/
/ww
w
.t
e
ns
o
r
f
l
o
w
.
o
r
g/
?
hl
=
f
r
[
26]
Z
.
Z
ha
ng
a
nd
M
.
R
.
S
a
bunc
u,
“
G
e
n
e
r
a
li
z
e
d
C
r
o
s
s
E
nt
r
o
p
y
L
o
s
s
f
o
r
T
r
a
in
in
g
D
e
e
p
N
e
u
r
a
l
N
e
tw
or
ks
w
it
h
N
o
is
y
L
a
b
e
l
s
,”
P
r
oc
e
e
di
ngs
of
t
he
32nd I
nt
e
r
nat
io
nal
C
on
f
e
r
e
nc
e
on N
e
u
r
al
I
nf
or
m
at
io
n P
r
oc
e
s
s
in
g Sy
s
te
m
s
, 2018, pp. 8778
–
8788.
[
27]
L
.
D
e
r
c
z
y
ns
ki
,
“
C
o
mpl
e
m
e
n
ta
r
it
y
,
F
-
s
c
or
e
,
a
nd
N
L
P
e
v
a
lu
a
ti
o
n,”
P
r
oc
e
e
di
ngs
of
th
e
T
e
nt
h
I
nt
e
r
nat
io
nal
C
on
f
e
r
e
nc
e
on
L
anguage
R
e
s
our
c
e
s
and E
v
al
uat
io
n (
L
R
E
C
'
16)
,
2016
, pp. 261
-
266
.
[
28]
H
.
A
L
-
S
a
i
f
a
nd
H
.
A
l
-
D
o
s
s
a
r
i,
“
D
e
t
e
c
t
in
g
a
nd
c
la
s
s
i
f
y
in
g
c
r
im
e
s
f
r
o
m
a
r
a
bi
c
twi
tt
e
r
p
o
s
ts
us
in
g
te
x
t
mi
ni
ng
t
e
c
hni
qu
e
s
,”
I
nt
e
r
nat
io
nal
J
our
nal
of
A
dv
anc
e
d
C
om
put
e
r
Sc
ie
nc
e
a
nd
A
ppl
ic
at
io
ns
,
vo
l.
9,
n
o
.
10,
pp.
377
–
387,
2018,
do
i:
10.14569/I
J
A
C
S
A
.2018.091046.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
I
S
S
N:
2502
-
4752
C
r
ime
pr
e
diction
us
ing
a
hy
br
id
s
e
nti
me
nt
analys
is
appr
oac
h
bas
e
d
on
the
…
(
M
ohamm
e
d
B
ouk
abo
us
)
1139
[
29]
S
.
A
gha
ba
ba
e
i
a
nd M
.
M
a
kr
e
hc
hi
,
“
M
in
in
g
T
w
it
t
e
r
da
ta
f
o
r
c
r
i
me
tr
e
nd
pr
e
di
c
ti
o
n,”
I
nt
e
ll
ig
e
nt
D
at
a
A
nal
y
s
is
,
vo
l.
22,
no
.
1,
pp.
117
-
141,
2018, do
i
:
10.3233/I
D
A
-
163183.
[
30]
H
.
E
l
H
a
nna
c
h
a
nd
M
.
B
e
nkha
li
f
a
,
“
W
o
r
dN
e
t
ba
s
e
d
i
mpl
ic
it
a
s
pe
c
t
s
e
n
ti
m
e
nt
a
na
l
y
s
is
f
o
r
c
r
im
e
id
e
nt
i
f
i
c
a
ti
o
n
f
r
o
m
T
w
it
t
e
r
,”
I
nt
e
r
nat
io
nal
J
our
nal
of
A
dv
anc
e
d
C
om
put
e
r
Sc
ie
nc
e
a
nd
A
ppl
ic
at
io
ns
,
vo
l.
9,
n
o
.
12,
pp.
150
-
159,
2018,
d
o
i
:
10.14569/I
J
A
C
S
A
.2018.091222.
[
31]
J
.
H
a
o
a
nd
H
.
D
a
i,
“
S
oc
ia
l
m
e
di
a
c
o
nt
e
nt
a
nd
s
e
nt
im
e
nt
a
na
l
y
s
is
o
n
c
o
ns
ume
r
s
e
c
ur
it
y
b
r
e
a
c
h
e
s
,”
J
our
nal
o
f
F
in
anc
ia
l
C
r
ime
,
vo
l.
23, n
o
. 4,
2016, d
o
i:
10.1108/J
FC
-
01
-
2016
-
0001.
B
I
OG
RA
P
HI
E
S
OF
AU
T
HO
RS
M
o
h
a
m
m
ed
Bo
uk
a
bo
u
s
i
s
a
Ph
.
D
.
c
an
d
i
d
at
e
i
n
Co
m
p
u
t
e
r
E
n
g
i
n
ee
ri
n
g
at
Mo
h
a
mme
d
Fi
rs
t
U
n
i
v
e
rs
i
t
y
i
n
O
u
j
d
a,
Mo
r
o
cc
o
,
w
h
e
r
e
h
e
i
s
c
o
n
d
u
c
t
i
n
g
r
e
s
e
ar
ch
i
n
s
ec
u
r
i
t
y
i
n
t
el
l
i
g
en
ce
u
s
i
n
g
d
ee
p
l
e
ar
n
i
n
g
a
l
g
o
ri
t
h
m
s
i
n
e
x
ch
an
g
e
d
me
s
s
ag
e
s
.
H
e
h
o
l
d
s
a
M.
S
c
.
d
eg
r
ee
i
n
i
n
t
e
r
n
e
t
o
f
t
h
i
n
g
s
fro
m
Si
d
i
Mo
h
a
me
d
B
e
n
A
b
d
e
l
l
ah
U
n
i
v
e
rs
i
t
y
i
n
F
e
z,
Mo
r
o
cc
o
(2
0
1
9
),
a
s
w
el
l
as
a
B.
S
c
.
d
e
g
r
ee
i
n
C
o
m
p
u
t
e
r
E
n
g
i
n
ee
r
i
n
g
fro
m
Mo
h
a
mme
d
F
i
rs
t
U
n
i
v
e
rs
i
t
y
(2
0
1
6
).
Fu
rt
h
e
r
mo
r
e
,
h
e
h
o
l
d
s
s
e
v
e
ra
l
ce
rt
i
f
i
c
at
i
o
n
s
i
n
n
a
t
u
ral
l
an
g
u
a
g
e
p
ro
ce
s
s
i
n
g
,
art
i
fi
ci
al
i
n
t
el
l
i
g
en
ce
,
s
ec
u
r
i
t
y
i
n
t
e
l
l
i
g
e
n
ce,
b
i
g
d
at
a,
a
n
d
cy
b
e
r
s
e
cu
ri
t
y
.
A
d
d
i
t
i
o
n
al
l
y
,
h
e
s
e
r
v
e
d
as
a
re
v
i
ew
e
r
fo
r
v
ari
o
u
s
i
n
t
e
rn
at
i
o
n
a
l
co
n
f
e
r
en
c
e
s
.
H
e
i
s
cu
rre
n
t
l
y
em
p
l
o
y
ed
at
Mo
h
a
mme
d
Fi
rs
t
U
n
i
v
e
rs
i
t
y
as
an
ad
mi
n
i
s
t
rat
i
v
e
.
H
e
c
a
n
b
e
c
o
n
t
ac
t
e
d
at
em
ai
l
:
m
.
b
o
u
k
ab
o
u
s
@
u
m
p
.
a
c
.
m
a
.
P
ro
f
.
D
r.
M
o
s
ta
f
a
A
zi
z
i
rece
i
v
e
d
a
St
at
e
E
n
g
i
n
eer
d
e
g
r
ee
i
n
A
u
t
o
m
at
i
o
n
a
n
d
I
n
d
u
s
t
ri
al
C
o
m
p
u
t
i
n
g
fro
m
t
h
e
E
n
g
i
n
ee
ri
n
g
S
c
h
o
o
l
E
MI
o
f
Rab
at
,
Mo
ro
cc
o
i
n
1
9
9
3
,
t
h
e
n
a
Mas
t
e
r
d
e
g
r
ee
i
n
A
u
t
o
m
at
i
o
n
an
d
I
n
d
u
s
t
ri
al
Co
m
p
u
t
i
n
g
fr
o
m
t
h
e
Fa
cu
l
t
y
o
f
S
ci
en
ce
s
o
f
O
u
j
d
a,
Mo
r
o
cc
o
i
n
1
9
9
5
,
an
d
a
Ph
.
D
.
d
eg
r
ee
i
n
Co
mp
u
t
e
r
Sc
i
en
ce
fro
m
t
h
e
U
n
i
v
e
rs
i
t
y
o
f
Mo
n
t
re
a
l
,
Can
ad
a
i
n
2
0
0
1
.
H
e
e
ar
n
e
d
al
s
o
t
en
s
o
f
o
n
l
i
n
e
ce
rt
i
fi
c
at
i
o
n
s
i
n
Pr
o
g
ra
mmi
n
g
,
N
e
t
w
o
rk
i
n
g
,
A
I,
Co
m
p
u
t
e
r
S
ecu
ri
t
y.
H
e
i
s
cu
rr
e
n
t
l
y
a
Pro
fe
s
s
o
r
at
t
h
e
E
ST
O
,
U
n
i
v
e
rs
i
t
y
Mo
h
a
mme
d
Fi
rs
t
o
f
O
u
j
d
a.
H
i
s
r
e
s
e
ar
c
h
i
n
t
e
r
e
s
t
s
i
n
c
l
u
d
e
S
ec
u
r
i
t
y
a
n
d
N
e
t
w
o
rk
i
n
g
,
A
I
,
So
ft
w
are
E
n
g
i
n
ee
ri
n
g
,
I
o
T
,
an
d
E
m
b
ed
d
ed
S
y
s
t
em
s
.
H
i
s
re
s
e
ar
ch
fi
n
d
i
n
g
s
w
i
t
h
h
i
s
t
e
a
m
are
p
u
b
l
i
s
h
ed
i
n
o
v
e
r
1
0
0
p
e
e
r
-
re
v
i
ew
ed
co
mm
u
n
i
c
at
i
o
n
s
an
d
p
ap
e
rs
.
H
e
al
s
o
s
e
rv
e
d
as
PC
mem
b
e
r
a
n
d
r
e
v
i
ew
e
r
i
n
s
ev
e
ral
i
n
t
e
rn
at
i
o
n
a
l
co
n
f
e
r
en
ce
s
an
d
j
o
u
rn
a
l
s
.
H
e
c
a
n
b
e
c
o
n
t
ac
t
e
d
at
em
ai
l
:
azi
z
i
.
m
o
s
@
u
m
p
.
a
c
.
m
a
.
Evaluation Warning : The document was created with Spire.PDF for Python.