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.
15
,
No
.
4
,
A
u
g
u
s
t 2
0
2
6
,
p
p
.
3
3
7
6
~
3
3
8
8
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
3
7
6
-
3
3
8
8
3376
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
D
e
te
ct
i
o
n
o
f
i
l
l
e
g
a
l
d
ru
g
a
ct
i
v
it
i
e
s
i
n
I
n
d
o
ne
s
i
a
v
i
a
s
o
ci
a
l
m
ed
i
a
X
Adi H
a
nif
Seda
r
1
,
L
uk
m
a
n Yudo
k
us
um
o
1
,
I
nd
ra
B
u
di
1
,
Am
a
na
h Ra
ma
dia
h
1
,
Aris B
ud
i Sa
nto
s
o
1
,
P
ra
bu
K
re
s
na
P
utr
a
2
1
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
F
a
c
u
l
t
y
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
U
n
i
v
e
r
si
t
a
s I
n
d
o
n
e
si
a
,
J
a
k
a
r
t
a
,
I
n
d
o
n
e
s
i
a
2
R
e
s
e
a
r
c
h
C
e
n
t
e
r
f
o
r
D
a
t
a
S
c
i
e
n
c
e
a
n
d
I
n
f
o
r
ma
t
i
o
n
,
N
a
t
i
o
n
a
l
R
e
se
a
r
c
h
a
n
d
I
n
n
o
v
a
t
i
o
n
A
g
e
n
c
y
(
B
R
I
N
)
,
Ta
n
g
e
r
a
n
g
,
I
n
d
o
n
e
s
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
an
2
7
,
2
0
2
5
R
ev
is
ed
Ma
y
1
7
,
2
0
2
6
Acc
ep
ted
Ma
y
2
4
,
2
0
2
6
In
t
o
d
a
y
'
s
d
ig
i
tal
wo
rl
d
,
so
c
ial
m
e
d
ia
p
latf
o
rm
s
su
c
h
a
s
X
(fo
rm
e
rl
y
Twit
ter
)
a
re
wid
e
ly
u
se
d
fo
r
c
o
m
m
u
n
ica
t
io
n
b
u
t
a
lso
fo
r
il
leg
a
l
a
c
ti
v
it
ies
,
su
c
h
a
s
se
ll
in
g
i
ll
icit
d
ru
g
s.
T
h
is
st
u
d
y
e
x
a
m
in
e
s
h
o
w
p
o
sts
fr
o
m
X
re
late
d
to
i
ll
e
g
a
l
d
ru
g
sa
les
in
In
d
o
n
e
sia
c
a
n
b
e
a
n
a
ly
z
e
d
u
sin
g
m
a
c
h
in
e
lea
rn
in
g
te
c
h
n
iq
u
e
s,
with
a
fo
c
u
s
o
n
i
n
c
o
r
p
o
ra
ti
n
g
fe
a
tu
re
s
re
late
d
to
p
e
rso
n
a
ll
y
i
d
e
n
ti
fiab
le
in
fo
rm
a
ti
o
n
(P
II)
a
n
d
we
b
URLs.
Th
e
m
a
in
q
u
e
stio
n
s
a
re
h
o
w
to
u
s
e
c
las
sifica
ti
o
n
m
e
th
o
d
s
to
d
e
tec
t
t
h
e
se
p
o
sts
a
n
d
w
h
ich
d
ru
g
s
a
re
m
o
st
o
ften
m
e
n
ti
o
n
e
d
.
A
to
tal
o
f
1
1
,
7
7
7
p
o
sts
fr
o
m
X
we
re
c
o
ll
e
c
ted
u
sin
g
we
b
sc
ra
p
in
g
.
Afte
r
c
lea
n
in
g
a
n
d
p
r
o
c
e
ss
in
g
th
e
d
a
ta,
fiv
e
m
a
c
h
in
e
lea
rn
in
g
m
o
d
e
ls
we
re
u
se
d
.
T
h
e
s
u
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
(S
VM)
m
o
d
e
l
s
h
o
we
d
th
e
b
e
st
re
su
lt
s
with
a
we
ig
h
te
d
F
1
-
sc
o
re
o
f
9
7
.
0
8
%
.
T
h
e
fin
d
in
g
s
s
h
o
w
t
h
a
t
a
b
o
rti
o
n
p
il
ls
a
re
t
h
e
m
o
st
fre
q
u
e
n
tl
y
m
e
n
ti
o
n
e
d
d
ru
g
s,
fo
ll
o
we
d
b
y
se
x
u
a
l
e
n
h
a
n
c
e
m
e
n
t
a
n
d
a
n
e
sth
e
ti
c
d
ru
g
s.
Th
is
stu
d
y
a
lso
u
se
d
v
is
u
a
l
to
o
ls
to
sh
o
w
th
e
m
o
st
c
o
m
m
o
n
h
a
sh
tag
s
,
p
o
st
s
d
istri
b
u
ti
o
n
o
v
e
r
t
ime
,
a
n
d
t
y
p
e
s
o
f
d
ru
g
s
d
isc
u
ss
e
d
.
Th
is
stu
d
y
u
n
d
e
rli
n
e
s
th
e
imp
o
rtan
c
e
o
f
m
o
n
i
to
ri
n
g
so
c
ial
m
e
d
ia
to
p
ro
tec
t
p
u
b
li
c
h
e
a
lt
h
a
n
d
h
ig
h
li
g
h
ts
h
o
w
in
c
o
rp
o
ra
ti
n
g
P
II
fe
a
tu
re
s
c
a
n
imp
ro
v
e
t
h
e
d
e
tec
ti
o
n
o
f
il
leg
a
l
a
c
ti
v
it
ies
.
F
u
tu
re
re
se
a
rc
h
sh
o
u
ld
l
o
o
k
a
t
p
o
sts
i
n
o
th
e
r
lan
g
u
a
g
e
s
a
n
d
fro
m
o
th
e
r
p
latfo
rm
s,
c
o
n
si
d
e
rin
g
P
II
fe
a
tu
re
s
fo
r
e
n
h
a
n
c
e
d
a
n
a
ly
sis.
K
ey
w
o
r
d
s
:
I
llicit d
r
u
g
s
Pu
b
lic
h
ea
lth
So
cial
m
ed
ia
T
ex
t c
lass
if
icatio
n
X
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
:
Ad
i H
an
if
Sed
ar
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
,
Facu
lty
o
f
C
o
m
p
u
ter
Scie
n
ce
,
Un
iv
er
s
itas
I
n
d
o
n
esia
J
ak
ar
ta,
I
n
d
o
n
esia
E
m
ail:
ad
i.h
an
if
@
u
i.a
c.
id
1.
I
NT
RO
D
UCT
I
O
N
I
n
to
d
a
y
'
s
d
ig
ital
er
a,
s
o
cial
m
ed
ia
is
a
b
ig
p
ar
t
o
f
o
u
r
d
ai
ly
liv
es,
it
lets
p
eo
p
le
s
h
a
r
e
i
n
f
o
r
m
atio
n
,
id
ea
s
,
an
d
o
p
in
io
n
s
[
1
]
.
X
(
f
o
r
m
er
ly
T
witter
)
,
o
n
e
o
f
th
e
m
o
s
t
p
o
p
u
lar
s
o
cial
m
ed
ia
p
l
atf
o
r
m
s
,
is
u
s
ed
b
y
m
illi
o
n
s
o
f
p
e
o
p
le
e
v
er
y
d
ay
f
o
r
d
if
f
er
en
t
r
ea
s
o
n
s
,
lik
e
talk
in
g
to
f
r
ien
d
s
o
r
s
p
r
ea
d
in
g
n
e
ws
[
2
]
.
B
u
t
s
o
cial
m
ed
ia'
s
ea
s
e
o
f
u
s
e
an
d
th
e
an
o
n
y
m
ity
it
o
f
f
er
s
h
av
e
b
r
o
u
g
h
t
u
p
n
ew
ch
allen
g
es,
esp
ec
ially
wh
en
it
co
m
es
to
illeg
al
ac
tiv
ities
lik
e
s
ell
in
g
f
ak
e
o
r
illeg
al
d
r
u
g
s
[
3
]
.
Sellin
g
illeg
al
d
r
u
g
s
is
a
b
ig
p
r
o
b
lem
wo
r
ld
wid
e,
it’s
b
ad
f
o
r
p
u
b
lic
h
ea
lth
,
s
af
ety
,
an
d
th
e
ec
o
n
o
m
y
,
th
e
W
o
r
ld
Hea
lth
Or
g
an
izatio
n
(
W
HO)
[
4
]
s
ay
s
illeg
al
d
r
u
g
s
ar
e
th
o
s
e
s
o
ld
with
o
u
t
a
p
p
r
o
v
al
f
r
o
m
h
ea
lth
au
t
h
o
r
ities
.
T
h
is
in
clu
d
es
f
a
k
e
d
r
u
g
s
o
r
s
ellin
g
d
r
u
g
s
with
o
u
t
a
p
r
escr
ip
tio
n
th
at
s
h
o
u
ld
n
'
t b
e
s
o
ld
o
v
e
r
th
e
co
u
n
ter
[
5
]
.
X
h
as
b
ec
o
m
e
a
f
av
o
r
e
d
s
p
a
ce
f
o
r
d
r
u
g
v
en
d
o
r
s
to
p
o
s
t
illeg
al
ad
v
er
tis
em
en
ts
,
r
ea
ch
i
n
g
a
wid
e
au
d
ien
ce
at
lo
w
c
o
s
t
[
6
]
–
[
8
]
.
Pre
v
io
u
s
r
esear
ch
f
o
u
n
d
th
at
ap
p
r
o
x
im
ately
1
3
in
1
0
0
s
o
cial
m
ed
ia
p
o
s
ts
ad
v
er
tis
e
illi
cit
d
r
u
g
s
[
9
]
.
T
h
e
an
o
n
y
m
ity
an
d
b
r
o
ad
r
ea
c
h
o
f
s
o
cial
m
ed
ia
p
r
o
v
id
e
an
ef
f
icien
t
m
ea
n
s
f
o
r
illeg
al
s
eller
s
to
p
r
o
m
o
te
th
eir
p
r
o
d
u
cts.
I
n
d
o
n
esia h
as
s
p
ec
if
ic
r
eg
u
latio
n
s
g
o
v
er
n
in
g
th
e
s
ale
o
f
d
r
u
g
s
,
w
h
ich
ar
e
o
u
tlin
ed
in
s
ev
e
r
al
laws.
T
h
e
p
r
im
ar
y
leg
is
latio
n
in
clu
d
es
Hea
lth
L
aw
No
.
3
6
o
f
2
0
0
9
(
Un
d
a
n
g
-
Un
d
a
n
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
Dete
ctio
n
o
f illeg
a
l d
r
u
g
s
a
cti
vities in
I
n
d
o
n
esia
via
s
o
ci
a
l
med
ia
X
(
A
d
i H
a
n
if S
ed
a
r
)
3377
K
eseh
a
ta
n
N
o
.
3
6
Ta
h
u
n
2
0
0
9
)
an
d
th
e
Nar
co
tics
L
aw
No
.
3
5
o
f
2
0
0
9
(
Un
d
a
n
g
-
Un
d
a
n
g
N
a
r
ko
tika
N
o
.
3
5
Ta
h
u
n
2
0
0
9
)
.
T
h
ese
laws
r
est
r
ict
th
e
s
ale
o
f
ce
r
tain
m
ed
icatio
n
s
an
d
n
ar
co
tics
with
o
u
t
p
r
o
p
er
au
th
o
r
izatio
n
.
Fo
r
in
s
tan
ce
,
d
r
u
g
s
lik
e
ab
o
r
t
io
n
p
ills
,
s
ed
ativ
es,
an
d
an
est
h
etics
ar
e
r
eg
u
lated
u
n
d
e
r
th
e
s
e
laws
to
p
r
ev
en
t
m
is
u
s
e
an
d
illeg
al
d
is
tr
ib
u
tio
n
.
Fo
r
ex
am
p
le,
ab
o
r
tio
n
in
I
n
d
o
n
esia
is
h
ig
h
ly
r
estricte
d
,
y
et
co
m
m
o
n
,
with
a
n
esti
m
ated
4
2
.
5
ab
o
r
tio
n
s
p
er
1
,
0
0
0
wo
m
en
ag
e
d
1
5
-
4
9
in
J
av
a,
I
n
d
o
n
esia
[
1
0
]
.
Un
s
af
e
ab
o
r
tio
n
s
co
n
tr
i
b
u
te
s
ig
n
if
ican
tly
to
m
ater
n
al
m
o
r
t
ality
,
ac
co
u
n
tin
g
f
o
r
u
p
to
3
0
%
o
f
m
ater
n
al
d
ea
t
h
s
[
1
1
]
.
On
lin
e
m
ar
k
etp
lace
s
h
av
e
b
ec
o
m
e
a
s
o
u
r
ce
f
o
r
o
b
tain
in
g
m
is
o
p
r
o
s
to
l,
an
ab
o
r
tio
n
m
ed
icatio
n
;
h
o
wev
er
,
m
an
y
s
eller
s
p
r
o
v
id
e
in
co
m
p
lete
in
f
o
r
m
atio
n
a
b
o
u
t
s
id
e
ef
f
ec
ts
an
d
co
m
p
licatio
n
s
,
an
d
s
o
m
e
p
ac
k
ag
es
co
n
t
ain
an
in
s
u
f
f
icien
t
q
u
an
tity
o
f
th
e
d
r
u
g
f
o
r
a
co
m
p
lete
ab
o
r
tio
n
[
1
2
]
.
E
n
f
o
r
c
em
en
t
r
em
ai
n
s
ch
allen
g
i
n
g
,
e
s
p
ec
ially
o
n
d
ig
ital
p
latf
o
r
m
s
wh
er
e
s
eller
s
ca
n
ea
s
ily
co
n
ce
al
th
eir
id
en
titi
es.
B
ec
au
s
e
o
f
th
e
h
ar
m
f
u
l
ef
f
ec
t
s
o
f
illeg
al
d
r
u
g
s
s
ales,
it
'
s
im
p
o
r
tan
t
to
f
in
d
ef
f
icien
t
way
s
t
o
s
p
o
t
an
d
s
to
p
th
ese
ac
tiv
ities
o
n
s
o
cial
m
ed
ia
[
8
]
,
[
1
3
]
.
I
n
I
n
d
o
n
esi
a,
th
e
s
ale
o
f
illeg
al
d
r
u
g
s
th
r
o
u
g
h
s
o
cial
m
ed
ia
p
latf
o
r
m
s
h
as
b
ec
o
m
e
a
m
aj
o
r
co
n
ce
r
n
[
1
4
]
.
I
t
p
r
o
v
id
es
an
ea
s
y
an
d
e
f
f
icien
t
m
ea
n
s
f
o
r
illeg
al
s
eller
s
to
p
r
o
m
o
te
t
h
eir
p
r
o
d
u
cts.
T
h
e
f
r
ee
d
o
m
o
f
co
m
m
u
n
icatio
n
a
n
d
th
e
ab
ilit
y
to
c
o
n
ce
al
id
e
n
titi
es
m
ak
e
th
ese
s
eller
s
f
ee
l
s
ec
u
r
e
o
p
e
r
atin
g
o
n
th
e
p
latf
o
r
m
[
1
5
]
.
I
n
d
o
n
e
s
ia
'
s
lar
g
e
p
o
p
u
latio
n
o
f
o
v
er
1
7
5
m
illi
o
n
ac
tiv
e
in
ter
n
et
u
s
er
s
(
2
0
2
1
–
2
0
2
4
)
,
co
m
b
in
ed
with
th
e
g
r
o
win
g
ad
o
p
tio
n
o
f
s
m
ar
tp
h
o
n
es,
h
a
s
cr
ea
ted
a
s
tr
o
n
g
f
o
u
n
d
atio
n
f
o
r
t
h
e
r
ap
i
d
g
r
o
wth
o
f
s
o
cial
m
e
d
ia
an
d
o
n
li
n
e
co
m
m
e
r
ce
[
1
6
]
.
Mo
r
eo
v
er
,
I
n
d
o
n
esia,
b
ein
g
a
co
u
n
tr
y
with
h
ig
h
s
o
cial
m
e
d
ia
u
s
ag
e,
s
er
v
es
as
a
p
o
ten
t
ial
m
ar
k
et
f
o
r
illeg
al
d
r
u
g
s
e
ller
s
to
r
ea
ch
m
o
r
e
co
n
s
u
m
er
s
.
W
ith
th
e
lack
o
f
ef
f
ec
tiv
e
s
u
p
er
v
is
io
n
an
d
law
en
f
o
r
ce
m
en
t,
th
e
ac
tiv
ity
o
f
s
ellin
g
illeg
al
d
r
u
g
s
ca
n
lead
to
s
er
io
u
s
h
ea
lth
co
n
s
eq
u
en
ce
s
f
o
r
th
e
p
u
b
lic.
Ad
v
an
ce
d
tech
n
o
lo
g
y
lik
e
d
a
ta
m
in
in
g
,
m
ac
h
in
e
lear
n
in
g
,
an
d
n
atu
r
al
la
n
g
u
a
g
e
p
r
o
ce
s
s
in
g
(
NL
P)
ca
n
b
e
u
s
ed
to
i
d
en
tify
a
n
d
a
n
aly
ze
co
m
m
u
n
icatio
n
p
atter
n
s
r
elate
d
to
illeg
al
ac
tiv
ity
s
ales
o
n
X
[
1
3
]
.
B
y
au
to
m
atin
g
tex
t
a
n
aly
s
is
an
d
u
s
in
g
tex
t
class
if
icatio
n
tech
n
iq
u
es,
r
esear
ch
e
r
s
an
d
law
en
f
o
r
ce
m
e
n
t
ca
n
q
u
ick
ly
id
e
n
tify
an
d
r
esp
o
n
d
t
o
illeg
al
ac
tiv
ities
o
n
s
o
cial
m
ed
ia
[
1
7
]
.
W
h
ile
th
ese
m
eth
o
d
o
lo
g
ies h
av
e
s
h
o
w
n
p
r
o
m
is
e,
th
er
e
is
a
n
o
tab
le
g
ap
in
th
e
co
n
s
id
er
atio
n
o
f
p
e
r
s
o
n
ally
id
en
tifia
b
le
in
f
o
r
m
atio
n
(
PII
)
d
etec
tio
n
with
in
th
e
f
ea
tu
r
e
s
ets
u
s
ed
f
o
r
class
if
icatio
n
.
PII
,
s
u
ch
as
p
h
o
n
e
n
u
m
b
e
r
s
,
em
ail
ad
d
r
ess
es,
an
d
web
UR
L
s
,
o
f
ten
s
er
v
e
as
d
ir
ec
t
in
d
icato
r
s
o
f
illi
cit
in
ten
t
[
5
]
,
esp
ec
iall
y
in
th
e
co
n
tex
t
o
f
illeg
al
d
r
u
g
s
ales
,
wh
er
e
s
e
ller
s
in
clu
d
e
co
n
tact
in
f
o
r
m
atio
n
f
o
r
p
o
ten
tial
b
u
y
er
s
.
Pre
v
io
u
s
s
tu
d
ies
[
6
]
,
[
1
3
]
,
[
1
7
]
h
av
e
p
r
im
ar
ily
f
o
c
u
s
ed
o
n
tex
tu
al
an
d
lin
g
u
is
tic
f
ea
tu
r
e
s
with
o
u
t
ex
ten
s
iv
ely
in
co
r
p
o
r
atin
g
PII
an
d
web
UR
L
s
d
etec
tio
n
in
to
th
eir
m
o
d
els.
T
h
is
m
ay
lim
it th
e
e
f
f
ec
tiv
en
ess
o
f
class
if
icatio
n
m
o
d
els in
ac
cu
r
ately
id
e
n
tify
in
g
illi
cit
ac
tiv
ities
.
T
h
is
s
tu
d
y
ad
d
r
ess
es
th
is
g
ap
th
r
o
u
g
h
two
r
esear
c
h
q
u
esti
o
n
s
:
i)
h
o
w
class
if
icatio
n
tech
n
iq
u
es
an
d
tex
t
m
in
in
g
ca
n
b
e
ap
p
lied
to
id
en
tify
p
o
s
ts
r
elate
d
to
th
e
s
ale
o
f
illeg
al
d
r
u
g
s
o
n
X,
an
d
ii)
wh
ich
ty
p
es
o
f
d
r
u
g
s
ar
e
m
o
s
t
f
r
e
q
u
en
tly
p
r
o
m
o
ted
o
n
X
in
I
n
d
o
n
esia.
T
o
an
s
wer
th
ese
q
u
esti
o
n
s
,
th
e
m
ain
co
n
tr
ib
u
tio
n
s
o
f
th
is
wo
r
k
ar
e
th
r
e
ef
o
ld
:
i)
s
y
s
tem
atica
lly
in
co
r
p
o
r
ate
PII
f
e
atu
r
es
lik
e
p
h
o
n
e
n
u
m
b
er
s
,
e
m
ail
ad
d
r
ess
es,
an
d
web
UR
L
s
in
to
illi
ci
t d
r
u
g
d
et
ec
tio
n
o
n
I
n
d
o
n
esian
X,
d
em
o
n
s
tr
atin
g
th
r
o
u
g
h
f
ea
tu
r
e
im
p
o
r
tan
ce
an
aly
s
is
th
at
th
ese
f
ea
tu
r
es
r
a
n
k
am
o
n
g
th
e
to
p
f
iv
e
m
o
s
t
in
f
l
u
en
tial
s
ig
n
als;
ii)
co
n
s
tr
u
ct
an
d
an
aly
z
e
a
d
o
m
ain
-
s
p
ec
if
i
c
d
ataset
I
n
d
o
n
esian
p
o
s
ts
co
v
er
in
g
ab
o
r
tio
n
p
ills
,
s
ex
u
al
en
h
an
ce
m
e
n
t,
an
d
an
esth
etic
d
r
u
g
s
,
co
n
tr
ib
u
tin
g
em
p
ir
ical
ev
id
en
ce
o
f
o
n
lin
e
illi
cit
p
h
ar
m
ac
eu
tical
ac
tiv
ity
in
th
is
r
eg
io
n
al
c
o
n
tex
t;
an
d
iii)
o
f
f
er
p
r
ac
tical
in
s
ig
h
ts
in
to
d
ep
lo
y
m
en
t
co
n
s
id
er
atio
n
s
,
in
clu
d
i
n
g
er
r
o
r
a
n
aly
s
is
an
d
et
h
ical
im
p
licatio
n
s
r
elev
an
t
to
law
en
f
o
r
ce
m
e
n
t
an
d
p
u
b
lic
h
ea
lt
h
ag
en
cies.
T
h
e
f
in
d
in
g
s
ca
n
ass
is
t
h
ea
lth
d
ep
ar
tm
e
n
ts
an
d
law
en
f
o
r
ce
m
e
n
t
in
id
en
tify
in
g
a
n
d
ad
d
r
ess
in
g
illi
cit
d
r
u
g
s
ales
o
n
s
o
cial
m
ed
i
a,
wh
ile
in
f
o
r
m
in
g
th
e
d
e
v
el
o
p
m
en
t
o
f
s
tr
o
n
g
e
r
r
eg
u
lato
r
y
an
d
m
o
n
ito
r
in
g
f
r
a
m
ewo
r
k
s
.
2.
M
E
T
H
O
D
I
n
th
is
s
tu
d
y
,
d
ata
wer
e
co
llected
u
s
in
g
web
s
cr
ap
in
g
,
ta
r
g
etin
g
X
'
s
s
ea
r
ch
r
esu
lts
to
f
in
d
p
o
s
ts
r
elate
d
to
th
e
k
ey
wo
r
d
s
illeg
al
d
r
u
g
s
.
T
h
e
f
ea
t
u
r
es
ex
tr
ac
t
ed
in
clu
d
e
u
s
er
ID
,
p
o
s
ts
tex
t
f
o
r
class
if
icatio
n
p
u
r
p
o
s
es,
an
d
th
e
p
o
s
ts
d
ate,
wh
ich
wer
e
th
e
n
ex
t
r
ac
ted
u
s
i
n
g
Py
th
o
n
.
Af
te
r
o
b
tain
in
g
th
e
co
llected
d
ata,
tex
t
p
r
e
-
p
r
o
ce
s
s
in
g
(
f
ilter
in
g
,
ca
s
e
f
o
ld
in
g
,
to
k
en
izatio
n
,
s
to
p
w
o
r
d
r
e
m
o
v
al,
an
d
d
ata
s
tem
m
i
n
g
)
was
p
e
r
f
o
r
m
ed
to
clea
n
it.
T
h
e
d
ata
was
th
en
m
an
u
ally
p
r
o
ce
s
s
ed
an
d
class
if
ied
.
Nex
t,
th
e
d
ata
was
d
iv
id
ed
in
to
test
in
g
an
d
tr
ain
in
g
s
ets.
C
o
n
s
tr
u
ct
a
m
ac
h
in
e
lear
n
in
g
m
o
d
el
,
th
en
co
n
d
u
cted
u
s
in
g
b
y
co
m
p
ar
in
g
th
e
m
o
d
els.
T
h
r
o
u
g
h
th
ese
m
eth
o
d
s
,
tex
t
class
if
icat
io
n
was
as
s
ess
ed
to
s
ee
wh
ich
p
o
s
ts
lead
to
th
e
s
ale
o
f
ill
icit
d
r
u
g
s
.
Fin
ally
,
th
e
d
ata
was
cr
ea
ted
a
n
d
v
is
u
alize
d
u
s
in
g
th
e
Py
th
o
n
lib
r
ar
y
f
o
r
cr
ea
tin
g
g
r
ap
h
s
m
at
p
lo
tlib
.
T
h
e
o
v
er
all
f
l
o
w
p
r
o
ce
s
s
o
f
th
is
s
tu
d
y
is
illu
s
tr
ated
in
Fig
u
r
e
1
.
2
.
1
.
Da
t
a
c
o
llect
io
n
T
h
e
d
a
t
a
s
e
t
u
s
ed
in
t
h
i
s
r
e
s
e
a
r
ch
wa
s
f
r
o
m
X
,
co
l
le
c
t
e
d
u
s
i
n
g
we
b
s
c
r
ap
i
n
g
,
t
a
r
g
e
t
i
n
g
X
's
s
e
a
r
c
h
r
e
s
u
l
t
s
.
W
e
b
s
c
r
a
p
i
n
g
c
a
n
b
e
d
e
s
ig
n
ed
t
o
co
l
l
e
c
t
c
o
n
v
e
r
s
a
t
i
o
n
a
l
d
a
t
a
f
r
o
m
s
o
c
i
a
l
m
e
d
i
a
p
l
a
t
f
o
r
m
s
i
n
J
av
a
S
c
r
ip
t
o
b
j
ec
t
n
o
t
a
t
io
n
(
J
S
O
N
)
f
o
r
m
a
t,
w
h
i
c
h
c
an
t
h
en
b
e
f
i
l
t
e
r
e
d
a
n
d
f
o
r
m
a
t
t
ed
f
o
r
s
p
e
c
i
f
i
c
u
s
e
ca
s
e
s
[
1
8
]
.
Us
i
n
g
P
y
th
o
n
s
c
r
ip
t
s
w
i
th
s
c
r
a
p
in
g
l
i
b
r
a
r
i
e
s
,
w
e
t
ar
g
e
te
d
d
a
t
a
in
v
o
l
v
in
g
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.
15
,
No
.
4
,
Au
g
u
s
t 2
0
2
6
:
3
3
7
6
-
3
3
8
8
3378
s
p
e
c
i
f
i
c
k
ey
w
o
r
d
s
in
I
n
d
o
n
e
s
i
a
n
,
s
u
c
h
a
s
"
j
u
a
l
o
b
a
t
"
(
s
e
l
l
d
r
u
g
s
)
,
"
o
b
a
t
a
b
o
r
s
i
"
(
ab
o
r
t
i
o
n
p
i
l
l
s
)
,
"
p
e
n
g
g
u
g
u
r
ka
n
d
u
n
g
a
n
"
(
p
r
e
g
n
an
cy
t
er
m
i
n
a
t
i
o
n
)
,
"
o
b
a
t
t
a
n
p
a
r
e
s
e
p
"
(
d
r
u
g
s
w
i
th
o
u
t
p
r
e
s
cr
i
p
t
io
n
)
,
"
p
i
l
p
en
e
n
a
n
g
"
(
s
e
d
a
t
i
v
e
p
i
lls
)
,
"
o
b
a
t
b
iu
s
"
(
a
n
e
s
th
e
t
i
c
d
r
u
g
s
)
,
an
d
a
l
s
o
s
l
an
g
o
r
c
o
d
e
w
o
r
d
s
u
s
ed
in
t
h
e
n
a
r
co
t
i
c
s
wo
r
ld
to
ev
ad
e
d
e
te
c
t
i
o
n
,
in
c
l
u
d
i
n
g
"
p
u
t
a
w
"
f
o
r
h
e
r
o
i
n
.
T
h
i
s
ap
p
r
o
a
ch
,
d
e
s
i
g
n
ed
to
co
v
e
r
t
h
e
b
r
o
a
d
s
p
e
c
t
r
u
m
o
f
i
l
l
eg
a
l
d
r
u
g
s
an
d
d
r
u
g
s
a
l
e
s
o
n
l
i
n
e
,
e
n
s
u
r
e
s
a
th
o
r
o
u
g
h
d
a
ta
co
ll
e
c
t
i
o
n
p
r
o
c
e
s
s
b
y
i
n
c
l
u
d
i
n
g
a
d
d
i
t
i
o
n
a
l
t
e
r
m
s
u
s
e
d
b
y
d
e
a
l
er
s
an
d
u
s
e
r
s
to
k
e
e
p
t
h
e
ir
a
c
t
i
v
i
t
ie
s
u
n
d
e
r
g
r
o
u
n
d
.
T
h
i
s
s
c
r
ap
i
n
g
p
r
o
c
e
s
s
w
a
s
c
o
n
d
u
ct
e
d
in
M
a
y
2
0
2
4
.
Du
r
i
n
g
th
e
d
a
ta
c
o
l
l
e
c
t
io
n
s
t
ag
e
s
,
s
e
v
e
r
a
l
c
h
a
l
l
en
g
e
s
w
er
e
e
n
c
o
u
n
te
r
ed
,
s
u
c
h
a
s
m
a
n
y
p
o
s
t
s
c
o
n
t
a
i
n
in
g
r
ep
e
t
i
t
iv
e
o
r
d
u
p
li
c
a
t
ed
,
r
e
q
u
i
r
e
d
a
d
d
i
t
i
o
n
a
l
s
t
e
p
s
to
r
e
m
o
v
e
t
h
e
s
e
t
o
av
o
id
s
k
e
w
in
g
th
e
r
es
u
l
t
s
.
B
e
f
o
r
e
f
i
l
t
er
i
n
g
d
u
p
l
i
ca
t
e
d
r
e
s
u
l
t
s
,
t
h
e
d
a
t
a
s
e
t
f
o
r
t
h
is
s
t
u
d
y
co
n
s
i
s
t
e
d
o
f
1
3
,
1
9
6
p
o
s
t
s
co
l
l
e
c
t
ed
o
v
er
a
o
n
e
-
y
e
a
r
p
er
i
o
d
(
Ap
r
i
l
2
0
2
3
–
2
0
2
4
)
.
O
u
t
o
f
t
h
e
i
n
i
t
ia
l
1
3
,
1
9
6
p
o
s
t
s
c
o
l
le
c
t
e
d
,
1
,
4
1
9
w
er
e
id
e
n
t
if
i
e
d
a
s
d
u
p
l
i
c
a
t
e
s
o
r
ir
r
e
l
ev
an
t
a
n
d
r
em
o
v
ed
d
u
r
in
g
p
r
e
-
p
r
o
c
e
s
s
i
n
g
,
r
e
s
u
l
t
i
n
g
i
n
a
f
i
n
a
l
d
a
t
a
s
e
t
o
f
1
1
,
7
7
7
p
o
s
t
s
.
T
h
e
d
a
t
a
c
o
l
le
c
t
e
d
i
s
t
h
e
n
s
to
r
ed
i
n
t
h
e
J
S
O
N
.
Af
ter
th
e
d
ata
was
co
llected
th
e
d
ata
was
lab
eled
m
an
u
ally
.
T
o
en
s
u
r
e
th
e
q
u
ality
o
f
m
an
u
al
class
if
icatio
n
,
th
e
lab
elin
g
p
r
o
ce
s
s
was
co
n
d
u
cted
b
y
two
an
n
o
tato
r
s
f
o
llo
win
g
s
p
ec
if
i
c
g
u
id
elin
es.
T
h
ese
g
u
id
elin
es
in
clu
d
e
d
cr
iter
ia
s
u
ch
as
m
en
tio
n
i
n
g
th
e
b
r
an
d
n
am
e
o
r
t
h
e
ty
p
e
o
f
d
r
u
g
(
e.
g
.
,
an
esth
etic)
o
r
its
ef
f
ec
ts
,
en
s
u
r
in
g
c
o
n
s
is
ten
t
a
n
d
ac
c
u
r
ate
ca
te
g
o
r
izatio
n
o
f
th
e
p
o
s
ts
.
T
h
is
c
o
llab
o
r
ativ
e
ap
p
r
o
ac
h
aim
ed
to
m
in
im
ize
b
ias
an
d
en
h
a
n
ce
t
h
e
r
eliab
ilit
y
o
f
th
e
lab
elin
g
p
r
o
ce
s
s
.
T
o
m
ea
s
u
r
e
in
ter
-
an
n
o
tato
r
a
g
r
ee
m
en
t,
C
o
h
en
'
s
Kap
p
a
(
)
was
co
m
p
u
ted
o
n
th
e
1
1
,
7
7
7
lab
eled
p
o
s
ts
,
y
ield
in
g
=0
.
8
7
,
wh
ich
in
d
icate
s
alm
o
s
t
p
er
f
ec
t a
g
r
ee
m
en
t
[
1
9
]
.
Dis
ag
r
ee
m
en
ts
wer
e
r
eso
lv
ed
th
r
o
u
g
h
d
is
cu
s
s
io
n
b
ef
o
r
e
t
r
ain
in
g
.
T
h
e
lab
elin
g
cr
iter
ia
f
o
r
in
ten
t c
lass
if
icatio
n
ar
e
d
escr
ib
ed
in
T
ab
le
1
.
T
ab
le
2
p
r
e
s
en
ts
ex
am
p
les o
f
p
o
s
ts
cla
s
s
if
ied
as in
ten
t to
s
el
l
an
d
n
o
t
in
ten
t to
s
ell.
E
x
am
p
le
s
o
f
p
o
s
ts
ca
teg
o
r
ized
b
y
d
r
u
g
ty
p
e
ar
e
s
h
o
wn
in
T
ab
le
3
.
Fig
u
r
e
1
.
Flo
w
p
r
o
ce
s
s
o
f
th
e
p
r
o
p
o
s
ed
r
esear
ch
T
ab
le
1
.
I
n
ten
t c
lass
if
icatio
n
La
b
e
l
D
e
scri
p
t
i
o
n
Y
e
s
Te
x
t
s
t
h
a
t
i
n
d
i
c
a
t
e
t
h
e
u
s
e
r
i
s
o
f
f
e
r
i
n
g
t
o
se
l
l
i
l
l
e
g
a
l
d
r
u
g
s.
T
h
i
s
c
o
u
l
d
i
n
c
l
u
d
e
s
p
e
c
i
f
i
c
o
f
f
e
r
s,
p
r
i
c
i
n
g
i
n
f
o
r
ma
t
i
o
n
,
d
e
t
a
i
l
s
a
b
o
u
t
d
r
u
g
t
y
p
e
s
o
r
q
u
a
n
t
i
t
i
e
s.
No
Te
x
t
s
t
h
a
t
me
n
t
i
o
n
i
l
l
e
g
a
l
d
r
u
g
s
b
u
t
d
o
n
o
t
d
e
m
o
n
st
r
a
t
e
a
c
l
e
a
r
i
n
t
e
n
t
t
o
s
e
l
l
.
T
h
i
s
c
o
u
l
d
i
n
c
l
u
d
e
d
i
sc
u
ss
i
o
n
s
a
b
o
u
t
n
e
w
s,
i
l
l
e
g
a
l
d
r
u
g
e
f
f
e
c
t
s,
l
e
g
a
l
i
ss
u
e
s
,
o
r
p
e
r
so
n
a
l
e
x
p
e
r
i
e
n
c
e
s
t
h
a
t
a
r
e
n
o
t
r
e
l
a
t
e
d
t
o
t
r
a
n
sa
c
t
i
o
n
s
.
T
ab
le
2
.
E
x
am
p
le
o
f
in
ten
t c
la
s
s
if
icatio
n
p
o
s
ts
f
r
o
m
X
La
b
e
l
P
o
st
s
f
r
o
m X
Y
e
s
“
J
u
a
l
o
b
a
t
a
b
o
rsi
p
e
n
g
g
u
g
u
r
k
a
n
d
u
n
g
a
n
m
i
s
o
p
r
o
s
t
o
l
G
a
st
r
u
l
a
m
p
u
h
u
n
t
u
k
t
u
n
t
a
sk
a
n
m
a
s
a
l
a
h
t
e
r
l
a
m
b
a
t
h
a
i
d
u
s
i
a
1
2
3
4
5
6
b
u
l
a
n
#
o
b
a
t
a
b
o
r
si
”
“
S
e
l
l
i
n
g
a
b
o
r
t
i
o
n
d
r
u
g
s
,
p
r
e
g
n
a
n
c
y
t
e
r
mi
n
a
t
o
r
s,
m
i
s
o
p
r
o
st
o
l
G
a
st
r
u
l
,
e
f
f
e
c
t
i
v
e
f
o
r
r
e
s
o
l
v
i
n
g
l
a
t
e
p
e
r
i
o
d
i
ss
u
e
s
f
o
r
1
t
o
6
mo
n
t
h
s
#
a
b
o
r
t
i
o
n
p
i
l
l
”
No
“
K
a
y
a
n
y
a
k
o
n
o
t
a
s
i
N
a
rk
o
b
a
i
t
u
n
e
g
a
t
i
f
.
T
a
p
i
m
e
m
a
n
g
a
d
a
o
b
a
t
f
l
u
y
g
d
i
d
a
l
a
m
n
y
a
a
d
a
g
o
l
o
n
g
a
n
n
a
rk
o
t
i
k
a
d
a
n
i
t
u
h
a
r
u
s
sesu
a
i
res
e
p
d
o
k
t
e
r
.
K
a
y
a
n
y
a
y
g
a
d
a
k
a
n
d
u
n
g
a
n
z
a
t
n
a
rk
o
t
i
k
a
g
a
k
se
g
a
m
p
a
n
g
i
t
u
d
i
e
d
a
r
k
a
n
d
i
p
a
sa
r t
a
n
p
a
r
e
se
p
.
C
m
i
i
w
”
“
I
t
see
ms
l
i
k
e
t
h
e
c
o
n
n
o
t
a
t
i
o
n
o
f
d
r
u
g
s
i
s
n
e
g
a
t
i
v
e
.
B
u
t
t
h
e
r
e
a
r
e
i
n
d
e
e
d
f
l
u
me
d
i
c
i
n
e
s
t
h
a
t
c
o
n
t
a
i
n
n
a
r
c
o
t
i
c
s
a
n
d
m
u
st
b
e
p
r
e
scr
i
b
e
d
b
y
a
d
o
c
t
o
r
.
I
t
se
e
ms
t
h
a
t
su
b
s
t
a
n
c
e
s
c
o
n
t
a
i
n
i
n
g
n
a
r
c
o
t
i
c
s
c
a
n
n
o
t
b
e
e
a
si
l
y
d
i
st
r
i
b
u
t
e
d
i
n
t
h
e
m
a
r
k
e
t
w
i
t
h
o
u
t
a
p
r
e
scr
i
p
t
i
o
n
.
C
o
r
r
e
c
t
me
i
f
I
'm w
r
o
n
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
Dete
ctio
n
o
f illeg
a
l d
r
u
g
s
a
cti
vities in
I
n
d
o
n
esia
via
s
o
ci
a
l
med
ia
X
(
A
d
i H
a
n
if S
ed
a
r
)
3379
T
ab
le
3
.
E
x
am
p
le
o
f
p
o
s
ts
with
ty
p
e
o
f
d
r
u
g
class
if
icatio
n
La
b
e
l
P
o
st
f
r
o
m X
A
b
o
r
t
i
o
n
p
i
l
l
s
“
J
u
a
l
o
b
a
t
c
y
t
o
t
e
c
a
b
o
rsi
p
e
n
g
g
u
g
u
r
k
a
n
d
u
n
g
a
n
a
sl
i
WA
”
“
S
e
l
l
i
n
g
o
r
i
g
i
n
a
l
a
b
o
r
t
i
o
n
d
r
u
g
c
y
t
o
t
e
c
W
h
a
t
sA
p
p
”
A
n
e
st
h
e
t
i
c
d
r
u
g
s
“
J
u
a
l
o
b
a
t
b
i
u
s
c
a
i
r
l
i
q
u
i
d
s
e
x
a
m
p
u
h
l
a
n
g
s
u
n
g
p
i
n
g
sa
n
g
W
A”
“
S
e
l
l
i
n
g
l
i
q
u
i
d
a
n
e
s
t
h
e
t
i
c
d
r
u
g
se
x
l
i
q
u
i
d
e
f
f
e
c
t
i
v
e
,
W
h
a
t
sA
p
p
”
S
e
x
u
a
l
e
n
h
a
n
c
e
me
n
t
d
r
u
g
s
“
J
u
a
l
o
b
a
t
v
i
a
g
ra
a
s
l
i
”
“
S
e
l
l
i
n
g
o
r
i
g
i
n
a
l
v
i
a
g
r
a
d
r
u
g
”
2
.
2
.
Da
t
a
pro
ce
s
s
ing
X
d
ata
is
h
ar
d
to
wo
r
k
with
b
ec
au
s
e
th
e
way
p
eo
p
le
wr
it
e
is
in
co
n
s
is
ten
t,
an
d
t
h
er
e
a
r
en
'
t
clea
r
ca
teg
o
r
ies,
m
ak
in
g
an
al
y
s
is
a
n
d
class
if
icatio
n
ch
allen
g
in
g
[
2
0
]
.
T
h
e
g
o
al
o
f
d
ata
p
r
e
-
p
r
o
c
ess
in
g
is
to
c
lean
s
e
an
d
s
tan
d
a
r
d
ize
t
h
e
d
ata
to
ai
d
s
u
b
s
eq
u
e
n
t
p
r
o
ce
s
s
es.
Var
io
u
s
d
ata
clea
n
in
g
tech
n
iq
u
es
e
n
s
u
r
e
th
at
ex
tr
ac
ted
in
f
o
r
m
atio
n
is
ac
cu
r
ate
an
d
d
ev
o
id
o
f
an
y
u
n
n
ec
ess
ar
y
n
o
i
s
e
o
r
m
is
s
in
g
in
f
o
r
m
atio
n
d
u
e
to
s
cr
ap
in
g
.
I
n
th
is
s
tu
d
y
,
th
e
au
th
o
r
u
s
es
s
ev
er
al
s
tep
s
f
o
r
d
ata
p
r
e
-
p
r
o
ce
s
s
in
g
.
T
h
e
in
itial
p
h
ase
in
v
o
lv
es
s
ele
ctin
g
an
d
c
u
r
atin
g
th
e
co
llected
X
d
ata
b
y
f
ilter
i
n
g
o
u
t
d
u
p
licated
p
o
s
ts
an
d
u
n
wan
ted
p
h
r
ases
n
o
t
r
elate
d
to
d
r
u
g
s
o
r
p
o
s
ts
f
r
o
m
ad
v
er
tis
er
s
,
r
esu
ltin
g
in
o
n
ly
1
1
,
7
7
7
p
o
s
ts
b
ein
g
u
s
ed
.
On
c
e
th
e
d
ata
is
f
ilter
ed
,
all
th
e
le
tter
s
in
th
e
p
o
s
ts
ar
e
tr
an
s
f
o
r
m
ed
i
n
to
lo
wer
ca
s
e
t
o
m
ak
e
th
e
d
ata
u
n
if
o
r
m
a
n
d
ea
s
ier
to
p
r
o
ce
s
s
in
th
e
s
u
b
s
eq
u
en
t
s
tep
s
.
T
h
e
to
k
en
izatio
n
p
r
o
ce
s
s
b
r
ea
k
s
d
o
wn
ea
ch
p
o
s
t
in
to
in
d
i
v
id
u
al
wo
r
d
s
,
o
r
“
to
k
e
n
s
”
,
wh
ile
s
im
u
ltan
eo
u
s
ly
r
em
o
v
in
g
p
u
n
ctu
atio
n
,
u
s
in
g
t
h
e
n
atu
r
al
lan
g
u
ag
e
to
o
lk
it
(
NL
T
K
)
'
s
wo
r
d
to
k
en
ize
f
u
n
ct
io
n
.
Sp
ec
if
ic
wo
r
d
s
th
at
d
o
n
o
t
co
n
tr
ib
u
te
t
o
th
e
o
v
er
all
m
ea
n
in
g
o
f
th
e
p
o
s
ts
,
k
n
o
wn
as
s
to
p
wo
r
d
s
,
ar
e
th
en
r
em
o
v
ed
.
Ad
d
itio
n
ally
,
UR
L
s
,
h
ash
tag
s
,
an
d
m
e
n
tio
n
s
ar
e
also
elim
in
a
ted
u
s
in
g
a
s
et
o
f
I
n
d
o
n
esian
a
n
d
X
-
s
p
ec
if
ic
s
to
p
wo
r
d
s
.
T
h
e
f
in
al
s
tep
in
th
e
p
r
e
-
p
r
o
ce
s
s
in
g
p
ip
elin
e
tr
a
n
s
f
o
r
m
s
all
wo
r
d
s
in
to
th
eir
r
o
o
t
f
o
r
m
s
o
r
s
tem
s
u
s
in
g
th
e
Sas
tr
awi
s
tem
m
er
f
o
r
th
e
I
n
d
o
n
esian
lan
g
u
ag
e
,
s
im
p
li
f
y
in
g
t
h
e
d
ata
an
d
r
ed
u
cin
g
its
co
m
p
lex
ity
f
o
r
s
u
b
s
eq
u
en
t
p
r
o
ce
s
s
in
g
.
E
ac
h
p
r
o
ce
s
s
ed
p
o
s
ts
ar
e
th
en
s
to
r
e
d
in
a
lis
t,
with
its
r
o
o
t
wo
r
d
s
an
d
a
v
er
s
io
n
o
f
th
e
p
o
s
ts
with
th
e
p
r
o
ce
s
s
ed
co
n
t
en
t.
T
h
e
p
r
o
ce
s
s
ed
d
ata
is
f
in
ally
s
av
ed
in
a
J
SON
f
ile
f
o
r
f
u
r
th
er
u
s
e.
T
a
b
le
4
d
is
p
lay
s
th
e
tr
an
s
f
o
r
m
atio
n
o
f
th
e
d
ata
b
ef
o
r
e
an
d
af
ter
t
h
e
p
r
e
-
p
r
o
ce
s
s
in
g
p
h
ase
f
o
r
t
h
e
co
ll
ec
ted
p
o
s
ts
.
T
ab
le
4
.
B
ef
o
r
e
an
d
a
f
ter
p
r
e
-
p
r
o
ce
s
s
in
g
La
b
e
l
P
o
st
s
f
r
o
m X
B
e
f
o
r
e
“
J
u
a
l
o
b
a
t
c
y
t
o
t
e
c
a
b
o
rsi
p
e
n
g
g
u
g
u
r
k
a
n
d
u
n
g
a
n
a
sl
i
WA
”
“
S
e
l
l
i
n
g
o
r
i
g
i
n
a
l
c
y
t
o
t
e
c
a
b
o
r
t
i
o
n
d
r
u
g
p
r
e
g
n
a
n
c
y
t
e
r
mi
n
a
t
o
r
W
A
”
A
f
t
e
r
j
u
a
l
o
b
a
t
c
y
t
o
t
e
c
a
b
o
rs
i
g
u
g
u
r k
a
n
d
u
n
g
a
sl
i
“
S
e
l
l
i
n
g
o
r
i
g
i
n
a
l
c
y
t
o
t
e
c
a
b
o
r
t
i
o
n
d
r
u
g
p
r
e
g
n
a
n
c
y
t
e
r
mi
n
a
t
o
r
”
2
.
3
.
F
e
a
t
ure
ex
t
r
a
ct
io
n
T
h
e
ter
m
f
r
eq
u
en
cy
-
i
n
v
er
s
e
d
o
cu
m
en
t
f
r
e
q
u
en
c
y
(
TF
-
I
DF
)
alg
o
r
ith
m
in
v
o
lv
es
two
ca
lcu
latio
n
s
:
T
F
an
d
I
DF
.
T
h
e
T
F
-
I
DF
s
co
r
e
h
ig
h
lig
h
ts
wh
ich
f
ea
tu
r
e
h
o
ld
s
th
e
g
r
ea
test
s
ig
n
if
ican
c
e
ac
r
o
s
s
th
e
e
n
tire
d
o
cu
m
e
n
t
co
llectio
n
[
2
1
]
.
Af
ter
th
e
weig
h
ts
ar
e
ass
ig
n
ed
,
t
h
e
lev
el
o
f
s
im
ilar
ity
is
d
eter
m
in
ed
b
y
a
r
r
an
g
i
n
g
th
e
wo
r
d
s
in
to
a
v
ec
t
o
r
a
n
d
u
tili
zin
g
co
s
in
e
s
im
ilar
ity
[
2
2
]
.
T
h
e
p
r
o
ce
s
s
ed
J
SON
f
ile
is
lo
ad
ed
,
an
d
t
h
e
co
n
ten
t
o
f
ea
ch
p
o
s
t
is
c
o
llected
in
to
s
ep
ar
ate
lis
ts
.
T
h
e
C
o
u
n
tVec
to
r
izer
an
d
T
f
id
f
Vec
t
o
r
izer
f
r
o
m
s
k
lear
n
ar
e
u
s
ed
to
b
u
ild
a
v
o
ca
b
u
lar
y
an
d
en
c
o
d
e
th
e
d
o
cu
m
en
ts
in
to
a
b
ag
o
f
wo
r
d
s
(
B
o
W
)
an
d
T
F
-
I
DF m
atr
ices.
2
.
4
.
Cla
s
s
if
ica
t
io
n a
nd
ev
a
lu
a
t
io
n
On
ce
th
e
d
ata
was
p
r
o
ce
s
s
ed
a
n
d
v
er
if
ied
,
r
esear
c
h
er
s
p
r
o
ce
ed
ed
to
b
u
ild
a
m
ac
h
in
e
lear
n
i
n
g
m
o
d
el.
T
o
p
er
f
o
r
m
f
u
t
u
r
e
class
if
icati
o
n
o
n
th
e
d
ata
,
f
iv
e
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
wer
e
test
ed
:
n
aïv
e
B
ay
es,
lo
g
is
tic
r
eg
r
ess
io
n
,
r
an
d
o
m
f
o
r
est
,
g
r
ad
ien
t
b
o
o
s
tin
g
,
an
d
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
(
SVM)
.
T
h
e
s
elec
tio
n
o
f
th
ese
m
o
d
els
was
in
f
o
r
m
ed
b
y
th
eir
s
tr
o
n
g
p
er
f
o
r
m
a
n
ce
in
s
im
ilar
s
tu
d
ies.
Pre
v
io
u
s
r
esear
ch
h
as
s
h
o
wn
th
at
SVM
is
p
ar
ticu
lar
ly
ef
f
ec
tiv
e
in
tex
t
class
if
icatio
n
task
s
d
u
e
to
its
ab
ilit
y
to
h
an
d
le
h
ig
h
-
d
im
en
s
io
n
al
f
ea
tu
r
e
s
p
ac
es
an
d
p
er
f
o
r
m
well
ev
e
n
with
s
m
all
d
ataset
s
[
6
]
,
[
7
]
.
Ad
d
itio
n
ally
,
o
th
e
r
alg
o
r
it
h
m
s
wer
e
u
s
ed
as
b
aselin
e
m
o
d
els
to
ev
alu
ate
th
e
r
elativ
e
p
er
f
o
r
m
an
ce
o
f
class
if
ier
s
.
W
h
ile
tr
an
s
f
o
r
m
er
-
b
ased
m
o
d
els
(
e.
g
.
,
b
id
ir
ec
tio
n
al
e
n
co
d
e
r
r
ep
r
esen
tatio
n
s
f
r
o
m
tr
an
s
f
o
r
m
er
s
(
B
E
R
T
)
)
an
d
d
ee
p
lea
r
n
in
g
a
r
ch
itectu
r
es
(
co
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
(
C
NN
)
-
lo
n
g
s
h
o
r
t
-
ter
m
m
em
o
r
y
(
L
STM
)
h
y
b
r
id
s
)
h
av
e
s
h
o
wn
s
tr
o
n
g
p
er
f
o
r
m
an
ce
in
te
x
t
class
if
icatio
n
[
7
]
,
th
is
s
tu
d
y
d
elib
er
ately
f
o
cu
s
es
o
n
class
ical
m
ac
h
in
e
lear
n
i
n
g
ap
p
r
o
ac
h
es
f
o
r
two
r
ea
s
o
n
s
:
i
n
ter
p
r
etab
ilit
y
o
f
f
ea
tu
r
e
co
n
tr
ib
u
tio
n
s
,
wh
ich
is
ess
en
tial
f
o
r
u
n
d
er
s
tan
d
in
g
th
e
r
o
le
o
f
PII
f
ea
tu
r
es
an
d
s
u
p
p
o
r
tin
g
h
u
m
an
m
o
d
er
ato
r
wo
r
k
f
lo
ws
[
2
3
]
,
[
2
4
]
a
n
d
l
o
wer
co
m
p
u
tatio
n
al
o
v
er
h
ea
d
,
en
a
b
lin
g
p
o
ten
tial
d
ep
lo
y
m
en
t
in
r
ea
l
-
tim
e
m
o
n
ito
r
in
g
s
y
s
tem
s
[
2
5
]
.
B
en
ch
m
ar
k
in
g
ag
ain
s
t
tr
an
s
f
o
r
m
er
-
b
ased
a
p
p
r
o
a
ch
es
is
lef
t
f
o
r
f
u
tu
r
e
wo
r
k
.
T
o
en
s
u
r
e
m
o
d
el
p
er
f
o
r
m
a
n
ce
,
f
r
o
m
1
1
,
7
7
7
p
o
s
ts
,
f
o
r
tr
ain
in
g
t
h
e
class
if
icatio
n
m
o
d
els,
8
0
%
o
f
th
e
d
ata
(
9
,
4
2
1
p
o
s
ts
)
was
u
s
ed
f
o
r
tr
ain
in
g
,
wh
ile
th
e
r
em
ain
in
g
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.
15
,
No
.
4
,
Au
g
u
s
t 2
0
2
6
:
3
3
7
6
-
3
3
8
8
3380
2
0
%
(
2
,
3
5
6
p
o
s
ts
)
was
r
eser
v
ed
f
o
r
test
in
g
.
T
h
e
s
ize
o
f
t
h
e
tr
ain
in
g
d
ata
was
ch
o
s
en
to
en
s
u
r
e
a
m
o
d
e
l
tr
ain
in
g
p
r
o
ce
s
s
wh
ile
m
ain
tain
in
g
s
u
f
f
icien
t
d
ata
f
o
r
u
n
b
iased
test
in
g
an
d
e
v
alu
atio
n
.
T
h
is
r
esear
ch
co
m
p
ar
ed
m
o
d
els
an
d
r
an
k
ed
th
e
m
o
d
els
b
ased
o
n
a
s
p
ec
if
ied
m
etr
ic,
in
th
is
ca
s
e,
ac
cu
r
ac
y
,
th
u
s
p
r
o
v
id
i
n
g
a
c
lear
in
d
icatio
n
o
f
wh
ich
m
o
d
el
p
er
f
o
r
m
e
d
b
est
f
o
r
o
u
r
s
p
ec
if
ic
task
.
T
h
en
co
m
p
ar
e
th
e
r
esu
lts
in
th
e
s
p
id
er
ch
ar
t
f
o
r
m
etr
ics
r
ec
all,
p
r
ec
is
io
n
,
an
d
F1
-
s
co
r
e.
Pre
ci
s
io
n
an
d
r
ec
all
ar
e
cr
u
cial
m
etr
ics
to
ev
alu
ate
wh
en
d
ea
lin
g
with
b
alan
ce
d
d
a
ta.
Ho
wev
er
,
f
o
r
u
n
b
alan
ce
d
d
ata,
th
e
F1
-
s
co
r
e
b
ec
o
m
es
th
e
m
o
s
t
s
ig
n
if
ican
t
m
etr
ic
to
co
n
s
id
er
[
2
6
]
.
T
h
is
e
v
alu
atio
n
p
r
o
v
i
d
es
in
s
ig
h
ts
in
to
th
e
s
tr
en
g
t
h
s
an
d
wea
k
n
ess
es
o
f
ea
ch
m
o
d
el
in
th
e
co
n
te
x
t
o
f
d
etec
tin
g
in
te
n
t
to
s
ell
illeg
al
d
r
u
g
ac
tiv
ities
o
n
X
.
T
o
en
h
a
n
ce
th
e
class
if
icatio
n
o
f
p
o
s
ts
in
d
icatin
g
in
ten
t
to
s
ell
illeg
al
d
r
u
g
s
,
r
e
g
u
lar
e
x
p
r
ess
io
n
(
r
eg
ex
)
p
atter
n
s
wer
e
d
ev
elo
p
e
d
to
id
en
tify
an
d
ex
tr
ac
t
PII
elem
en
ts
s
u
ch
as
p
h
o
n
e
n
u
m
b
er
s
,
em
ail
ad
d
r
ess
es,
an
d
web
UR
L
s
f
r
o
m
th
e
p
o
s
t
tex
ts
.
R
eg
u
lar
ex
p
r
ess
io
n
s
ar
e
a
p
o
wer
f
u
l
to
o
l
f
o
r
p
atter
n
m
atch
in
g
with
in
tex
t,
allo
win
g
f
o
r
th
e
ef
f
icien
t
d
etec
tio
n
o
f
s
p
ec
if
ic
p
atter
n
s
co
r
r
esp
o
n
d
in
g
to
d
if
f
er
en
t
ty
p
es
o
f
PII
.
C
o
ld
ea
et
a
l.
[
2
7
]
d
ev
elo
p
ed
a
r
eg
ex
-
b
ased
ap
p
r
o
ac
h
f
o
r
f
la
g
g
in
g
s
en
s
itiv
e
d
ata
in
lar
g
e
h
ea
lth
d
atasets
,
im
p
r
o
v
in
g
d
a
ta
q
u
ality
r
e
p
o
r
tin
g
.
T
ab
le
5
s
u
m
m
ar
izes th
e
r
eg
ex
p
atter
n
s
im
p
lem
en
ted
.
B
y
ap
p
ly
in
g
th
ese
r
eg
ex
p
at
ter
n
s
to
th
e
p
o
s
ts
d
ata,
th
e
o
cc
u
r
r
e
n
ce
s
o
f
p
h
o
n
e
n
u
m
b
er
s
,
em
ail
ad
d
r
ess
es,
an
d
web
UR
L
s
we
r
e
s
y
s
tem
atica
lly
id
en
tifie
d
an
d
co
u
n
ted
.
T
h
ese
elem
en
ts
wer
e
th
en
en
co
d
e
d
as
ad
d
itio
n
al
b
in
ar
y
f
ea
tu
r
es
in
t
h
e
d
ataset,
in
d
icatin
g
th
e
p
r
es
en
ce
o
r
ab
s
en
ce
o
f
ea
c
h
ty
p
e
o
f
PII
in
a
p
o
s
t
.
B
y
in
clu
d
in
g
PII
an
d
web
UR
L
s
f
ea
tu
r
es
alo
n
g
s
id
e
tr
ad
itio
n
al
tex
tu
al
f
ea
tu
r
es
(
e.
g
.
,
T
F
-
I
DF
v
ec
to
r
s
)
,
th
e
im
p
ac
t
o
f
PII
d
etec
tio
n
o
n
th
e
p
e
r
f
o
r
m
an
ce
o
f
t
h
e
class
if
icatio
n
m
o
d
els wa
s
ass
e
s
s
ed
.
T
ab
le
5
.
I
m
p
lem
en
ted
r
eg
e
x
p
atter
n
N
a
me
R
e
g
e
x
p
a
t
t
e
r
n
P
u
r
p
o
se
p
h
o
n
e
_
p
a
t
t
e
r
N
\
b
(
?
:
\
+?(
\
d
{
1
,
3
})
?
[
-
.
●
]
?
(
\
d
{1
,
4
})
?
[
-
.
●
]
?
(
\
d
{
1
,
4
})
?
[
-
.
●
]
?
(
\
d
{
1
,
9
})
)
\
b
D
e
t
e
c
t
s
p
h
o
n
e
n
u
m
b
e
r
s
,
i
n
c
l
u
d
i
n
g
i
n
t
e
r
n
a
t
i
o
n
a
l
f
o
r
mat
s
w
i
t
h
-
,
.
,
o
r
●
a
s
se
p
a
r
a
t
o
r
s
e
mai
l
_
p
a
t
t
e
r
N
\
b
[
A
-
Za
-
z0
-
9
.
_
%+
-
]
+
@
[
A
-
Za
-
z0
-
9.
-
]+
\
.
[
A
-
Z|
a
-
z
]
{2
,
}
\
b
D
e
t
e
c
t
s
e
ma
i
l
a
d
d
r
e
sses
u
r
l
_
p
a
t
t
e
r
n
\
b
(
?
:
h
t
t
p
[
s]
?
:
/
/
)
?
(
?
:
w
w
w
\
.
)
?
[
a
-
zA
-
Z0
-
9.
-
]+
\
.
[
a
-
zA
-
Z]
{2
,
6
}
\
b
D
e
t
e
c
t
s w
e
b
s
i
t
e
U
R
Ls
(
w
i
t
h
o
r
w
i
t
h
o
u
t
h
t
t
p
(
s)
:
/
/
a
n
d
w
w
w
.
p
r
e
f
i
x
)
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
A
t
o
t
a
l
o
f
1
1
,
7
7
7
p
o
s
t
s
w
e
r
e
p
r
o
c
e
s
s
e
d
a
n
d
u
s
e
d
t
o
t
r
a
i
n
a
n
d
e
v
a
l
u
a
t
e
f
i
v
e
c
l
ass
i
f
i
e
r
s
.
SV
M
a
c
h
i
e
v
e
d
t
h
e
h
i
g
h
e
s
t
o
v
e
r
a
ll
F
1
-
s
c
o
r
e
o
f
9
7
.
0
8
%
,
o
u
t
p
e
r
f
o
r
m
i
n
g
n
a
ï
v
e
B
a
y
es
(
9
4
.
8
5
%
)
,
l
o
g
i
s
ti
c
r
eg
r
e
s
s
i
o
n
(
9
6
.
6
5
%
)
,
r
a
n
d
o
m
f
o
r
e
s
t
(
9
6
.
8
3
%
)
,
a
n
d
g
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
(
9
6
.
4
9
%
)
.
S
V
M
w
a
s
t
h
e
r
e
f
o
r
e
s
e
l
e
ct
e
d
a
s
t
h
e
b
e
s
t
-
p
e
r
f
o
r
m
i
n
g
c
l
a
s
s
i
f
ie
r
f
o
r
f
u
r
t
h
e
r
a
n
a
l
y
s
i
s
.
F
i
g
u
r
e
2
p
r
e
s
e
n
t
s
t
h
e
c
o
m
p
a
r
is
o
n
o
f
t
e
x
t
c
l
a
s
s
i
f
ic
a
t
i
o
n
p
e
r
f
o
r
m
a
n
c
e
a
c
r
o
s
s
a
ll
m
o
d
e
l
s
u
s
i
n
g
a
s
p
i
d
e
r
c
h
a
r
t
.
T
ab
l
e
6
s
u
m
m
a
r
i
z
es
t
h
e
cl
a
s
s
i
f
i
c
at
i
o
n
p
e
r
f
o
r
m
a
n
c
e
m
e
t
r
i
cs
a
c
r
o
s
s
a
l
l
m
o
d
el
s
.
Fig
u
r
e
2
.
C
o
m
p
a
r
is
o
n
o
f
tex
t c
lass
if
icatio
n
p
er
f
o
r
m
an
ce
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
Dete
ctio
n
o
f illeg
a
l d
r
u
g
s
a
cti
vities in
I
n
d
o
n
esia
via
s
o
ci
a
l
med
ia
X
(
A
d
i H
a
n
if S
ed
a
r
)
3381
T
ab
le
6
.
C
lass
if
icatio
n
p
er
f
o
r
m
an
ce
o
n
in
ten
t to
s
ell
(
p
o
s
itiv
e)
class
o
n
th
e
v
alid
atio
n
s
et
M
o
d
e
l
P
r
e
c
i
s
i
o
n
(
%)
R
e
c
a
l
l
(
%)
F1
-
sc
o
r
e
(
i
n
t
e
n
t
)
(
%)
F1
-
sc
o
r
e
(
w
e
i
g
h
t
e
d
)
(
%)
N
a
ï
v
e
B
a
y
e
s
9
0
.
0
5
9
3
.
3
1
9
1
.
6
6
9
4
.
8
5
S
V
M
9
8
.
5
1
9
1
.
9
2
9
5
.
1
0
9
7
.
0
8
Lo
g
i
s
t
i
c
r
e
g
r
e
ssi
o
n
9
7
.
9
0
9
1
.
0
9
9
4
.
3
7
9
6
.
6
5
R
a
n
d
o
m
f
o
r
e
s
t
9
7
.
9
2
9
1
.
6
4
9
4
.
6
8
9
6
.
8
3
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
9
7
.
3
2
9
1
.
0
9
9
4
.
1
0
9
6
.
4
9
Fo
llo
win
g
th
e
co
m
p
ar
is
o
n
o
f
all
m
o
d
els,
a
clo
s
er
ex
am
in
atio
n
o
f
th
e
b
est
-
p
er
f
o
r
m
i
n
g
class
if
ier
was
co
n
d
u
cte
d
.
Fig
u
r
e
3
p
r
esen
ts
th
e
co
n
f
u
s
io
n
m
atr
ix
o
f
th
e
SVM
clas
s
if
ier
o
n
th
e
test
in
g
d
ata
(
2
,
3
5
6
p
o
s
ts
)
,
s
h
o
win
g
1
,
6
3
3
tr
u
e
n
eg
ativ
es,
1
4
f
alse
p
o
s
itiv
es,
3
9
f
alse
n
eg
ativ
es,
an
d
6
7
0
tr
u
e
p
o
s
itiv
es.
Ov
er
all,
th
e
m
o
d
el
c
o
r
r
ec
tly
class
if
ied
2
,
3
0
3
o
u
t
o
f
2
,
3
5
6
p
o
s
ts
,
with
o
n
ly
5
3
m
is
class
if
icatio
n
s
.
Fig
u
r
e
3
p
r
esen
ts
th
e
co
n
f
u
s
io
n
m
atr
ix
.
An
er
r
o
r
a
n
aly
s
is
o
f
t
h
e
m
is
class
if
ied
p
o
s
ts
in
Fig
u
r
e
3
p
r
esen
ts
th
e
c
o
n
f
u
s
io
n
m
atr
i
x
r
ev
ea
le
d
d
is
tin
ct
p
atter
n
s
f
o
r
ea
ch
er
r
o
r
ty
p
e.
T
h
e
3
9
f
alse
n
eg
ativ
e
s
wer
e
d
o
m
in
ated
b
y
s
h
o
r
t,
c
o
n
v
er
s
atio
n
al
p
o
s
ts
th
at
m
en
tio
n
ed
d
r
u
g
n
am
es
with
o
u
t
ex
p
licit
co
m
m
er
cial
cu
es,
s
u
ch
as
b
r
ief
r
ep
lies
r
ef
er
en
cin
g
“
o
b
a
t
b
iu
s
”
(
an
esth
etic
d
r
u
g
s
)
o
r
h
ash
tag
-
o
n
ly
p
o
s
ts
with
o
u
t
s
u
p
p
o
r
ti
n
g
c
o
n
tex
t.
Ma
n
y
o
f
th
ese
p
o
s
ts
lack
ed
p
h
o
n
e
n
u
m
b
er
s
,
UR
L
s
,
o
r
p
r
icin
g
in
f
o
r
m
atio
n
,
wh
ich
m
ad
e
th
e
s
ellin
g
in
ten
t
d
if
f
icu
lt
to
d
etec
t
f
r
o
m
tex
tu
al
f
ea
tu
r
es
alo
n
e.
T
h
e
1
4
f
alse
p
o
s
itiv
es,
in
co
n
t
r
ast,
wer
e
ty
p
ically
p
r
o
m
o
tio
n
al
-
s
ty
le
p
o
s
ts
co
n
tain
in
g
s
tr
o
n
g
PII
s
ig
n
als
s
u
ch
as
p
h
o
n
e
n
u
m
b
e
r
s
an
d
U
R
L
s
,
in
d
icatin
g
th
at
th
e
m
o
d
e
l
o
cc
asio
n
ally
o
v
er
-
r
elied
o
n
th
ese
f
ea
tu
r
es
wh
en
tex
tu
al
co
n
te
x
t
was
am
b
ig
u
o
u
s
.
T
h
e
h
ig
h
er
n
u
m
b
er
o
f
f
alse
n
eg
ativ
es
co
m
p
a
r
ed
t
o
f
alse
p
o
s
itiv
es
also
ex
p
lain
s
th
e
m
o
d
el'
s
lo
wer
r
e
ca
ll
(
9
4
.
5
0
%)
r
elativ
e
to
its
p
r
ec
is
io
n
(
9
7
.
9
5
%)
o
n
t
h
e
test
in
g
s
et
(
2
,
3
5
6
p
o
s
ts
)
.
T
ab
le
7
p
r
esen
ts
r
ep
r
esen
tativ
e
ex
am
p
les o
f
m
is
class
if
ied
p
o
s
ts
f
r
o
m
ea
ch
er
r
o
r
ty
p
e.
Fig
u
r
e
3
.
C
o
n
f
u
s
io
n
m
atr
i
x
o
f
SVM
class
if
ier
o
n
test
in
g
d
ata
T
ab
le
1
.
E
x
am
p
les o
f
m
is
class
if
ied
p
o
s
ts
b
y
SVM
Ty
p
e
P
o
st
s
f
r
o
m X
Tr
u
e
l
a
b
e
l
P
r
e
d
i
c
t
e
d
F
a
l
se
n
e
g
a
t
i
v
e
"
@
u
s
e
r
o
b
a
t
b
i
u
s k
a
k
c
e
p
e
t
k
o
k
"
(
A
n
e
st
h
e
t
i
c
d
r
u
g
s
b
r
o
,
w
o
r
k
s
f
a
st
)
I
n
t
e
n
t
t
o
s
e
l
l
N
o
n
-
i
n
t
e
n
t
F
a
l
se
n
e
g
a
t
i
v
e
"#
p
e
n
g
g
u
g
u
rk
a
n
d
u
n
g
a
n
#
m
i
s
o
p
r
o
st
o
l
#
c
y
t
o
t
e
c
o
ri
g
i
n
a
l
#
a
b
o
rs
i
[U
RL]
"
I
n
t
e
n
t
t
o
s
e
l
l
N
o
n
-
i
n
t
e
n
t
F
a
l
se
p
o
si
t
i
v
e
"
A
r
t
o
f
t
h
e
d
a
y
:
'
c
y
t
o
t
e
c
m
i
s
o
p
r
o
st
o
l
'
.
V
i
e
w
a
t
:
[
U
R
L]
"
N
o
n
-
i
n
t
e
n
t
I
n
t
e
n
t
t
o
s
e
l
l
F
a
l
se
p
o
si
t
i
v
e
"
T
u
n
t
a
s
Bo
y
o
l
a
l
i
2
b
l
n
l
e
b
i
h
[U
RL]
"
(
C
o
m
p
l
e
t
e
B
o
y
o
l
a
l
i
2
+
m
o
n
t
h
s)
N
o
n
-
i
n
t
e
n
t
I
n
t
e
n
t
t
o
s
e
l
l
T
h
ese
f
in
d
in
g
s
s
u
g
g
est
th
at
in
co
r
p
o
r
atin
g
co
n
tex
t
u
al
cu
es
b
ey
o
n
d
k
e
y
wo
r
d
s
an
d
PII
f
o
r
ex
am
p
le
,
lev
er
ag
in
g
co
n
v
er
s
atio
n
al
co
n
tex
t
o
r
s
lan
g
-
awa
r
e
em
b
e
d
d
in
g
s
,
co
u
ld
f
u
r
th
e
r
im
p
r
o
v
e
class
if
icatio
n
ac
cu
r
ac
y
.
Af
ter
th
e
class
if
icatio
n
an
d
e
v
alu
atio
n
p
r
o
ce
s
s
,
m
atp
lo
tlib
an
d
p
an
d
a'
s
lib
r
ar
ies
wer
e
u
s
ed
to
v
is
u
alize
th
e
d
is
tr
ib
u
tio
n
an
d
tr
e
n
d
s
o
f
th
e
p
o
s
ts
.
T
h
e
d
is
tr
ib
u
tio
n
o
f
p
o
s
ts
class
if
ied
as
in
ten
t
to
s
ell
v
s
.
n
o
n
-
in
ten
t
is
s
h
o
wn
in
Fig
u
r
e
4
.
Fig
u
r
e
4
r
ep
r
esen
ts
th
e
tex
t
class
if
icatio
n
o
f
p
o
s
ts
th
at
in
ten
d
to
s
ell
illi
cit
d
r
u
g
s
u
s
in
g
a
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.
15
,
No
.
4
,
Au
g
u
s
t 2
0
2
6
:
3
3
7
6
-
3
3
8
8
3382
v
is
u
al
r
ep
r
esen
tatio
n
.
3
,
392
p
o
s
ts
s
h
o
wed
in
ten
t
to
s
ell
an
d
8
,
385
p
o
s
ts
ar
e
n
eu
tr
al
o
r
in
d
icate
d
no
in
ten
t
to
s
ell.
T
h
e
d
is
tr
ib
u
tio
n
o
f
p
o
s
ts
o
v
er
tim
e
th
at
e
x
p
r
ess
an
in
ten
t to
s
ell
illeg
al
d
r
u
g
s
is
p
r
esen
t
ed
in
Fig
u
r
e
5
.
Fig
u
r
e
4
.
I
n
ten
t to
s
ell
class
if
icatio
n
Fig
u
r
e
5
.
Dis
tr
ib
u
tio
n
s
o
f
p
o
s
t
o
v
er
tim
e
Fig
u
r
e
5
s
h
o
ws
th
e
d
is
tr
ib
u
tio
n
o
f
in
ten
t
-
to
-
s
ell
p
o
s
ts
f
r
o
m
Ma
y
2
0
2
3
t
o
Ma
y
2
0
2
4
.
Activ
ity
r
em
ain
ed
l
o
w
an
d
s
tead
y
u
n
til
late
No
v
em
b
er
2
0
2
3
,
f
o
llo
we
d
b
y
a
s
h
a
r
p
s
p
i
k
e
in
Dec
em
b
er
2
0
2
3
(
p
ea
k
i
n
g
at
1
6
0
p
o
s
ts
)
,
with
s
m
aller
f
lu
c
tu
atio
n
s
th
r
o
u
g
h
Ma
y
2
0
2
4
.
T
h
is
s
p
ik
e
m
a
y
in
d
icate
h
ei
g
h
ten
ed
illi
cit
s
ales
ac
tiv
ity
d
u
r
in
g
th
at
p
er
io
d
.
T
h
e
n
u
m
b
er
o
f
p
o
s
ts
g
r
o
u
p
ed
b
y
to
p
ic
is
p
r
esen
ted
i
n
Fig
u
r
e
6
.
Fig
u
r
e
6
illu
s
tr
ates
th
e
n
u
m
b
er
o
f
p
o
s
ts
g
r
o
u
p
e
d
b
y
th
e
t
y
p
e
o
f
d
r
u
g
m
e
n
tio
n
ed
i
n
th
e
p
o
s
ts
.
T
h
e
ch
ar
t
co
m
p
a
r
es
th
r
ee
ca
teg
o
r
ies:
ab
o
r
tio
n
p
ills
,
s
ex
u
al
en
h
an
ce
m
en
t
d
r
u
g
s
,
an
d
an
esth
etic
d
r
u
g
s
.
T
h
e
d
ata
s
h
o
ws
th
at
a
b
o
r
tio
n
p
ills
ar
e
th
e
m
ajo
r
ity
,
with
a
to
tal
o
f
3
,
0
6
4
p
o
s
ts
.
T
h
is
is
s
ig
n
if
ica
n
tly
h
ig
h
er
th
an
th
e
o
th
er
ca
te
g
o
r
ies,
in
d
icatin
g
a
p
r
ed
o
m
i
n
an
t
f
o
cu
s
o
n
th
is
ty
p
e
o
f
d
r
u
g
in
t
h
e
d
ataset.
Sex
u
a
l
en
h
an
ce
m
e
n
t
d
r
u
g
s
ar
e
m
en
t
io
n
ed
in
1
3
9
p
o
s
ts
,
wh
ile
an
esth
etic
d
r
u
g
s
a
p
p
ea
r
in
1
8
9
p
o
s
ts
.
T
h
e
s
tar
k
d
if
f
er
e
n
ce
in
th
e
n
u
m
b
er
o
f
p
o
s
ts
b
et
wee
n
ab
o
r
tio
n
p
ills
an
d
th
e
o
th
er
two
ca
teg
o
r
ies
s
u
g
g
ests
th
at
d
is
cu
s
s
io
n
s
r
el
ated
to
ab
o
r
tio
n
p
ills
ar
e
f
ar
m
o
r
e
p
r
ev
ale
n
t
o
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
Dete
ctio
n
o
f illeg
a
l d
r
u
g
s
a
cti
vities in
I
n
d
o
n
esia
via
s
o
ci
a
l
med
ia
X
(
A
d
i H
a
n
if S
ed
a
r
)
3383
s
o
cial
m
ed
ia.
T
h
is
co
u
ld
r
ef
lect
cu
r
r
en
t
ev
en
ts
,
p
u
b
lic
in
ter
est,
o
r
o
n
g
o
in
g
d
is
cu
s
s
io
n
s
in
s
o
ciety
r
elate
d
to
th
is
d
r
u
g
.
T
h
e
co
u
n
t o
f
p
o
s
ts
co
n
tain
in
g
PI
I
is
p
r
esen
ted
in
F
ig
u
r
e
7
.
Fig
u
r
e
7
illu
s
tr
ates
th
e
co
u
n
t
o
f
p
o
s
ts
th
at
co
n
tain
PII
,
ca
teg
o
r
ized
in
to
p
h
o
n
e
n
u
m
b
er
s
,
em
ail
ad
d
r
ess
es,
an
d
web
UR
L
s
.
T
h
e
d
ata
r
ev
ea
ls
a
s
ig
n
if
ican
t
p
r
esen
ce
o
f
PII
in
p
o
s
ts
,
with
p
h
o
n
e
n
u
m
b
er
s
ap
p
ea
r
in
g
in
4
,
0
2
4
p
o
s
ts
an
d
web
UR
L
s
in
4
,
8
2
2
p
o
s
ts
.
I
t
is
im
p
o
r
tan
t
to
n
o
te
t
h
at
s
o
m
e
p
o
s
ts
m
ay
co
n
tain
b
o
th
p
h
o
n
e
n
u
m
b
er
s
an
d
web
UR
L
s
,
lead
in
g
to
o
v
er
lap
p
in
g
co
u
n
ts
.
T
h
e
to
tal
n
u
m
b
er
o
f
p
o
s
ts
co
n
tain
in
g
PII
is
n
o
t
th
e
s
u
m
o
f
th
ese
two
ca
teg
o
r
ies,
as
s
o
m
e
p
o
s
ts
in
clu
d
ed
b
o
t
h
ty
p
es
o
f
PII
.
N
o
tab
ly
,
n
o
p
o
s
ts
wer
e
f
o
u
n
d
to
co
n
tain
em
ail
ad
d
r
ess
es.
T
h
is
h
ig
h
lig
h
ts
th
e
e
x
ten
t
to
wh
ich
u
s
er
s
m
ay
b
e
s
h
ar
in
g
s
en
s
itiv
e
in
f
o
r
m
atio
n
o
n
s
o
cial
m
ed
ia.
Fig
u
r
e
1
.
Nu
m
b
er
o
f
p
o
s
ts
g
r
o
u
p
ed
b
y
to
p
ic
Fig
u
r
e
2
.
C
o
u
n
t o
f
p
o
s
ts
c
o
n
ta
in
in
g
PII
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.
15
,
No
.
4
,
Au
g
u
s
t 2
0
2
6
:
3
3
7
6
-
3
3
8
8
3384
T
o
ev
alu
ate
th
e
s
ig
n
if
ican
ce
o
f
th
e
PII
f
ea
tu
r
es
in
o
u
r
cl
ass
if
icatio
n
task
,
we
p
er
f
o
r
m
ed
f
ea
tu
r
e
im
p
o
r
tan
ce
an
aly
s
is
u
s
in
g
t
h
e
r
an
d
o
m
f
o
r
est
class
if
ier
,
wh
ich
p
r
o
v
id
es
i
n
h
er
e
n
t
m
ea
s
u
r
es
o
f
f
ea
t
u
r
e
im
p
o
r
tan
ce
b
ased
o
n
th
e
Gin
i
im
p
u
r
ity
r
ed
u
cti
o
n
.
T
h
e
f
ea
tu
r
e
im
p
o
r
tan
ce
a
n
aly
s
is
id
en
tifi
ed
th
e
f
ea
tu
r
es
th
at
d
id
n
o
t
co
n
tr
ib
u
te
to
th
e
class
if
ier
'
s
p
er
f
o
r
m
an
ce
,
en
a
b
lin
g
i
m
p
r
o
v
e
m
en
ts
in
ac
cu
r
ac
y
[
2
3
]
,
[
2
4
]
.
T
h
e
r
esu
lts
in
d
icate
d
th
at
PII
f
ea
tu
r
es,
p
a
r
ticu
lar
ly
th
e
p
r
esen
ce
o
f
p
h
o
n
e
n
u
m
b
er
s
an
d
web
UR
L
s
,
wer
e
am
o
n
g
t
h
e
to
p
co
n
tr
ib
u
to
r
s
to
th
e
m
o
d
el'
s
ab
ilit
y
to
ac
cu
r
ately
class
if
y
p
o
s
ts
with
in
ten
t
to
s
ell
illeg
al
d
r
u
g
s
.
T
ab
le
8
s
h
o
ws
th
e
f
ea
tu
r
e
im
p
o
r
tan
ce
s
co
r
e.
T
ab
le
8
.
Featu
r
e
im
p
o
r
tan
ce
s
co
r
e
R
a
n
k
F
e
a
t
u
r
e
I
mp
o
r
t
a
n
c
e
s
c
o
r
e
1
W
o
r
d
:
a
b
o
rs
i
(
a
b
o
r
t
i
o
n
)
0
.
0
8
1
6
2
W
o
r
d
:
c
y
t
o
t
e
c
0
.
0
7
0
6
3
W
o
r
d
:
p
e
n
g
g
u
g
u
r
(
a
b
o
r
t
i
o
n
i
n
d
u
c
e
r
)
0
.
0
4
9
2
4
P
r
e
sen
c
e
o
f
p
h
o
n
e
n
u
m
b
e
r
0
.
0
4
8
0
5
P
r
e
sen
c
e
o
f
U
R
L
0
.
0
4
0
3
6
W
o
r
d
:
m
i
so
p
r
o
st
o
l
0
.
0
2
4
2
7
W
o
r
d
:
t
u
n
t
a
s
(
c
o
m
p
l
e
t
e
)
0
.
0
2
2
1
8
W
o
r
d
:
a
sl
i
(
o
r
i
g
i
n
a
l
)
0
.
0
1
4
9
T
h
e
f
ea
tu
r
e
im
p
o
r
tan
ce
s
co
r
es
r
ev
ea
l
th
at
th
e
p
r
esen
ce
o
f
a
p
h
o
n
e
n
u
m
b
er
an
d
UR
L
ar
e
o
n
to
p
f
iv
e
s
ig
n
if
ican
t
f
ea
tu
r
es
,
in
d
icatin
g
th
at
p
o
s
ts
co
n
tain
in
g
p
h
o
n
e
n
u
m
b
e
r
s
ar
e
s
tr
o
n
g
in
d
icato
r
s
o
f
in
ten
t
to
s
ell
illeg
al
d
r
u
g
s
.
Similar
ly
,
th
e
p
r
esen
ce
o
f
a
web
UR
L
is
also
h
ig
h
ly
in
f
lu
en
tial.
T
r
ad
itio
n
al
t
ex
tu
al
f
ea
tu
r
es
lik
e
s
p
ec
if
ic
k
ey
wo
r
d
s
(
e.
g
.
,
“
a
b
o
r
s
i
(
ab
o
r
tio
n
)
”,
“
cy
to
t
ec
(
a
b
o
r
tio
n
p
ill)”
,
“
p
e
n
g
g
u
g
u
r
(
ab
o
r
tio
n
in
d
u
ce
r
)
”
)
r
em
ain
im
p
o
r
tan
t,
b
u
t th
e
in
c
o
r
p
o
r
atio
n
o
f
PI
I
f
ea
tu
r
es e
n
h
an
ce
s
th
e
m
o
d
el'
s
ab
ilit
y
to
d
etec
t illicit in
ten
t.
W
h
ile
th
e
p
r
esen
ce
o
f
p
h
o
n
e
n
u
m
b
er
s
an
d
o
th
er
PII
s
ig
n
if
i
ca
n
tly
b
o
o
s
ts
th
e
ac
cu
r
ac
y
o
f
d
etec
tin
g
illi
cit
d
r
u
g
s
ale
p
o
s
ts
,
it
d
o
es
n
o
t
au
to
m
atica
lly
lead
to
ea
s
y
ap
p
r
eh
e
n
s
io
n
o
f
o
f
f
en
d
e
r
s
.
C
r
itical
ch
allen
g
e
in
u
s
in
g
p
h
o
n
e
n
u
m
b
er
s
an
d
PII
f
o
r
in
v
esti
g
atin
g
n
a
r
co
tics
cr
im
es
in
I
n
d
o
n
esia
is
d
u
e
to
wea
k
n
ess
es
in
SIM
ca
r
d
r
e
g
is
tr
atio
n
p
o
licies
an
d
d
ata
s
h
ar
in
g
b
etwe
en
telec
o
m
p
r
o
v
id
er
s
a
n
d
law
e
n
f
o
r
ce
m
en
t
[
2
8
]
.
Peo
p
le
en
g
ag
in
g
in
d
r
u
g
tr
an
s
ac
tio
n
s
o
n
s
o
cial
m
ed
ia
also
g
e
n
e
r
ally
r
ep
o
r
t
lo
w
co
n
ce
r
n
f
o
r
law
en
f
o
r
ce
m
e
n
t
ex
p
o
s
u
r
e
,
b
u
t
s
till
em
p
lo
y
r
is
k
m
an
ag
em
en
t
s
tr
ateg
ies,
s
u
ch
as
u
s
in
g
s
elf
-
d
eletin
g
m
ess
ag
es
an
d
co
d
e
lan
g
u
ag
e
[
2
9
]
.
T
h
ese
f
ac
to
r
s
m
ak
e
it
d
if
f
icu
lt
f
o
r
au
th
o
r
itie
s
to
lin
k
p
h
o
n
e
n
u
m
b
er
s
to
ac
tu
al
in
d
iv
id
u
als
o
r
d
is
m
an
tle
b
r
o
a
d
er
n
etwo
r
k
s
,
d
esp
ite
th
e
a
v
ailab
ilit
y
o
f
PII
.
T
h
e
wo
r
d
clo
u
d
o
f
m
o
s
t
u
s
ed
h
ash
tag
s
is
p
r
esen
ted
in
Fig
u
r
e
8
.
Fig
u
r
e
8
.
W
o
r
d
clo
u
d
o
f
t
h
e
m
o
s
t u
s
ed
h
ash
tag
s
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
Dete
ctio
n
o
f illeg
a
l d
r
u
g
s
a
cti
vities in
I
n
d
o
n
esia
via
s
o
ci
a
l
med
ia
X
(
A
d
i H
a
n
if S
ed
a
r
)
3385
Fig
u
r
e
8
s
h
o
ws
a
wo
r
d
clo
u
d
o
f
th
e
m
o
s
t
u
s
ed
h
ash
tag
s
in
th
e
p
o
s
ts
we
co
llected
.
T
h
e
b
ig
g
e
r
th
e
wo
r
d
,
th
e
m
o
r
e
o
f
ten
it
ap
p
ea
r
ed
.
T
h
e
b
ig
g
est
wo
r
d
s
,
lik
e
“
o
b
a
t
a
b
o
r
s
i
,
”
“c
y
to
tec,
”
an
d
“
p
en
g
g
u
g
u
r
k
a
n
d
u
n
g
a
n
,
”
ar
e
th
e
m
o
s
t
c
o
m
m
o
n
h
ash
ta
g
s
.
T
h
is
s
h
o
ws
th
at
p
o
s
ts
ar
e
m
o
s
tly
ab
o
u
t
ab
o
r
tio
n
p
ills
,
esp
ec
ially
th
e
“c
y
to
tec”
b
r
an
d
.
Oth
er
c
o
m
m
o
n
h
ash
tag
s
ar
e
“ju
alo
b
atcy
t
o
tec”
an
d
“
m
is
o
p
r
o
s
to
l,”
wh
ic
h
also
r
elate
to
s
p
ec
if
ic
d
r
u
g
s
a
n
d
th
eir
s
ale.
T
h
e
r
esu
lts
f
r
o
m
th
is
h
ig
h
lig
h
t
th
e
im
p
o
r
tan
ce
an
d
f
ea
s
ib
ilit
y
o
f
u
tili
zin
g
s
o
cial
m
ed
ia
d
ata,
s
p
ec
if
ically
X
d
ata,
as
a
v
alu
ab
le
s
o
u
r
ce
o
f
tex
t
class
if
icatio
n
an
aly
s
is
.
Su
ch
an
aly
s
is
ca
n
b
e
cr
u
cial
in
g
u
id
in
g
d
ec
is
io
n
-
m
a
k
in
g
p
r
o
ce
s
s
es in
v
ar
io
u
s
s
ec
to
r
s
,
lik
e
p
o
licy
f
o
r
m
u
latio
n
.
T
h
e
p
e
r
f
o
r
m
an
ce
o
f
o
u
r
SV
M
m
o
d
el,
ac
h
ie
v
in
g
an
o
v
e
r
a
ll
F1
-
s
co
r
e
o
f
9
7
.
0
8
%,
is
c
o
n
s
is
ten
t
with
f
in
d
in
g
s
f
r
o
m
r
elate
d
s
tu
d
ies
o
n
d
r
u
g
-
r
elate
d
tex
t
class
if
icatio
n
in
s
o
cial
m
ed
ia.
Alr
u
waili
et
a
l.
[
3
0
]
r
ep
o
r
ted
an
ac
cu
r
ac
y
o
f
9
6
%
wh
e
n
u
s
in
g
SVM
to
d
etec
t
d
r
u
g
-
s
ellin
g
b
e
h
av
io
r
s
o
n
Ar
ab
ic
p
o
s
ts
,
wh
ich
clo
s
ely
alig
n
s
with
o
u
r
r
esu
lts
an
d
s
u
g
g
ests
th
at
SVM
is
r
o
b
u
s
t
ac
r
o
s
s
d
if
f
er
en
t
lan
g
u
ag
es
f
o
r
th
is
task
d
o
m
ain
.
Similar
ly
,
Kan
an
et
a
l.
[
7
]
d
ev
elo
p
ed
a
n
in
tellig
en
t
h
ea
lth
ca
r
e
s
y
s
tem
b
ased
o
n
l
o
w
-
r
eso
u
r
ce
lan
g
u
ag
e
p
r
o
ce
s
s
in
g
an
d
d
em
o
n
s
tr
ated
t
h
e
ef
f
ec
tiv
en
e
s
s
o
f
m
ac
h
in
e
lear
n
in
g
m
o
d
e
ls
f
o
r
d
r
u
g
ab
u
s
e
d
etec
tio
n
i
n
s
o
cial
m
ed
ia.
I
n
co
n
tr
ast,
T
ass
o
n
e
et
a
l.
[
2
0
]
f
o
u
n
d
th
at
SVM
ac
h
iev
ed
o
n
ly
5
9
.
3
3
%
ac
cu
r
ac
y
o
n
a
b
r
o
ad
er
d
r
u
g
u
s
e
d
etec
tio
n
task
u
s
in
g
E
n
g
lis
h
p
o
s
ts
,
h
ig
h
lig
h
tin
g
th
at
task
-
s
p
e
cif
ic
k
ey
wo
r
d
tar
g
etin
g
a
n
d
d
o
m
ai
n
-
f
o
c
u
s
ed
p
r
ep
r
o
ce
s
s
in
g
,
as e
m
p
lo
y
e
d
in
th
is
s
tu
d
y
,
m
ay
s
u
b
s
tan
tially
im
p
r
o
v
e
class
if
icatio
n
p
er
f
o
r
m
an
ce
.
T
h
is
s
t
u
d
y
als
o
c
o
n
tr
ib
u
t
es
a
n
o
v
el
m
et
h
o
d
o
l
o
g
i
ca
l
a
d
v
a
n
ce
m
e
n
t
b
y
i
n
c
o
r
p
o
r
a
ti
n
g
P
I
I
f
e
at
u
r
es
(
p
h
o
n
e
n
u
m
b
e
r
s
a
n
d
we
b
UR
L
s
)
in
to
t
h
e
c
lass
i
f
i
ca
t
io
n
p
i
p
el
in
e.
W
h
i
le
p
r
e
v
i
o
u
s
s
t
u
d
ies
[
8
]
,
[
1
3
]
,
[
1
7
]
p
r
i
m
ar
ily
r
elied
o
n
tex
tu
al
an
d
lin
g
u
is
tic
f
e
atu
r
es
f
o
r
d
etec
tin
g
d
r
u
g
-
r
el
ated
co
n
ten
t,
o
u
r
f
ea
tu
r
e
i
m
p
o
r
tan
ce
an
aly
s
is
d
em
o
n
s
tr
ates
th
at
PII
elem
e
n
ts
r
an
k
am
o
n
g
t
h
e
to
p
f
iv
e
m
o
s
t
in
f
lu
en
tial
f
ea
tu
r
es.
T
h
is
f
in
d
in
g
s
u
g
g
ests
th
at
s
eller
s
o
f
illeg
al
d
r
u
g
s
o
n
I
n
d
o
n
esian
X
f
r
e
q
u
en
tly
i
n
clu
d
e
co
n
tact
in
f
o
r
m
atio
n
to
f
ac
i
litate
tr
an
s
ac
tio
n
s
,
m
ak
in
g
PII
d
etec
tio
n
a
v
alu
ab
le
co
m
p
lem
en
t
to
tr
a
d
itio
n
al
NL
P
f
ea
tu
r
es.
Mo
r
eo
v
er
,
th
e
PII
-
en
h
a
n
ce
d
f
r
am
ewo
r
k
is
g
en
er
aliza
b
le:
p
h
o
n
e
n
u
m
b
e
r
s
an
d
UR
L
s
ar
e
l
an
g
u
ag
e
-
ag
n
o
s
tic
s
ig
n
als,
allo
win
g
th
is
ap
p
r
o
ac
h
to
b
e
ad
ap
ted
t
o
o
th
er
r
eg
io
n
s
an
d
lan
g
u
a
g
es
with
m
in
im
al
m
o
d
if
icatio
n
to
th
e
r
eg
ex
p
a
tter
n
s
.
Fu
tu
r
e
wo
r
k
co
u
ld
ex
p
lo
r
e
co
m
b
i
n
in
g
o
u
r
PII
-
en
h
a
n
ce
d
f
ea
tu
r
e
e
n
g
in
e
er
in
g
with
tr
a
n
s
f
o
r
m
er
-
b
ased
m
o
d
els
to
f
u
r
th
e
r
im
p
r
o
v
e
d
etec
tio
n
ac
c
u
r
ac
y
.
B
ey
o
n
d
tech
n
ical
p
e
r
f
o
r
m
an
ce
,
th
e
d
ep
lo
y
m
en
t
o
f
th
is
d
etec
tio
n
m
o
d
el
r
e
q
u
ir
es
ca
r
e
f
u
l
eth
ical
co
n
s
id
er
atio
n
s
.
As
s
h
o
wn
in
th
e
er
r
o
r
an
aly
s
is
,
f
alse
p
o
s
itiv
es
m
ay
in
co
r
r
ec
tly
f
la
g
p
o
s
ts
d
is
cu
s
s
in
g
d
r
u
g
s
in
n
o
n
-
co
m
m
er
cial
co
n
tex
ts
s
u
ch
as
p
er
s
o
n
al
ex
p
er
ien
ce
s
o
r
h
ar
m
-
r
e
d
u
ctio
n
d
is
co
u
r
s
e.
T
h
e
s
en
s
itiv
e
n
atu
r
e
o
f
ce
r
tain
ca
teg
o
r
ies,
p
ar
ticu
lar
ly
ab
o
r
tio
n
p
ills
,
also
in
ter
s
ec
ts
with
r
ep
r
o
d
u
ctiv
e
h
ea
lth
an
d
p
er
s
o
n
al
s
af
ety
.
T
h
er
ef
o
r
e,
t
h
e
m
o
d
el
is
r
ec
o
m
m
en
d
ed
f
o
r
u
s
e
as
a
d
ec
is
io
n
-
s
u
p
p
o
r
t
to
o
l
f
o
r
h
u
m
an
m
o
d
e
r
ato
r
s
r
ath
er
th
a
n
as
a
s
tan
d
alo
n
e
e
n
f
o
r
ce
m
e
n
t
m
ec
h
a
n
is
m
,
with
tr
an
s
p
ar
en
t
o
v
er
s
ig
h
t
an
d
co
m
p
lian
ce
with
d
ata
p
r
o
tectio
n
r
eg
u
latio
n
s
.
4.
CO
NCLU
SI
O
N
T
h
e
s
tu
d
y
ex
am
in
e
d
p
o
s
ts
r
el
ated
to
illeg
al
d
r
u
g
s
ales.
Utilizin
g
m
ac
h
in
e
lear
n
in
g
tech
n
iq
u
es,
th
e
r
esear
ch
er
s
an
aly
ze
d
1
1
,
7
7
7
p
o
s
ts
.
Usi
n
g
n
aïv
e
B
ay
es,
lo
g
is
tic
r
eg
r
ess
io
n
,
r
an
d
o
m
f
o
r
est
,
g
r
ad
ien
t
b
o
o
s
tin
g
,
an
d
SVM
class
if
ier
s
to
id
en
tify
p
o
s
ts
in
d
icatin
g
in
te
n
t
to
s
ell
illeg
al
d
r
u
g
s
.
T
h
e
r
esu
lts
s
h
o
wed
th
at
th
e
SVM
m
o
d
el
p
er
f
o
r
m
ed
b
est,
ac
h
iev
in
g
an
F1
-
s
co
r
e
o
f
9
7
.
0
8
%.
T
h
ese
r
esu
lts
s
u
g
g
est
th
at
th
e
SVM
alg
o
r
ith
m
is
m
o
r
e
s
u
itab
le
f
o
r
f
u
tu
r
e
task
s
o
n
th
is
d
ataset.
T
h
e
f
ea
tu
r
e
im
p
o
r
tan
ce
an
aly
s
is
d
em
o
n
s
tr
ated
th
at
PII
elem
e
n
ts
wer
e
cr
itical
in
en
h
an
cin
g
th
e
ac
cu
r
ac
y
o
f
t
h
e
class
if
icatio
n
m
o
d
els.
T
h
ese
f
in
d
in
g
s
s
u
g
g
est
th
at
in
teg
r
atin
g
PII
d
etec
tio
n
in
to
th
e
an
aly
s
is
ca
n
g
r
ea
tly
e
n
h
an
ce
t
h
e
ef
f
ec
tiv
en
ess
o
f
m
o
n
ito
r
i
n
g
an
d
d
e
tectin
g
illi
cit
d
r
u
g
s
ales
ac
tiv
itie
s
o
n
lin
e.
T
h
is
s
tu
d
y
an
aly
ze
s
th
at
p
o
s
ts
m
en
tio
n
in
g
ab
o
r
tio
n
p
ills
wer
e
th
e
m
o
s
t
co
m
m
o
n
,
f
o
llo
wed
b
y
t
h
o
s
e
ab
o
u
t
s
ex
u
al
en
h
an
ce
m
en
t
d
r
u
g
s
an
d
an
esth
etic
d
r
u
g
s
.
T
h
er
e
was
a
n
o
tab
le
in
cr
ea
s
e
in
p
o
s
ts
r
elate
d
to
illeg
al
d
r
u
g
s
ales
in
Dec
em
b
er
2
0
2
3
,
s
u
g
g
e
s
tin
g
m
o
r
e
ac
tiv
ity
d
u
r
i
n
g
th
a
t
p
er
io
d
.
T
h
e
s
tu
d
y
in
clu
d
ed
v
is
u
aliza
tio
n
s
to
s
h
o
w
th
e
d
ata,
s
u
ch
as
th
e
d
is
tr
ib
u
tio
n
o
f
p
o
s
ts
o
v
er
tim
e,
th
e
n
u
m
b
er
o
f
p
o
s
ts
b
y
d
r
u
g
t
y
p
e,
a
n
d
a
wo
r
d
clo
u
d
o
f
th
e
m
o
s
t
u
s
ed
h
ash
tag
s
.
T
h
e
s
e
v
is
u
aliza
tio
n
s
s
h
o
wed
tr
en
d
s
in
th
e
d
ata,
with
ab
o
r
tio
n
p
ills
b
ein
g
th
e
m
o
s
t
d
is
cu
s
s
ed
,
esp
ec
ially
th
e
“
cy
to
t
ec
”
.
T
h
is
s
tu
d
y
h
as
s
o
m
e
lim
itatio
n
s
.
T
h
e
d
ata
m
ig
h
t
n
o
t
f
u
lly
r
e
f
lect
th
e
p
u
b
lic
b
ec
au
s
e
it
f
o
cu
s
ed
o
n
s
p
ec
if
ic
p
o
s
ts
an
d
to
p
ics.
O
n
ly
p
o
s
ts
in
th
e
I
n
d
o
n
esian
lan
g
u
ag
e
wer
e
in
clu
d
e
d
,
wh
ic
h
m
ig
h
t
m
is
s
r
elev
an
t
d
ata
in
o
th
er
lan
g
u
ag
es.
Fu
tu
r
e
r
esear
ch
s
h
o
u
ld
lo
o
k
at
a
wid
er
r
an
g
e
o
f
to
p
ics
an
d
in
c
lu
d
e
d
ata
f
r
o
m
o
th
er
s
o
cial
m
ed
ia
p
latf
o
r
m
s
,
f
o
r
u
m
s
,
an
d
co
m
m
en
t
s
ec
tio
n
s
.
T
h
is
wo
u
ld
g
iv
e
a
b
etter
u
n
d
er
s
tan
d
in
g
o
f
p
u
b
lic
o
p
in
io
n
.
Stu
d
y
in
g
d
ata
in
d
if
f
er
e
n
t
lan
g
u
ag
es
an
d
r
eg
io
n
s
co
u
ld
p
r
o
v
id
e
a
m
o
r
e
g
l
o
b
al
v
iew.
Ad
d
in
g
d
ata
f
r
o
m
n
ews
r
ep
o
r
ts
an
d
law
en
f
o
r
ce
m
e
n
t
co
u
ld
h
elp
co
n
f
ir
m
th
e
f
in
d
in
g
s
an
d
g
iv
e
m
o
r
e
in
s
ig
h
ts
in
to
illeg
al
d
r
u
g
s
ales
o
n
s
o
cial
m
ed
ia.
T
h
e
r
esu
lts
o
f
th
is
s
tu
d
y
ca
n
b
e
ap
p
lied
in
s
ev
er
al
r
ea
l
-
wo
r
ld
co
n
tex
ts
,
p
ar
ticu
lar
ly
f
o
r
law
en
f
o
r
ce
m
en
t
an
d
p
u
b
lic
h
ea
lth
ag
en
cies.
Fo
r
in
s
tan
ce
,
th
e
class
if
icatio
n
m
o
d
el
co
u
ld
b
e
in
teg
r
ate
d
in
to
a
m
o
n
ito
r
in
g
s
y
s
tem
th
at
f
lag
s
p
o
ten
tial illeg
al
d
r
u
g
s
ale
p
o
s
ts
in
r
ea
l
-
tim
e,
allo
win
g
f
o
r
f
aster
r
esp
o
n
s
e
a
n
d
in
t
er
v
en
tio
n
.
Pu
b
lic
h
ea
lth
ca
m
p
aig
n
s
ca
n
also
b
e
tailo
r
ed
to
ad
d
r
ess
th
e
r
is
e
i
n
d
is
cu
s
s
io
n
s
ab
o
u
t
ce
r
tai
n
d
r
u
g
s
,
s
u
ch
as
a
b
o
r
tio
n
p
ills
,
p
r
o
v
id
i
n
g
tar
g
eted
in
f
o
r
m
atio
n
an
d
r
eso
u
r
ce
s
to
c
o
m
b
at
th
eir
m
is
u
s
e.
Evaluation Warning : The document was created with Spire.PDF for Python.