I
AE
S In
t
er
na
t
io
na
l J
o
urna
l o
f
Ro
bo
t
ics a
nd
Aut
o
m
a
t
io
n
(
I
J
RA)
Vo
l.
1
4
,
No
.
1
,
Ma
r
ch
20
2
5
,
p
p
.
47
~
57
I
SS
N:
2722
-
2
5
8
6
,
DOI
:
1
0
.
1
1
5
9
1
/i
jr
a
.
v
1
4
i
1
.
pp
47
-
57
47
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
r
a
.
ia
esco
r
e.
co
m
Integ
ra
tion o
f
na
t
ura
l lang
ua
g
e process
ing
methods
a
nd
ma
chine learning
mo
del f
o
r ma
licio
us webpa
g
e de
tec
tion ba
sed
o
n web
co
n
tent
s
Sh
a
heet
ha
L
ia
qu
a
t
ha
li,
Va
di
v
a
zha
g
a
n K
a
dirv
elu
D
e
p
a
r
t
me
n
t
o
f
C
o
mp
u
t
e
r
a
n
d
I
n
f
o
r
ma
t
i
o
n
S
c
i
e
n
c
e
,
A
n
n
a
ma
l
a
i
U
n
i
v
e
r
si
t
y
,
Ta
mi
l
N
a
d
u
,
I
n
d
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Au
g
1
,
2
0
2
4
R
ev
is
ed
Sep
8
,
2
0
2
4
Acc
ep
ted
Oct
6
,
2
0
2
4
M
a
li
c
io
u
s
a
c
to
rs
c
o
n
ti
n
u
a
ll
y
e
x
p
l
o
it
v
u
l
n
e
ra
b
il
it
ies
i
n
we
b
s
y
ste
m
s
to
d
istri
b
u
te
m
a
lwa
re
,
lau
n
c
h
p
h
ish
i
n
g
a
tt
a
c
k
s,
ste
a
l
se
n
sit
iv
e
i
n
fo
rm
a
ti
o
n
,
a
n
d
p
e
rp
e
trate
v
a
rio
u
s
f
o
rm
s
o
f
c
y
b
e
r
c
rime
.
Trad
it
io
n
a
l
sig
n
a
tu
re
-
b
a
se
d
m
e
th
o
d
s
fo
r
d
e
tec
ti
n
g
m
a
li
c
io
u
s
we
b
p
a
g
e
s
o
ften
stru
g
g
le
to
k
e
e
p
p
a
c
e
with
th
e
ra
p
id
e
v
o
lu
ti
o
n
o
f
m
a
lwa
re
a
n
d
c
y
b
e
r
th
re
a
ts.
As
a
re
su
lt
,
th
e
re
is
a
g
ro
win
g
d
e
m
a
n
d
f
o
r
m
o
re
a
d
v
a
n
c
e
d
a
n
d
p
ro
a
c
ti
v
e
a
p
p
ro
a
c
h
e
s
t
h
a
t
c
a
n
e
ffe
c
ti
v
e
ly
id
e
n
ti
f
y
m
a
li
c
io
u
s
we
b
c
o
n
ten
t
b
a
se
d
o
n
it
s
c
h
a
ra
c
teristics
a
n
d
b
e
h
a
v
io
r.
De
tec
ti
o
n
b
a
se
d
o
n
we
b
c
o
n
ten
t
is
c
ru
c
ial
b
e
c
a
u
se
m
a
li
c
io
u
s
we
b
p
a
g
e
s
c
a
n
b
e
d
e
sig
n
e
d
to
m
imi
c
leg
it
ima
t
e
o
n
e
s,
m
a
k
i
n
g
t
h
e
m
d
iffi
c
u
lt
t
o
i
d
e
n
ti
f
y
th
ro
u
g
h
trad
it
i
o
n
a
l
m
e
a
n
s.
By
a
n
a
ly
z
in
g
th
e
c
o
n
ten
t
o
f
we
b
p
a
g
e
s,
it
b
e
c
o
m
e
s
p
o
ss
ib
le
to
u
n
c
o
v
e
r
p
a
tt
e
rn
s,
a
n
o
m
a
li
e
s,
a
n
d
m
a
li
c
io
u
s in
ten
t
t
h
a
t
m
a
y
n
o
t
b
e
e
v
id
e
n
t
fro
m
s
u
rfa
c
e
-
lev
e
l
in
sp
e
c
ti
o
n
.
Th
e
p
r
o
p
o
se
d
a
p
p
ro
a
c
h
in
teg
ra
tes
a
p
re
train
e
d
Wo
r
d
2
Ve
c
m
o
d
e
l
wit
h
se
v
e
n
d
isti
n
c
t
m
a
c
h
in
e
lea
rn
in
g
c
las
sifiers
to
e
n
h
a
n
c
e
m
a
li
c
io
u
s
we
b
p
a
g
e
d
e
tec
ti
o
n
.
In
i
ti
a
ll
y
,
we
b
c
o
n
ten
ts
(d
o
c
u
m
e
n
ts)
a
re
e
n
c
o
d
e
d
u
sin
g
th
e
W
o
rd
2
Ve
c
m
o
d
e
l,
f
o
ll
o
we
d
b
y
th
e
c
o
m
p
u
tatio
n
o
f
a
v
e
ra
g
e
Wo
r
d
2
Ve
c
e
m
b
e
d
d
in
g
s
fo
r
e
a
c
h
d
o
c
u
m
e
n
t.
S
u
b
se
q
u
e
n
tl
y
,
e
a
c
h
c
las
sifier
is
train
e
d
o
n
th
e
e
x
t
ra
c
ted
a
v
e
ra
g
e
Wo
r
d
2
Ve
c
e
m
b
e
d
d
in
g
fe
a
tu
re
s.
Th
e
re
su
lt
s d
e
m
o
n
stra
te t
h
a
t
th
e
Wo
rd
2
Ve
c
m
o
d
e
l
si
g
n
ifi
c
a
n
tl
y
e
n
h
a
n
c
e
s
th
e
d
e
tec
ti
o
n
a
c
c
u
ra
c
y
,
a
c
h
ie
v
in
g
a
n
a
c
c
u
ra
c
y
o
f
9
4
.
8
%
a
n
d
a
n
F
1
-
sc
o
re
o
f
9
4
.
9
%
wit
h
t
h
e
ra
n
d
o
m
fo
re
st
c
l
a
ss
ifi
e
r,
a
n
d
a
n
a
c
c
u
ra
c
y
o
f
9
4
.
6
%
a
n
d
a
n
F1
-
sc
o
re
o
f
9
4
.
7
%
wit
h
t
h
e
e
x
tre
m
e
g
ra
d
ien
t
b
o
o
stin
g
c
las
sifier.
K
ey
w
o
r
d
s
:
C
y
b
er
cr
im
e
E
n
s
em
b
le
Ma
ch
in
e
lear
n
in
g
Ma
licio
u
s
web
p
ag
e
d
etec
tio
n
Ph
is
h
in
g
Stack
in
g
W
o
r
d
2
Vec
em
b
ed
d
in
g
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
:
Sh
ah
ee
th
a
L
iaq
u
ath
ali
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
a
n
d
I
n
f
o
r
m
atio
n
Scien
ce
,
An
n
am
a
lai
Un
iv
er
s
ity
T
am
il Na
d
u
,
I
n
d
ia
E
m
ail: sh
ah
ee
.
aa
s
c@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
I
n
o
u
r
m
o
d
er
n
d
ig
ital
lan
d
s
ca
p
e,
th
e
in
ter
n
et
p
lay
s
an
in
d
is
p
en
s
ab
le
r
o
le
in
co
m
m
u
n
icatio
n
,
co
m
m
er
ce
,
an
d
in
f
o
r
m
atio
n
s
h
ar
in
g
.
Ho
wev
e
r
,
alo
n
g
s
id
e
i
ts
m
y
r
iad
ad
v
an
tag
es,
th
e
in
t
er
n
et
also
p
r
esen
ts
s
ig
n
if
ican
t
cy
b
er
s
ec
u
r
ity
r
is
k
s
,
p
ar
ticu
lar
ly
th
r
o
u
g
h
m
alicio
u
s
web
p
ag
es
[
1
]
.
T
h
ese
web
p
ag
es
h
ar
b
o
r
v
ar
i
o
u
s
f
o
r
m
s
o
f
h
ar
m
f
u
l
co
n
ten
t,
in
c
lu
d
in
g
m
alwa
r
e
a
n
d
p
h
is
h
in
g
s
ca
m
s
,
d
esig
n
ed
to
d
ec
eiv
e
u
s
er
s
,
co
m
p
r
o
m
is
e
th
eir
p
r
i
v
ac
y
,
o
r
in
f
lict
d
a
m
a
g
e
o
n
th
eir
d
ev
ices
[
2
]
.
T
r
a
d
itio
n
al
m
eth
o
d
s
o
f
d
etec
tin
g
m
alicio
u
s
web
p
ag
es
o
f
ten
r
ely
o
n
UR
L
-
b
ased
b
lac
k
lis
tin
g
,
h
eu
r
is
tic
ap
p
r
o
ac
h
es,
o
r
UR
L
-
b
ased
f
ea
tu
r
e
ex
tr
ac
ti
o
n
.
Ho
wev
er
,
t
h
ese
m
eth
o
d
s
h
a
v
e
lim
itatio
n
s
in
a
cc
u
r
ately
id
e
n
tify
in
g
an
d
p
r
ev
en
tin
g
ac
ce
s
s
to
all
ty
p
es
o
f
m
alicio
u
s
web
p
ag
es
[
3
]
.
UR
L
-
b
ased
b
lack
lis
tin
g
in
v
o
lv
es
m
ain
tain
i
n
g
a
d
atab
ase
o
f
k
n
o
wn
m
alicio
u
s
UR
L
s
an
d
b
lo
c
k
in
g
ac
ce
s
s
to
web
s
ites
with
m
atch
in
g
UR
L
s
.
W
h
ile
ef
f
ec
tiv
e
t
o
s
o
m
e
e
x
ten
t,
th
is
a
p
p
r
o
ac
h
f
ails
to
ad
d
r
ess
n
ewly
cr
ea
te
d
m
alicio
u
s
web
p
ag
es
o
r
th
o
s
e
th
at
d
y
n
am
ically
g
en
er
ate
U
R
L
s
.
Mo
r
eo
v
er
,
cy
b
er
cr
im
i
n
a
ls
ca
n
ea
s
i
ly
ev
ad
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
7
2
2
-
2
5
8
6
I
AE
S
I
n
t
J
R
o
b
&
A
u
to
m
,
Vo
l
.
1
4
,
No
.
1
,
Ma
r
ch
20
2
5
:
47
-
57
48
d
etec
tio
n
b
y
f
r
eq
u
en
tly
ch
a
n
g
in
g
UR
L
s
o
r
u
s
in
g
UR
L
o
b
f
u
s
ca
tio
n
tech
n
iq
u
es.
Heu
r
is
tic
ap
p
r
o
ac
h
es
r
ely
o
n
p
r
ed
ef
in
e
d
r
u
les
o
r
p
atter
n
s
to
id
en
tify
p
o
ten
tially
m
alicio
u
s
web
p
ag
es
b
ased
o
n
ch
ar
ac
ter
is
tics
s
u
ch
as
U
R
L
s
tr
u
ctu
r
e,
HT
ML
co
n
te
n
t,
o
r
b
eh
av
io
r
.
W
h
ile
h
eu
r
is
tic
m
eth
o
d
s
ca
n
d
etec
t
s
o
m
e
ty
p
es
o
f
m
alicio
u
s
ac
tiv
ity
,
th
ey
o
f
te
n
s
u
f
f
e
r
f
r
o
m
h
ig
h
f
al
s
e
p
o
s
itiv
e
r
ates
an
d
m
ay
o
v
er
lo
o
k
s
o
p
h
is
ticated
th
r
ea
ts
th
at
d
o
n
o
t
co
n
f
o
r
m
to
p
r
ed
ef
in
e
d
p
atter
n
s
[
4
]
.
UR
L
-
b
ased
f
ea
t
u
r
e
e
x
tr
ac
tio
n
in
v
o
lv
es
an
aly
zin
g
f
ea
tu
r
es
d
er
i
v
ed
f
r
o
m
th
e
UR
L
its
elf
,
s
u
ch
as
d
o
m
ain
r
ep
u
tatio
n
,
UR
L
le
n
g
th
,
o
r
th
e
p
r
esen
ce
o
f
s
p
ec
if
ic
k
ey
wo
r
d
s
.
W
h
ile
th
is
m
eth
o
d
ca
n
p
r
o
v
id
e
s
o
m
e
in
s
ig
h
t
in
to
th
e
n
atu
r
e
o
f
a
web
p
a
g
e,
it
m
ay
n
o
t
ca
p
tu
r
e
th
e
f
u
ll
co
n
tex
t
o
r
s
em
an
tic
m
ea
n
in
g
o
f
th
e
co
n
ten
t
h
o
s
ted
o
n
th
e
web
p
ag
e
[
5
]
.
T
h
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
,
th
e
w
eb
c
o
n
ten
t
an
aly
s
is
with
th
e
W
o
r
d
2
Vec
mod
el
(
W
C
A
-
W
2
V
)
,
ad
d
r
ess
es
th
ese
lim
itatio
n
s
b
y
lev
er
ag
in
g
th
e
ac
t
u
al
co
n
ten
t
o
f
web
p
ag
es
f
o
r
m
alicio
u
s
web
p
ag
e
d
etec
tio
n
.
I
n
p
ar
ticu
lar
,
th
e
tex
t
u
al
co
n
ten
t
s
(
d
o
cu
m
en
ts
)
ar
e
r
etr
iev
ed
f
r
o
m
p
ar
a
g
r
ap
h
,
d
iv
is
io
n
,
a
n
d
m
eta
tag
s
em
b
ed
d
e
d
with
in
web
p
ag
es.
T
h
ese
co
n
te
n
ts
ar
e
th
en
s
u
b
jecte
d
to
W
o
r
d
2
Vec
em
b
ed
d
i
n
g
tech
n
i
q
u
es
to
en
ca
p
s
u
late
th
e
m
in
to
d
en
s
e
v
ec
to
r
r
e
p
r
esen
tatio
n
s
[
6
]
.
Su
b
s
eq
u
en
tly
,
th
e
av
er
ag
e
W
o
r
d
2
Vec
em
b
e
d
d
in
g
s
ar
e
co
m
p
u
ted
f
o
r
ea
ch
d
o
cu
m
e
n
t,
en
ab
lin
g
a
m
o
r
e
co
m
p
ac
t
an
d
s
em
an
tically
en
r
ich
ed
r
ep
r
esen
tatio
n
o
f
th
e
web
p
ag
e'
s
tex
tu
al
co
n
ten
t.
B
y
ca
p
tu
r
in
g
s
em
a
n
tic
r
elatio
n
s
h
ip
s
b
etwe
en
wo
r
d
s
,
th
e
ap
p
r
o
ac
h
en
h
an
c
es
th
e
ac
cu
r
ac
y
o
f
m
alicio
u
s
web
p
ag
e
d
etec
tio
n
.
Un
lik
e
UR
L
-
b
ased
m
eth
o
d
s
,
wh
ich
f
o
c
u
s
s
o
lely
o
n
th
e
ad
d
r
ess
(
UR
L
)
o
f
th
e
web
p
ag
e,
th
e
p
r
o
p
o
s
ed
a
p
p
r
o
ac
h
co
n
s
id
er
s
th
e
ac
tu
al
co
n
t
en
t
(
tex
tu
al)
h
o
s
ted
o
n
th
e
w
eb
p
ag
e.
T
h
is
allo
ws
th
e
d
etec
tio
n
o
f
m
alicio
u
s
co
n
ten
t
r
eg
a
r
d
less
o
f
ch
an
g
es
to
UR
L
s
o
r
ev
asio
n
tactics
em
p
lo
y
ed
b
y
cy
b
er
cr
im
in
als.
R
e
c
e
n
t
r
es
e
a
r
c
h
i
n
m
a
li
c
i
o
u
s
w
e
b
p
a
g
e
d
e
t
e
c
t
i
o
n
h
as
e
x
p
l
o
r
e
d
v
a
r
i
o
u
s
v
e
c
t
o
r
i
z
a
ti
o
n
m
e
t
h
o
d
s
i
n
n
a
t
u
r
al
l
a
n
g
u
a
g
e
p
r
o
c
e
s
s
i
n
g
,
i
n
cl
u
d
i
n
g
c
o
u
n
t
,
T
F
-
I
D
F
,
a
n
d
h
a
s
h
i
n
g
v
e
c
t
o
r
i
z
e
r
s
[
7
]
.
H
o
w
e
v
e
r
,
t
h
e
p
r
o
p
o
s
e
d
a
p
p
r
o
a
c
h
d
i
v
e
r
g
e
s
b
y
u
t
i
l
i
z
i
n
g
W
o
r
d
2
V
e
c
e
m
b
e
d
d
i
n
g
,
w
h
i
c
h
c
a
p
t
u
r
e
s
s
e
m
a
n
ti
c
r
e
l
at
i
o
n
s
h
i
p
s
b
e
tw
e
e
n
w
o
r
d
s
.
U
n
l
i
k
e
c
o
u
n
t
-
b
a
s
e
d
m
e
t
h
o
d
s
,
w
h
ic
h
r
e
l
y
s
o
l
e
l
y
o
n
w
o
r
d
o
c
c
u
r
r
e
n
c
es
[
8
]
,
W
o
r
d
2
V
e
c
e
m
b
e
d
d
i
n
g
s
r
e
p
r
e
s
e
n
t
w
o
r
d
s
i
n
a
h
i
g
h
-
d
i
m
e
n
s
i
o
n
a
l
v
e
c
t
o
r
s
p
ac
e
,
e
n
a
b
l
i
n
g
a
m
o
r
e
s
u
b
t
l
e
a
n
d
c
o
n
t
e
x
t
u
a
l
l
y
m
e
a
n
i
n
g
f
u
l
r
ep
r
e
s
e
n
t
a
ti
o
n
o
f
w
e
b
c
o
n
t
e
n
t
.
S
i
m
il
a
r
l
y
,
w
h
i
le
T
F
-
I
D
F
a
s
s
i
g
n
s
we
i
g
h
ts
b
a
s
e
d
o
n
w
o
r
d
f
r
e
q
u
e
n
c
y
,
i
t
l
a
c
k
s
i
n
h
e
r
e
n
t
s
e
m
a
n
t
i
c
u
n
d
e
r
s
t
a
n
d
i
n
g
[
9
]
,
u
n
l
i
k
e
W
o
r
d
2
V
e
c
e
m
b
e
d
d
i
n
g
s
w
h
i
c
h
e
n
c
o
d
e
s
e
m
a
n
t
i
c
i
n
f
o
r
m
a
t
i
o
n
,
t
h
er
e
b
y
e
n
h
a
n
c
i
n
g
t
h
e
a
c
c
u
r
a
c
y
o
f
m
a
l
i
c
i
o
u
s
c
o
n
t
e
n
t
d
e
t
e
c
t
i
o
n
.
M
o
r
e
o
v
e
r
,
u
n
l
i
k
e
h
as
h
i
n
g
t
e
c
h
n
i
q
u
e
s
t
h
a
t
c
o
n
v
e
r
t
te
x
t
i
n
t
o
f
i
x
e
d
-
l
e
n
g
t
h
h
a
s
h
c
o
d
e
s
[
1
0
]
,
W
o
r
d
2
V
e
c
e
m
b
e
d
d
i
n
g
s
p
r
e
s
e
r
v
e
s
e
m
a
n
ti
c
r
e
l
a
t
i
o
n
s
h
i
p
s
,
f
a
c
il
i
ta
t
i
n
g
a
d
e
e
p
e
r
u
n
d
e
r
s
t
a
n
d
i
n
g
o
f
w
e
b
c
o
n
t
e
n
t
s
e
m
a
n
t
i
cs
a
n
d
e
n
ab
l
i
n
g
t
h
e
d
e
t
e
c
ti
o
n
o
f
s
u
b
t
l
e
p
at
t
e
r
n
s
i
n
d
ic
a
t
i
v
e
o
f
m
al
i
c
i
o
u
s
in
t
e
n
t
[
1
1
]
.
T
h
e
m
ajo
r
c
o
n
tr
ib
u
tio
n
s
o
f
th
is
wo
r
k
ar
e
i)
l
e
v
e
r
a
g
i
n
g
a
c
t
u
a
l
w
e
b
p
a
g
e
c
o
n
t
e
n
t
f
o
r
m
a
l
i
c
i
o
u
s
w
e
b
p
a
g
e
d
e
t
e
c
ti
o
n
,
i
i
)
e
x
t
r
ac
t
i
o
n
o
f
c
o
n
t
e
n
t
f
r
o
m
p
a
r
a
g
r
a
p
h
,
d
i
v
i
s
i
o
n
,
a
n
d
m
e
t
a
t
a
g
s
f
o
r
a
n
a
l
y
s
i
s
,
i
i
i
)
u
t
il
i
z
at
i
o
n
o
f
A
v
e
r
a
g
e
W
o
r
d
2
V
e
c
e
m
b
e
d
d
i
n
g
t
e
c
h
n
i
q
u
e
s
f
o
r
d
e
n
s
e
a
n
d
m
e
a
n
i
n
g
f
u
l
r
e
p
r
e
s
e
n
t
a
ti
o
n
o
f
w
e
b
p
a
g
e
c
o
n
t
e
n
t
,
i
v
)
e
n
h
a
n
c
em
e
n
t
o
f
d
e
t
e
ct
i
o
n
a
c
c
u
r
ac
y
th
r
o
u
g
h
c
a
p
t
u
r
i
n
g
s
e
m
a
n
t
i
c
r
e
la
t
i
o
n
s
h
i
p
s
b
e
tw
e
e
n
w
o
r
d
s
,
v
)
c
o
n
s
i
d
e
r
a
t
i
o
n
o
f
w
eb
p
a
g
e
c
o
n
t
e
n
t
r
a
t
h
e
r
t
h
a
n
s
o
l
el
y
r
e
l
y
i
n
g
o
n
UR
L
-
b
as
e
d
m
eth
o
d
s
,
v
i
)
a
b
i
li
t
y
t
o
d
e
t
e
c
t
m
al
i
c
i
o
u
s
c
o
n
t
e
n
t
d
e
s
p
it
e
c
h
a
n
g
e
s
t
o
UR
L
s
o
r
e
v
as
i
o
n
t
a
c
t
ic
s
,
a
n
d
v
i
i
)
i
n
t
e
g
r
a
t
i
o
n
o
f
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
c
l
a
s
s
i
f
ie
r
s
wi
t
h
A
v
e
r
a
g
e
W
o
r
d
2
V
e
c
e
m
b
e
d
d
i
n
g
f
o
r
i
m
p
r
o
v
e
d
d
e
t
e
c
t
i
o
n
c
a
p
a
b
il
i
ti
e
s
I
n
th
is
p
ap
e
r
,
we
aim
t
o
en
h
a
n
ce
th
e
d
etec
tio
n
ac
cu
r
ac
y
o
f
m
alicio
u
s
web
p
ag
es
u
tili
zin
g
ad
v
an
ce
d
m
ac
h
in
e
lear
n
i
n
g
tech
n
iq
u
es
a
n
d
co
n
ten
t
an
aly
s
is
.
Ou
r
r
esear
ch
o
b
jectiv
e
is
to
d
e
v
elo
p
a
p
r
o
ac
tiv
e
ap
p
r
o
ac
h
th
at
ef
f
ec
tiv
ely
i
d
en
tifie
s
m
al
icio
u
s
web
co
n
te
n
t
b
ased
o
n
its
ch
ar
ac
ter
is
tics
an
d
b
eh
av
i
o
r
.
Sp
ec
if
ically
,
we
f
o
cu
s
o
n
in
te
g
r
atin
g
a
p
r
etr
ai
n
ed
W
o
r
d
2
Vec
m
o
d
el
with
m
ac
h
in
e
lear
n
in
g
class
if
ier
s
t
o
en
h
an
ce
m
alicio
u
s
web
p
ag
e
d
etec
tio
n
.
B
y
an
al
y
zin
g
th
e
co
n
ten
t
o
f
web
p
ag
e
s
,
we
aim
to
u
n
co
v
er
p
atter
n
s
,
an
o
m
alies,
an
d
m
alicio
u
s
in
ten
t
th
at
m
ay
n
o
t
b
e
e
v
id
en
t
f
r
o
m
s
u
r
f
ac
e
-
lev
el
in
s
p
ec
tio
n
.
T
h
is
will
co
n
tr
ib
u
te
to
th
e
ad
v
an
ce
m
e
n
t
o
f
cy
b
er
s
ec
u
r
ity
ef
f
o
r
ts
b
y
p
r
o
v
id
in
g
m
o
r
e
r
o
b
u
s
t
an
d
ac
c
u
r
ate
m
eth
o
d
s
f
o
r
d
etec
tin
g
m
alicio
u
s
web
co
n
ten
t.
T
h
e
r
em
ai
n
d
er
o
f
th
e
p
ap
er
is
o
u
tlin
ed
as
f
o
ll
o
ws:
Sectio
n
1
h
ig
h
lig
h
ts
th
e
s
ig
n
if
ican
ce
o
f
th
e
r
esear
ch
wo
r
k
an
d
el
u
cid
ates
its
d
is
tin
ctio
n
s
f
r
o
m
o
th
er
ex
is
tin
g
r
esear
ch
e
n
d
ea
v
o
r
s
.
Sectio
n
2
d
el
v
es
in
to
th
e
ex
is
tin
g
r
esear
ch
wo
r
k
s
with
in
th
e
d
o
m
ain
o
f
m
alicio
u
s
web
p
ag
e
d
etec
tio
n
,
p
r
o
v
i
d
in
g
a
co
m
p
r
eh
e
n
s
iv
e
o
v
er
v
iew
o
f
th
e
cu
r
r
en
t
lan
d
s
ca
p
e.
Sectio
n
3
f
u
r
n
is
h
es
b
a
ck
g
r
o
u
n
d
in
f
o
r
m
atio
n
p
er
tin
e
n
t
to
th
e
r
esear
ch
,
wh
ich
is
ess
en
tial
f
o
r
u
n
d
e
r
s
tan
d
in
g
th
e
p
r
o
p
o
s
ed
a
p
p
r
o
ac
h
.
Sectio
n
4
e
x
p
o
u
n
d
s
u
p
o
n
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
,
d
elin
ea
tin
g
its
m
et
h
o
d
o
lo
g
y
an
d
i
n
tr
icac
ies
in
d
etail.
Sect
io
n
5
p
r
esen
ts
th
e
ex
p
er
i
m
en
tal
r
esu
lts
o
b
tain
ed
f
r
o
m
t
h
e
ap
p
licatio
n
o
f
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
,
s
h
o
wca
s
in
g
its
ef
f
icac
y
a
n
d
p
er
f
o
r
m
an
ce
m
etr
ics.
Fin
ally
,
Sectio
n
6
en
ca
p
s
u
lates th
e
f
in
d
in
g
s
an
d
i
n
s
ig
h
ts
d
er
iv
ed
f
r
o
m
th
e
r
esear
ch
,
p
r
o
v
id
in
g
co
n
clu
d
in
g
r
e
m
ar
k
s
.
2.
RE
L
AT
E
D
WO
RK
S
Dete
ctin
g
m
alicio
u
s
web
s
ites
is
cr
u
cial
f
o
r
s
af
eg
u
a
r
d
in
g
u
s
er
s
f
r
o
m
o
n
li
n
e
r
is
k
s
lik
e
m
alwa
r
e,
p
h
is
h
in
g
,
an
d
d
ata
b
r
ea
ch
es.
R
esear
ch
er
s
h
av
e
d
ev
el
o
p
ed
v
ar
io
u
s
m
eth
o
d
s
to
id
en
tif
y
th
ese
h
ar
m
f
u
l
web
s
ites
.
W
an
et
a
l.
[
1
2
]
o
u
tlin
ed
k
ey
ch
ar
ac
ter
is
tics
o
f
m
alicio
u
s
w
eb
s
ites
,
s
u
ch
as
HT
ML
an
d
J
av
aScr
ip
t
f
ea
tu
r
es.
Me
an
wh
ile,
Ma
lak
et
a
l.
[
1
3
]
an
aly
ze
d
a
d
ataset
c
o
n
tain
in
g
6
6
,
5
0
6
UR
L
s
to
id
en
tif
y
d
a
n
g
er
o
u
s
s
ites
u
s
in
g
m
ac
h
in
e
lear
n
in
g
(
ML
)
an
d
d
ee
p
lear
n
in
g
(
DL
)
m
o
d
els.
T
h
ey
em
p
lo
y
ed
th
r
ee
ty
p
es
o
f
f
ea
tu
r
es
—
lex
ical,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
AE
S
I
n
t
J
R
o
b
&
A
u
to
m
I
SS
N:
2722
-
2
5
8
6
I
n
teg
r
a
tio
n
o
f n
a
t
u
r
a
l la
n
g
u
a
g
e
p
r
o
ce
s
s
in
g
meth
o
d
s
a
n
d
m
a
ch
in
e
lea
r
n
in
g
mo
d
el
… (
S
h
a
h
ee
th
a
Lia
q
u
a
th
a
li
)
49
n
etwo
r
k
-
b
ased
,
an
d
co
n
ten
t
-
b
ased
—
to
d
etec
t
m
alicio
u
s
web
s
ites
.
T
ec
h
n
iq
u
es
lik
e
c
o
r
r
elatio
n
a
n
aly
s
is
,
ANOV
A,
an
d
ch
i
-
s
q
u
ar
e
h
elp
ed
ex
tr
ac
t
th
e
m
o
s
t
s
ig
n
if
ican
t
f
ea
tu
r
es
f
r
o
m
th
e
d
ataset.
T
h
eir
f
in
d
in
g
s
s
h
o
we
d
th
at
th
e
n
aiv
e
B
ay
es
(
NB
)
m
o
d
el
ac
h
iev
e
d
th
e
h
i
g
h
est
ac
cu
r
ac
y
,
d
etec
tin
g
m
alicio
u
s
UR
L
s
w
ith
9
6
%
p
r
ec
is
io
n
.
Sira
g
eld
in
et
a
l.
[
1
4
]
p
r
esen
te
d
a
f
r
am
ewo
r
k
f
o
r
id
en
tify
i
n
g
m
alicio
u
s
web
p
ag
es,
f
o
cu
s
in
g
o
n
UR
L
k
ey
wo
r
d
s
an
d
p
a
g
e
co
n
te
n
t.
T
h
eir
d
ataset
co
n
s
is
ted
o
f
b
o
th
l
eg
itima
te
an
d
h
ar
m
f
u
l w
eb
s
ite
s
f
r
o
m
s
o
u
r
ce
s
lik
e
Alex
a
an
d
Ma
lwar
eURL
.
B
y
s
p
litt
in
g
th
e
f
ea
tu
r
es
in
to
tr
ain
in
g
an
d
test
in
g
s
ets,
th
ey
ac
h
iev
ed
a
9
7
%
ac
cu
r
ac
y
r
ate
with
o
u
t
a
n
y
f
al
s
e
p
o
s
itiv
es.
Saleem
R
aja
et
a
l.
[
1
5
]
also
u
s
ed
ML
an
d
D
L
m
o
d
els
to
class
if
y
h
ar
m
f
u
l
web
s
ites
b
ased
o
n
c
o
n
ten
t,
p
ar
ticu
la
r
ly
f
o
c
u
s
in
g
o
n
HT
ML
tag
s
,
ev
en
t
m
eth
o
d
s
,
DOM
k
ey
wo
r
d
s
,
an
d
J
av
aScr
ip
t
f
u
n
ctio
n
s
.
T
h
ei
r
ex
p
er
im
e
n
ts
,
wh
ich
u
s
ed
th
e
Kag
g
le
d
ataset
an
d
em
p
lo
y
ed
o
v
er
2
0
6
f
ea
tu
r
es,
r
ev
ea
led
th
at
th
e
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
(
SVM)
ac
h
ie
v
ed
8
8
% a
cc
u
r
ac
y
,
wh
ile
th
e
r
a
n
d
o
m
f
o
r
est
(
R
F)
m
o
d
el
r
ea
ch
ed
9
3
%.
Desai
et
a
l.
[
1
6
]
d
e
v
elo
p
ed
a
C
h
r
o
m
e
ex
ten
s
io
n
to
h
elp
u
s
e
r
s
id
en
tify
p
h
is
h
in
g
web
s
ites
,
u
s
in
g
th
e
UC
I
d
ataset
co
n
tain
in
g
1
1
,
0
5
5
r
ec
o
r
d
s
with
3
0
f
ea
tu
r
es,
o
f
wh
ich
o
n
ly
2
2
wer
e
u
s
ed
in
t
h
e
ex
p
er
im
en
t.
T
h
e
r
esu
lts
d
em
o
n
s
tr
ated
th
at
th
e
R
an
d
o
m
Fo
r
est
m
o
d
el
ac
h
i
ev
ed
9
6
%
ac
cu
r
ac
y
,
wh
ile
S
VM
an
d
k
-
n
e
ar
est
n
eig
h
b
o
r
(
k
-
NN
)
m
o
d
els
ac
h
iev
ed
9
3
.
5
%
a
n
d
9
3
%,
r
es
p
ec
tiv
ely
.
Saleem
R
aja
et
a
l.
[
9
]
p
r
o
p
o
s
ed
a
lig
h
tweig
h
t
alg
o
r
ith
m
f
o
r
d
et
ec
tin
g
m
alicio
u
s
UR
L
s
,
ad
d
r
e
s
s
in
g
th
e
lim
itatio
n
s
o
f
b
lack
l
is
t
an
d
r
ep
u
tatio
n
-
b
ased
m
eth
o
d
s
in
d
etec
tin
g
n
e
w
th
r
ea
ts
.
B
y
u
tili
zin
g
lex
ical
f
ea
tu
r
es
s
u
c
h
as
UR
L
len
g
t
h
an
d
th
e
n
u
m
b
er
o
f
s
u
b
d
o
m
ain
s
,
th
eir
ap
p
r
o
ac
h
tr
ain
ed
ML
m
o
d
els
to
id
en
tif
y
h
ar
m
f
u
l
UR
L
s
,
with
k
-
NN
an
d
R
an
d
o
m
Fo
r
est
m
o
d
els ac
h
iev
in
g
9
8
% a
n
d
9
9
% a
cc
u
r
ac
y
,
r
esp
ec
tiv
ely
.
Pra
d
ee
p
a
et
a
l.
[
1
7
]
d
ev
is
ed
a
m
eth
o
d
to
id
en
tify
m
alicio
u
s
UR
L
s
b
ased
o
n
r
an
k
-
b
ased
,
b
ag
-
of
-
wo
r
d
s
,
web
p
a
g
e,
an
d
lex
ical
f
ea
tu
r
es,
u
tili
zin
g
d
atasets
f
r
o
m
Ph
is
tan
k
an
d
Kag
g
le.
T
h
eir
r
esu
lts
s
h
o
wed
th
at
th
e
r
an
d
o
m
f
o
r
est
m
o
d
el
o
f
f
e
r
ed
9
9
%
ac
cu
r
ac
y
.
I
n
an
o
th
e
r
s
tu
d
y
,
Saleem
R
aja
e
t
a
l.
[
1
8
]
ap
p
lied
n
atu
r
al
lan
g
u
ag
e
p
r
o
ce
s
s
in
g
(
NL
P)
to
v
ec
to
r
ize
UR
L
ter
m
s
f
o
r
ML
an
d
DL
m
o
d
el
class
if
ica
tio
n
.
T
h
ey
u
s
ed
two
d
atasets
(
D1
an
d
D2
)
an
d
th
r
e
e
v
ec
to
r
izatio
n
tech
n
iq
u
es
—
co
u
n
t
,
T
F
-
I
DF,
an
d
h
ash
in
g
.
T
h
e
r
esu
lts
in
d
icate
d
th
at
th
e
d
ec
is
io
n
tr
ee
(
DT
)
m
o
d
el
with
a
co
u
n
t
v
ec
to
r
izer
a
n
d
th
e
RF
m
o
d
el
with
a
T
F
-
I
DF
v
ec
to
r
izer
b
o
t
h
ac
h
iev
ed
9
2
.
4
%
ac
cu
r
ac
y
with
th
e
D1
d
ataset.
Fo
r
th
e
D2
d
a
taset,
th
e
DT
with
th
e
T
F
-
I
DF
v
ec
to
r
izer
r
ea
c
h
ed
a
h
ig
h
er
ac
cu
r
ac
y
o
f
9
9
.
5
%,
wh
ile
th
e
ar
tific
ial
n
eu
r
al
n
et
wo
r
k
(
ANN)
m
o
d
el
ac
h
iev
ed
8
9
.
6
%
ac
cu
r
ac
y
with
D1
an
d
9
9
.
2
% with
D2
.
T
h
e
c
o
m
p
ar
is
o
n
o
f
v
a
r
io
u
s
m
alicio
u
s
UR
L
d
etec
tio
n
m
eth
o
d
s
ca
n
b
e
s
ee
n
in
T
ab
le
1
.
T
ab
le
1
.
C
o
m
p
a
r
is
o
n
o
f
v
ar
io
u
s
m
alicio
u
s
UR
L
d
etec
tio
n
m
eth
o
d
s
b
ased
o
n
f
ea
tu
r
es a
n
d
a
cc
u
r
ac
y
No
M
e
t
h
o
d
F
e
a
t
u
r
e
s
A
c
c
u
r
a
c
y
R
e
mar
k
s
1
M
a
l
a
k
et
al
.
[
1
3
]
U
R
L
l
e
x
i
c
a
l
f
e
a
t
u
r
e
s,
c
o
n
t
e
n
t
f
e
a
t
u
r
e
s
,
r
e
p
u
t
a
t
i
o
n
f
e
a
t
u
r
e
s,
n
e
t
w
o
r
k
f
e
a
t
u
r
e
s
N
a
v
e
B
a
y
e
s:
9
6
%
Th
e
d
a
t
a
se
t
h
a
s
n
o
t
b
e
e
n
f
u
l
l
y
l
e
v
e
r
a
g
e
d
,
a
n
d
t
h
e
r
e
s
u
l
t
s m
a
y
d
i
f
f
e
r
d
e
p
e
n
d
i
n
g
o
n
t
h
e
se
l
e
c
t
e
d
f
e
a
t
u
r
e
s.
2
S
i
r
a
g
e
l
d
i
n
e
t
a
l
.
[
1
4
]
U
R
L
k
e
y
w
o
r
d
s a
n
d
p
a
g
e
c
o
n
t
e
n
t
9
7
%
Th
e
f
e
a
t
u
r
e
s a
v
a
i
l
a
b
l
e
a
r
e
l
i
mi
t
e
d
,
a
n
d
t
h
e
r
e
su
l
t
s m
a
y
v
a
r
y
b
a
se
d
o
n
t
h
e
sp
e
c
i
f
i
c
f
e
a
t
u
r
e
s s
e
l
e
c
t
e
d
f
o
r
a
n
a
l
y
s
i
s.
3
S
a
l
e
e
m
R
a
j
a
e
t
a
l
.
[
1
5
]
H
TM
L
t
a
g
s,
e
v
e
n
t
me
t
h
o
d
s,
D
O
M
k
e
y
w
o
r
d
s,
a
n
d
Ja
v
a
S
c
r
i
p
t
f
u
n
c
t
i
o
n
s
S
V
M
:
8
8
%
R
F
:
9
3
%
F
e
a
t
u
r
e
s
a
r
e
l
i
mi
t
e
d
.
l
o
n
g
e
r
p
r
o
c
e
ss
i
n
g
t
i
m
e
.
u
n
a
b
l
e
t
o
b
e
e
x
t
e
n
d
e
d
.
4
D
e
sai
e
t
a
l
.
[
1
6
]
U
R
L
l
e
x
i
c
a
l
f
e
a
t
u
r
e
s,
c
o
n
t
e
n
t
f
e
a
t
u
r
e
s
,
r
e
p
u
t
a
t
i
o
n
f
e
a
t
u
r
e
s,
n
e
t
w
o
r
k
f
e
a
t
u
r
e
s
R
F
:
9
6
%,
S
V
M
:
9
3
.
5
%
k
-
NN
9
3
%
F
e
a
t
u
r
e
s
a
r
e
l
i
mi
t
e
d
.
d
e
p
e
n
d
i
n
g
o
n
t
h
e
f
e
a
t
u
r
e
s
c
h
o
s
e
n
,
t
h
e
o
u
t
c
o
me
c
o
u
l
d
v
a
r
y
.
5
S
a
l
e
e
m
R
a
j
a
e
t
a
l
.
[
5
]
U
R
L
l
e
x
i
c
a
l
f
e
a
t
u
r
e
s,
c
o
n
t
e
n
t
f
e
a
t
u
r
e
s
,
r
e
p
u
t
a
t
i
o
n
f
e
a
t
u
r
e
s,
n
e
t
w
o
r
k
f
e
a
t
u
r
e
s
k
-
N
N
:
9
8
%,
R
F
:
9
9
%
Th
e
a
v
a
i
l
a
b
l
e
f
e
a
t
u
r
e
s
a
r
e
c
o
n
s
t
r
a
i
n
e
d
,
a
n
d
t
h
e
r
e
s
u
l
t
s m
a
y
f
l
u
c
t
u
a
t
e
d
e
p
e
n
d
i
n
g
o
n
w
h
i
c
h
f
e
a
t
u
r
e
s
a
r
e
s
e
l
e
c
t
e
d
.
6
P
r
a
d
e
e
p
a
e
t
a
l
.
[
1
7
]
R
a
n
k
-
b
a
s
e
d
f
e
a
t
u
r
e
s,
b
a
g
-
of
-
w
o
r
d
s
f
e
a
t
u
r
e
s,
w
e
b
p
a
g
e
c
o
n
t
e
n
t
f
e
a
t
u
r
e
s,
U
R
L
l
e
x
i
c
a
l
f
e
a
t
u
r
e
s
R
F
:
9
9
%
F
e
a
t
u
r
e
s
a
r
e
l
i
mi
t
e
d
.
D
e
p
e
n
d
i
n
g
o
n
t
h
e
f
e
a
t
u
r
e
s
c
h
o
s
e
n
,
t
h
e
o
u
t
c
o
me
c
o
u
l
d
v
a
r
y
7
S
a
l
e
e
m
R
a
j
a
e
t
a
l
.
[
1
8
]
U
R
L
v
e
c
t
o
r
i
z
e
d
f
e
a
t
u
r
e
s
u
si
n
g
N
LP
met
h
o
d
D
T
w
i
t
h
TF
-
I
D
F
v
e
c
t
o
r
i
z
e
r
:
9
9
.
5
%.
U
R
Ls
a
l
o
n
e
m
a
y
n
o
t
b
e
a
d
e
q
u
a
t
e
f
o
r
b
e
t
t
e
r
c
l
a
ssi
f
i
c
a
t
i
o
n
.
Ma
ch
in
e
lear
n
in
g
alg
o
r
ith
m
s
p
lay
a
p
iv
o
tal
r
o
le
in
th
e
d
etec
tio
n
o
f
m
alicio
u
s
web
p
ag
es,
co
n
tr
ib
u
tin
g
s
ig
n
if
ican
tly
to
cy
b
e
r
s
ec
u
r
ity
ef
f
o
r
ts
[
1
9
]
.
E
ac
h
alg
o
r
ith
m
o
f
f
er
s
u
n
iq
u
e
ca
p
ab
ilit
ies
an
d
ap
p
r
o
ac
h
es
to
id
en
tify
in
g
p
o
ten
tially
h
ar
m
f
u
l
web
co
n
ten
t,
t
h
er
eb
y
en
h
a
n
cin
g
th
e
o
v
er
all
ef
f
ec
tiv
e
n
ess
o
f
th
r
ea
t
d
etec
tio
n
s
y
s
tem
s
.
L
o
g
is
tic
r
eg
r
ess
io
n
,
f
o
r
in
s
tan
ce
,
is
ad
ep
t
at
m
o
d
elin
g
th
e
p
r
o
b
a
b
ilit
y
o
f
a
web
p
a
g
e
b
ein
g
m
alicio
u
s
b
ased
o
n
in
p
u
t
f
ea
tu
r
es.
B
y
f
itti
n
g
a
lo
g
is
tic
f
u
n
ctio
n
t
o
th
e
d
ata,
lo
g
is
tic
r
eg
r
ess
io
n
(
L
o
g
R
)
p
r
o
v
id
es
in
ter
p
r
etab
le
r
esu
lts
,
allo
win
g
cy
b
er
s
ec
u
r
ity
e
x
p
er
ts
to
u
n
d
er
s
tan
d
th
e
co
n
tr
ib
u
tin
g
f
ac
t
o
r
s
to
web
p
ag
e
t
h
r
ea
ts
.
SVM
ex
ce
l
in
clas
s
if
y
in
g
web
p
ag
es
b
y
s
ep
ar
atin
g
th
e
m
in
t
o
d
if
f
er
en
t
ca
teg
o
r
ies
u
s
in
g
h
y
p
er
p
lan
es
in
h
ig
h
-
d
im
en
s
io
n
al
s
p
ac
e.
SVMs
ar
e
p
ar
ticu
lar
ly
u
s
ef
u
l
f
o
r
id
en
tify
in
g
co
m
p
lex
p
atter
n
s
a
n
d
o
u
tlier
s
,
en
h
an
cin
g
th
e
d
etec
tio
n
o
f
s
o
p
h
is
ticated
m
alicio
u
s
ac
tiv
ities
o
n
th
e
web
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
7
2
2
-
2
5
8
6
I
AE
S
I
n
t
J
R
o
b
&
A
u
to
m
,
Vo
l
.
1
4
,
No
.
1
,
Ma
r
ch
20
2
5
:
47
-
57
50
An
o
th
er
v
alu
a
b
le
to
o
l
f
o
r
m
alicio
u
s
web
p
ag
e
d
etec
tio
n
is
th
e
s
im
ilar
ity
-
b
ased
class
if
icatio
n
p
r
o
v
id
e
d
b
y
k
-
NN
.
B
y
ex
am
i
n
in
g
th
e
ch
ar
ac
ter
is
tics
o
f
n
ei
g
h
b
o
r
in
g
we
b
p
ag
es,
k
-
NN
ca
n
ac
cu
r
ately
class
if
y
n
ew
in
s
tan
ce
s
b
ased
o
n
t
h
eir
p
r
o
x
im
ity
to
k
n
o
wn
m
alicio
u
s
o
r
b
en
ig
n
ex
a
m
p
les.
Dec
i
s
io
n
tr
ee
s
o
f
f
er
a
tr
an
s
p
ar
en
t
an
d
in
tu
itiv
e
ap
p
r
o
ac
h
to
m
alicio
u
s
web
p
ag
e
d
etec
tio
n
,
b
r
ea
k
in
g
d
o
wn
t
h
e
c
lass
if
icatio
n
p
r
o
ce
s
s
in
to
a
s
er
ies
o
f
b
in
a
r
y
d
ec
is
i
o
n
s
b
ased
o
n
f
ea
tu
r
es
s
u
ch
a
s
UR
L
s
tr
u
ctu
r
e,
co
n
ten
t,
an
d
b
eh
av
io
r
.
R
an
d
o
m
f
o
r
ests
(
R
F)
ex
ten
d
d
ec
is
io
n
tr
ee
s
b
y
ag
g
r
e
g
atin
g
m
u
ltip
le
tr
ee
s
an
d
r
ed
u
ci
n
g
o
v
er
f
itti
n
g
,
lead
in
g
t
o
im
p
r
o
v
ed
g
en
er
aliza
tio
n
p
er
f
o
r
m
an
ce
.
Gr
ad
ien
t
b
o
o
s
tin
g
(
GB
)
alg
o
r
ith
m
s
,
in
clu
d
in
g
g
r
ad
ien
t
b
o
o
s
tin
g
an
d
e
x
tr
em
e
g
r
ad
ien
t
b
o
o
s
tin
g
(
XGBo
o
s
t)
,
s
eq
u
en
tially
b
u
ild
an
en
s
em
b
le
o
f
wea
k
lear
n
er
s
to
iter
ati
v
ely
co
r
r
ec
t
er
r
o
r
s
,
r
esu
ltin
g
in
r
o
b
u
s
t p
r
e
d
ictiv
e
m
o
d
els f
o
r
i
d
en
tify
in
g
m
alicio
u
s
web
co
n
ten
t.
3.
M
E
T
H
O
DS
T
h
e
p
r
o
p
o
s
ed
W
C
A
-
W
2
V
f
o
r
m
alicio
u
s
web
p
ag
e
d
etec
t
io
n
co
n
s
is
ts
o
f
s
ev
er
al
s
tag
es.
First,
it
in
v
o
lv
es
ex
tr
ac
tin
g
an
d
clea
n
in
g
web
co
n
ten
t
to
en
s
u
r
e
d
ata
q
u
ality
.
Nex
t,
th
e
co
n
te
n
t
is
v
ec
to
r
ized
to
tr
an
s
f
o
r
m
it
i
n
to
a
s
u
itab
le
f
o
r
m
at
f
o
r
a
n
aly
s
is
.
Fin
ally
,
th
e
v
ec
to
r
ize
d
d
ata
is
class
if
i
ed
u
s
in
g
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
to
en
h
an
c
e
d
etec
tio
n
ac
cu
r
ac
y
,
as d
e
p
icted
in
Fig
u
r
e
1.
Fig
u
r
e
1
.
Pro
p
o
s
ed
a
p
p
r
o
ac
h
(
W
C
A
-
W
2
V)
3
.
1
.
Da
t
a
s
et
T
h
e
ex
p
er
i
m
en
t
in
c
o
r
p
o
r
a
t
ed
r
en
o
wn
e
d
d
a
t
a
s
e
t
s
in
c
l
u
d
in
g
th
e
U
R
L
d
a
t
a
s
e
t
(
I
S
C
X
-
U
R
L
2
0
1
6
)
[
2
0
]
,
U
N
B
[
2
1
]
,
a
n
d
P
h
i
s
t
a
n
k
[
2
2
]
.
I
n
m
a
c
h
in
e
l
e
ar
n
in
g
,
im
b
a
l
an
c
ed
d
a
t
a
i
s
a
c
o
m
m
o
n
h
u
r
d
l
e
,
wh
e
r
e
o
n
e
c
l
a
s
s
d
o
m
in
a
t
e
s
w
i
th
p
o
in
t
e
d
l
y
m
o
r
e
t
r
i
a
l
s
th
a
n
t
h
e
o
th
e
r
s
.
T
h
i
s
d
i
s
p
ar
i
t
y
c
an
r
e
s
u
l
t
in
b
i
a
s
e
d
m
o
d
e
l
s
th
a
t
l
i
s
t
t
h
e
c
o
m
m
o
n
c
la
s
s
,
l
e
ad
i
n
g
t
o
i
n
f
e
r
io
r
p
er
f
o
r
m
a
n
ce
co
n
c
e
r
n
in
g
t
h
e
m
in
o
r
i
t
y
c
l
a
s
s
[
1
0
]
.
T
o
ad
d
r
e
s
s
t
h
i
s
i
s
s
u
e
a
n
d
u
p
h
o
l
d
im
p
ar
t
i
a
l
o
u
t
c
o
m
e
s
,
an
e
q
u
a
l
n
u
m
b
er
o
f
b
e
n
ig
n
a
n
d
m
a
l
i
c
io
u
s
U
R
L
s
w
e
r
e
th
o
u
g
h
t
f
u
l
ly
n
o
m
in
a
t
ed
f
o
r
e
x
p
e
r
im
e
n
ta
t
i
o
n
.
T
a
b
l
e
2
o
f
f
e
r
s
a
s
u
c
c
in
c
t
o
v
e
r
v
i
e
w
o
f
th
e
to
t
a
l
b
e
n
ig
n
an
d
m
a
l
i
c
i
o
u
s
U
R
L
s
u
t
i
l
i
z
e
d
i
n
t
h
i
s
s
t
u
d
y
.
T
ab
le
2
.
Data
s
et
s
u
m
m
ar
y
No
Ty
p
e
C
o
u
n
t
1
B
e
n
i
g
n
5
5
3
1
2
M
a
l
i
c
i
o
u
s
5
8
8
3
Evaluation Warning : The document was created with Spire.PDF for Python.
I
AE
S
I
n
t
J
R
o
b
&
A
u
to
m
I
SS
N:
2722
-
2
5
8
6
I
n
teg
r
a
tio
n
o
f n
a
t
u
r
a
l la
n
g
u
a
g
e
p
r
o
ce
s
s
in
g
meth
o
d
s
a
n
d
m
a
ch
in
e
lea
r
n
in
g
mo
d
el
… (
S
h
a
h
ee
th
a
Lia
q
u
a
th
a
li
)
51
3
.
2
.
I
nfo
r
m
a
t
io
n
e
x
t
ra
ct
i
o
n
Du
r
in
g
th
e
in
f
o
r
m
atio
n
ex
tr
a
ctio
n
s
tag
e,
th
e
p
r
im
ar
y
aim
is
to
r
etr
iev
e
tex
tu
al
co
n
ten
t
f
r
o
m
th
e
web
p
ag
e,
with
p
ar
ticu
lar
att
en
tio
n
to
p
ar
ag
r
ap
h
(
p
ar
a
)
,
d
iv
is
io
n
(
d
iv
)
,
an
d
m
eta
ta
g
s
.
T
h
is
en
d
ea
v
o
r
is
f
ac
ilit
ated
b
y
u
tili
zin
g
Py
th
o
n
p
ac
k
ag
es
lik
e
r
eq
u
ests
an
d
B
ea
u
tifu
lSo
u
p
,
wh
ic
h
s
tr
ea
m
lin
e
th
e
p
r
o
ce
s
s
es
o
f
web
s
cr
ap
in
g
an
d
HT
ML
p
a
r
s
in
g
.
I
n
itially
,
th
e
r
eq
u
ests
p
ac
k
ag
e
is
u
tili
ze
d
to
ac
q
u
ir
e
th
e
HT
ML
co
n
ten
t
o
f
th
e
d
esig
n
ated
web
p
ag
e
th
r
o
u
g
h
an
HT
T
P
r
eq
u
est.
Su
b
s
eq
u
en
tly
,
B
ea
u
tifu
lSo
u
p
is
em
p
lo
y
ed
to
n
av
ig
ate
th
e
HT
ML
s
tr
u
ctu
r
e
an
d
e
x
tr
ac
t
p
er
tin
en
t
tag
s
co
n
tain
i
n
g
tex
t
u
al
d
ata.
T
h
r
o
u
g
h
iter
ativ
e
p
r
o
ce
s
s
in
g
o
f
ea
ch
ex
tr
ac
ted
tag
,
th
e
r
aw
tex
tu
a
l
co
n
ten
t
is
s
y
s
tem
atica
lly
g
ath
er
ed
a
n
d
r
ea
d
y
f
o
r
s
u
b
s
e
q
u
en
t
a
n
aly
s
is
an
d
p
r
o
ce
s
s
in
g
.
Alg
o
r
ith
m
1
s
u
m
m
ar
izes th
e
en
tire
p
r
o
ce
s
s
o
f
i
n
f
o
r
m
atio
n
ex
tr
ac
tio
n
.
Alg
o
r
ith
m
1
.
I
n
f
o
r
m
atio
n
ex
tr
ac
tio
n
Input: Set of Webpage URLs
Output: Textual content extracted from paragraph, division, and meta tags
ℎ
∈
⃪
(
)
⃪
(
)
⃪
∅
ℎ
∈
ℎ
⃪
∪
(
)
3
.
3
.
Da
t
a
clea
nin
g
Du
r
in
g
t
h
e
p
r
ep
r
o
ce
s
s
in
g
s
tag
e
f
o
r
ea
ch
web
p
a
g
e'
s
tex
tu
al
co
n
ten
t,
a
s
er
ies
o
f
tr
an
s
f
o
r
m
atio
n
s
ar
e
ap
p
lied
to
p
r
ep
ar
e
it
f
o
r
f
u
r
t
h
er
an
al
y
s
is
.
Firstl
y
,
th
e
tex
t
f
r
o
m
ea
ch
we
b
p
ag
e
is
to
k
e
n
ized
in
to
in
d
iv
i
d
u
al
wo
r
d
s
.
Su
b
s
eq
u
e
n
tly
,
ea
c
h
to
k
en
is
co
n
v
er
ted
t
o
lo
wer
ca
s
e,
an
d
a
n
y
n
o
n
-
alp
h
a
b
etica
l
to
k
en
s
ar
e
r
em
o
v
ed
.
Fin
ally
,
a
f
ilter
in
g
p
r
o
ce
s
s
eli
m
in
ates
an
y
r
em
ain
in
g
to
k
en
s
th
at
b
elo
n
g
to
a
p
r
ed
ef
in
e
d
s
et
o
f
s
to
p
wo
r
d
s
,
r
esu
ltin
g
in
a
clea
n
a
n
d
s
tan
d
ar
d
ized
r
ep
r
esen
tatio
n
o
f
th
e
tex
tu
al
co
n
ten
t
o
f
ea
ch
web
p
ag
e.
Alg
o
r
ith
m
2
o
u
tlin
es th
e
co
m
p
lete
p
r
o
ce
d
u
r
e
f
o
r
d
ata
clea
n
s
in
g
.
Alg
o
r
ith
m
2
.
Data
clea
n
in
g
Input: Textual content of each webpage
Output: Cleaned textual content of each webpage
ℎ
•
⃪∅
•
ℎ
∈
(
)
o
(
)
ℎ
▪
⃪
∪
{
(
)
}
3
.
3
.
Av
er
a
g
e
Wo
rd2
Vec
e
mb
eddi
ng
Fo
llo
win
g
th
e
clea
n
s
in
g
o
f
web
co
n
ten
ts
,
th
e
s
u
b
s
eq
u
en
t
s
tag
e
o
f
p
r
o
ce
s
s
in
g
en
ab
les
th
e
r
ep
r
esen
tatio
n
o
f
th
is
co
n
ten
t
in
a
s
tr
u
ct
u
r
ed
an
d
m
ea
n
in
g
f
u
l
m
an
n
e
r
.
T
h
r
o
u
g
h
th
e
ap
p
licatio
n
o
f
W
o
r
d
2
Vec
em
b
ed
d
in
g
,
in
d
iv
id
u
al
wo
r
d
s
with
in
th
e
tex
tu
al
co
n
ten
t
ar
e
co
n
v
er
ted
in
t
o
d
en
s
e
v
ec
to
r
r
ep
r
esen
tatio
n
s
,
ef
f
ec
tiv
ely
ca
p
tu
r
in
g
s
em
an
ti
c
r
elatio
n
s
h
ip
s
b
etwe
e
n
wo
r
d
s
.
T
h
is
tr
an
s
f
o
r
m
ativ
e
p
r
o
ce
s
s
n
o
t
o
n
ly
p
r
eser
v
es
th
e
co
n
tex
tu
al
in
tr
icac
ies
in
h
e
r
en
t
in
th
e
tex
tu
al
d
ata
b
u
t
also
f
ac
ilit
ates
co
m
p
u
tatio
n
al
an
aly
s
is
b
y
en
co
d
in
g
s
em
an
tic
in
f
o
r
m
atio
n
in
to
n
u
m
er
ical
v
ec
to
r
s
.
Fu
r
th
er
m
o
r
e
,
to
r
ef
in
e
th
e
r
e
p
r
esen
tatio
n
o
f
th
e
web
p
a
g
e'
s
tex
tu
al
co
n
ten
t,
t
h
e
d
e
r
iv
ed
W
o
r
d
2
Vec
em
b
ed
d
in
g
s
u
n
d
er
g
o
ag
g
r
eg
atio
n
th
r
o
u
g
h
th
e
co
m
p
u
tatio
n
o
f
av
e
r
ag
e
W
o
r
d
2
Vec
e
m
b
ed
d
in
g
s
f
o
r
ea
c
h
d
o
cu
m
e
n
t.
B
y
av
e
r
ag
in
g
th
e
W
o
r
d
2
Vec
v
ec
to
r
s
co
r
r
esp
o
n
d
in
g
to
th
e
wo
r
d
s
with
in
ea
ch
d
o
cu
m
e
n
t,
a
co
n
d
en
s
ed
an
d
s
em
an
tically
en
r
ich
ed
r
ep
r
esen
tatio
n
o
f
th
e
web
p
ag
e'
s
co
n
ten
t
is
ac
h
ie
v
ed
.
T
h
is
av
er
ag
e
d
em
b
ed
d
in
g
e
n
ca
p
s
u
lates
th
e
s
em
an
tic
ess
en
ce
o
f
th
e
d
o
cu
m
en
t,
f
o
s
ter
in
g
a
co
m
p
r
eh
en
s
iv
e
u
n
d
er
s
tan
d
in
g
o
f
its
tex
tu
al
co
n
ten
t.
T
h
e
c
o
m
p
l
ete
p
r
o
ce
s
s
is
o
u
tlin
ed
in
A
lg
o
r
ith
m
3
.
3
.
4
.
T
ra
ini
ng
a
nd
t
esting
Fo
llo
win
g
th
e
co
m
p
letio
n
o
f
th
e
av
er
ag
e
wo
r
d
em
b
ed
d
in
g
s
tag
e,
th
e
s
u
b
s
eq
u
en
t
p
h
as
e
in
v
o
lv
es
tr
ain
in
g
an
d
test
in
g
th
e
d
ata
u
s
in
g
r
ep
ea
te
d
s
tr
atif
ied
k
-
f
o
ld
cr
o
s
s
-
v
alid
atio
n
.
T
h
is
m
eth
o
d
o
lo
g
y
e
n
s
u
r
es
r
o
b
u
s
tn
ess
in
m
o
d
el
ev
alu
ati
o
n
b
y
r
ep
ea
ted
ly
s
p
litt
in
g
t
h
e
d
ataset
in
to
k
s
u
b
s
ets
wh
ile
m
ain
tain
in
g
class
b
alan
ce
.
E
ac
h
f
o
ld
is
u
tili
ze
d
as
b
o
t
h
tr
ain
i
n
g
a
n
d
test
in
g
d
ata
iter
ativ
el
y
,
allo
win
g
f
o
r
a
co
m
p
r
eh
en
s
iv
e
ass
es
s
m
en
t
o
f
th
e
m
o
d
el'
s
p
er
f
o
r
m
a
n
ce
ac
r
o
s
s
m
u
ltip
le
ite
r
atio
n
s
.
B
y
em
p
lo
y
in
g
t
h
is
ap
p
r
o
ac
h
,
th
e
r
is
k
o
f
o
v
er
f
itti
n
g
is
m
itig
ated
,
an
d
th
e
g
en
er
aliza
tio
n
ca
p
ab
ilit
y
o
f
th
e
m
o
d
el
is
en
h
an
ce
d
,
th
u
s
en
ab
lin
g
m
o
r
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
7
2
2
-
2
5
8
6
I
AE
S
I
n
t
J
R
o
b
&
A
u
to
m
,
Vo
l
.
1
4
,
No
.
1
,
Ma
r
ch
20
2
5
:
47
-
57
52
r
eliab
le
d
etec
tio
n
o
f
m
alicio
u
s
web
p
ag
es
[
2
3
]
–
[
2
5
]
.
I
n
th
e
f
i
n
al
s
tag
e
o
f
th
e
p
r
o
ce
s
s
,
th
e
f
o
cu
s
s
h
if
ts
to
war
d
s
ap
p
ly
in
g
s
ev
e
n
d
is
tin
ct
m
ac
h
i
n
e
lear
n
in
g
al
g
o
r
ith
m
s
—
lo
g
is
tic
r
eg
r
ess
io
n
,
SVM,
k
-
NN
,
d
e
cisi
o
n
tr
ee
,
r
an
d
o
m
f
o
r
est,
g
r
ad
ie
n
t
b
o
o
s
tin
g
,
an
d
XGBo
o
s
t.
T
h
ese
alg
o
r
ith
m
s
ar
e
s
elec
ted
f
o
r
th
eir
d
i
v
er
s
e
ap
p
r
o
ac
h
es
to
class
if
icatio
n
an
d
th
eir
p
o
ten
t
ial
to
ca
p
tu
r
e
d
if
f
e
r
en
t
asp
ec
ts
o
f
th
e
d
ata.
B
y
m
ea
s
u
r
in
g
v
ar
io
u
s
p
er
f
o
r
m
an
c
e
m
etr
ics
s
u
ch
as
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all,
an
d
F1
-
s
co
r
e
ac
r
o
s
s
th
ese
alg
o
r
ith
m
s
,
a
co
m
p
r
eh
en
s
iv
e
ev
alu
atio
n
o
f
th
eir
ef
f
ec
tiv
en
ess
in
m
alicio
u
s
web
p
ag
e
d
etec
tio
n
is
ac
h
i
ev
ed
.
Alg
o
r
ith
m
3
.
Av
er
ag
e
W
o
r
d
2
Vec
Input: Cleaned textual content of webpages (Text)
Output: Average Word2Vec embeddings for each document
1. Word2Vec_Embeddings[]
⃪
[]
2. Avg_Word2Vec_Embeddings[]
⃪
[]
3. for each word w in Text:
•
Word2Vec_Embeddings[w]
⃪
Word2Vec(w)
4. for each document d in Text:
•
Sum_Word2Vec_Embeddings
⃪
0
•
for each word w in d:
o
Sum_Word2Vec_Embeddings
+
⃪
Word2Vec_Embeddings[w]
o
Avg_Word2Vec_Embeddings[d]
⃪
Sum_Word2Vec_Embeddings / |d|
5. return Avg_Word2Vec_Embeddings
4.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
e
ex
p
er
im
en
tal
r
esu
lts
wer
e
co
n
d
u
cted
with
a
s
y
s
tem
c
o
n
f
ig
u
r
ed
with
a
W
in
d
o
ws
1
5
o
p
er
atin
g
s
y
s
tem
r
u
n
n
in
g
o
n
a
2
GHz
p
r
o
ce
s
s
o
r
.
T
h
e
ex
p
e
r
im
en
tati
o
n
en
v
ir
o
n
m
en
t
was
f
ac
ilit
ated
th
r
o
u
g
h
J
u
p
y
ter
No
teb
o
o
k
,
p
r
o
v
i
d
in
g
an
in
ter
a
ctiv
e
p
latf
o
r
m
f
o
r
co
d
in
g
an
d
an
aly
s
is
.
T
h
e
p
r
o
g
r
am
m
in
g
ta
s
k
s
wer
e
ex
ec
u
ted
u
tili
zin
g
th
e
s
cik
it
-
lear
n
p
ac
k
ag
e,
a
co
m
p
r
eh
e
n
s
iv
e
lib
r
a
r
y
f
o
r
m
ac
h
in
e
lear
n
in
g
task
s
i
n
Py
th
o
n
.
L
ev
er
a
g
in
g
th
is
s
etu
p
,
th
e
p
er
f
o
r
m
an
c
e
o
f
v
ar
io
u
s
m
ac
h
in
e
lear
n
in
g
al
g
o
r
ith
m
s
was
ass
ess
ed
f
o
r
m
alicio
u
s
web
p
ag
e
d
etec
tio
n
.
T
h
e
e
v
alu
atio
n
p
r
o
ce
s
s
in
v
o
lv
ed
test
in
g
th
e
tex
tu
al
co
n
te
n
ts
ex
tr
ac
ted
f
r
o
m
in
d
iv
id
u
al
HT
ML
tag
s
s
u
ch
a
s
p
ar
ag
r
ap
h
(
Par
a)
,
m
eta,
a
n
d
d
iv
is
io
n
(
Div
)
,
as
well
as
co
m
b
in
ed
tex
t
f
r
o
m
all
tag
s
.
T
h
e
r
esu
lts
o
f
th
ese
ex
p
er
im
en
ts
ar
e
p
r
esen
ted
i
n
T
ab
les
3
to
6
,
ea
c
h
s
h
o
wca
s
in
g
th
e
p
er
f
o
r
m
an
ce
m
etr
ics
o
f
v
ar
io
u
s
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
in
clu
d
in
g
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all
,
an
d
F1
-
s
co
r
e
.
A
r
ig
o
r
o
u
s
ev
alu
atio
n
was
co
n
d
u
cted
u
s
in
g
r
e
p
ea
ted
s
tr
atif
ied
k
-
f
o
ld
c
r
o
s
s
-
v
alid
atio
n
,
e
n
s
u
r
in
g
r
o
b
u
s
tn
ess
an
d
r
eliab
ilit
y
in
ass
ess
in
g
t
h
e
ef
f
ec
tiv
en
ess
o
f
ea
ch
alg
o
r
ith
m
f
o
r
m
alicio
u
s
web
p
a
g
e
d
etec
tio
n
.
T
h
e
ass
es
s
m
en
t
o
f
t
h
e
r
a
n
d
o
m
f
o
r
est
alg
o
r
ith
m
is
d
o
n
e
th
r
o
u
g
h
co
n
f
u
s
io
n
m
atr
ices
an
d
R
OC
-
AUC
cu
r
v
es
f
o
r
v
ar
io
u
s
HT
ML
tag
s
,
in
clu
d
in
g
p
ar
ag
r
a
p
h
as
s
h
o
wn
in
Fig
u
r
e
s
2
an
d
3
,
Div
as
s
h
o
wn
in
Fi
g
u
r
es
4
an
d
5
,
an
d
Me
ta
as
s
h
o
wn
in
Fig
u
r
es
6
a
n
d
7
.
A
d
d
itio
n
ally
,
t
h
e
co
m
b
i
n
ed
tex
tu
al
c
o
n
ten
ts
'
p
er
f
o
r
m
a
n
ce
is
illu
s
tr
ated
in
Fig
u
r
es 8
an
d
9
.
T
ab
le
3
.
Per
f
o
r
m
an
ce
o
f
tex
t
u
al
co
n
ten
t o
f
p
ar
a
g
r
ap
h
T
ag
A
l
g
o
r
i
t
h
m
A
c
c
u
r
a
c
y
(
%)
P
r
e
c
i
s
i
o
n
(
%)
R
e
c
a
l
l
(
%)
F1
-
S
c
o
r
e
(
%)
S
p
e
c
i
f
i
c
i
t
y
(
%)
Lo
g
R
8
4
.
5
8
5
.
6
8
4
.
7
8
5
.
1
8
1
.
8
S
V
C
8
6
.
5
8
8
.
1
8
6
.
4
8
7
.
3
8
3
.
5
k
-
NN
9
0
.
1
9
4
.
1
9
3
.
7
9
3
.
9
8
8
.
2
DT
9
1
.
4
9
9
.
3
9
9
.
8
9
9
.
5
8
9
.
6
RF
9
3
.
9
9
9
.
2
9
9
.
9
9
9
.
5
9
1
.
2
GB
8
9
.
5
9
2
.
3
9
0
.
7
9
1
.
5
8
7
.
1
XGB
9
3
.
9
9
9
.
2
9
9
.
8
9
9
.
5
9
1
.
4
T
ab
le
4
.
Per
f
o
r
m
an
ce
o
f
tex
t
u
al
co
n
ten
t o
f
Div
T
ag
A
l
g
o
r
i
t
h
m
A
c
c
u
r
a
c
y
(
%)
P
r
e
c
i
s
i
o
n
(
%)
R
e
c
a
l
l
(
%)
F1
-
S
c
o
r
e
(
%)
S
p
e
c
i
f
i
c
i
t
y
(
%)
Lo
g
R
8
3
.
2
8
3
.
1
8
4
.
6
8
3
.
8
8
0
.
9
S
V
C
8
5
.
8
8
5
.
6
8
8
.
2
8
6
.
9
8
2
.
7
k
-
NN
8
9
.
9
9
3
.
1
9
4
.
3
9
3
.
7
8
8
.
4
DT
9
1
.
8
9
8
.
6
9
9
.
4
9
9
.
0
8
9
.
2
RF
9
3
.
7
9
8
.
6
9
9
.
4
9
9
.
0
9
1
.
5
GB
8
8
.
8
9
0
.
5
9
1
.
6
9
1
.
1
8
6
.
3
XGB
9
3
.
5
9
8
.
5
9
9
.
4
9
9
.
0
9
0
.
7
Evaluation Warning : The document was created with Spire.PDF for Python.
I
AE
S
I
n
t
J
R
o
b
&
A
u
to
m
I
SS
N:
2722
-
2
5
8
6
I
n
teg
r
a
tio
n
o
f n
a
t
u
r
a
l la
n
g
u
a
g
e
p
r
o
ce
s
s
in
g
meth
o
d
s
a
n
d
m
a
ch
in
e
lea
r
n
in
g
mo
d
el
… (
S
h
a
h
ee
th
a
Lia
q
u
a
th
a
li
)
53
T
ab
le
5
.
Per
f
o
r
m
an
ce
o
f
tex
t
u
al
co
n
ten
t o
f
Me
ta
T
ag
A
l
g
o
r
i
t
h
m
A
c
c
u
r
a
c
y
(
%)
P
r
e
c
i
s
i
o
n
(
%)
R
e
c
a
l
l
(
%)
F1
-
S
c
o
r
e
(
%)
S
p
e
c
i
f
i
c
i
t
y
(
%)
Lo
g
R
7
6
.
9
8
3
.
6
6
8
.
2
7
5
.
1
7
2
.
4
S
V
C
7
8
.
3
8
7
.
4
6
8
.
2
7
6
.
6
7
4
.
1
k
-
NN
8
7
.
1
9
0
.
7
9
1
.
8
9
1
.
2
8
5
.
6
DT
8
9
.
7
9
7
.
7
9
9
.
7
9
8
.
7
9
0
.
3
RF
9
2
.
4
9
7
.
6
9
9
.
8
9
8
.
7
9
1
.
7
GB
8
6
.
5
8
9
.
1
8
7
.
8
8
8
.
5
8
4
.
6
XGB
9
1
.
9
9
7
.
6
9
9
.
5
9
8
.
5
9
0
.
8
T
ab
le
6
.
Per
f
o
r
m
an
ce
o
f
C
o
m
b
in
ed
(
Par
a,
Div
an
d
Me
ta
ta
g
s
)
tex
tu
al
co
n
ten
ts
A
l
g
o
r
i
t
h
m
A
c
c
u
r
a
c
y
(
%)
P
r
e
c
i
s
i
o
n
(
%)
R
e
c
a
l
l
(
%)
F1
-
S
c
o
r
e
(
%)
S
p
e
c
i
f
i
c
i
t
y
(
%)
Lo
g
R
8
7
.
2
8
7
.
6
8
7
.
7
8
7
.
6
8
5
.
0
S
V
C
8
9
.
6
9
0
.
0
8
9
.
8
8
9
.
9
8
7
.
5
k
-
NN
9
1
.
7
9
2
.
5
9
1
.
2
9
1
.
9
8
9
.
7
DT
9
2
.
1
9
1
.
3
9
3
.
6
9
2
.
4
9
0
.
5
RF
9
4
.
7
9
4
.
9
9
4
.
7
9
4
.
8
9
2
.
3
GB
9
1
.
1
9
1
.
0
9
1
.
7
9
1
.
4
8
9
.
8
XGB
9
4
.
6
9
4
.
4
9
5
.
1
9
4
.
7
9
2
.
0
Fig
u
r
e
2
.
C
o
n
f
u
s
io
n
m
atr
i
x
o
f
r
an
d
o
m
f
o
r
est
alg
o
r
ith
m
f
o
r
p
a
r
ag
r
a
p
h
tag
Fig
u
r
e
3
.
R
OC
-
AUC o
f
r
an
d
o
m
f
o
r
est alg
o
r
ith
m
f
o
r
p
ar
ag
r
a
p
h
tag
Fig
u
r
e
4
.
C
o
n
f
u
s
io
n
m
atr
i
x
o
f
r
an
d
o
m
f
o
r
est
alg
o
r
ith
m
f
o
r
Div
T
a
g
Fig
u
r
e
5
.
R
OC
-
AUC o
f
r
an
d
o
m
f
o
r
est alg
o
r
ith
m
f
o
r
Div
T
ag
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
7
2
2
-
2
5
8
6
I
AE
S
I
n
t
J
R
o
b
&
A
u
to
m
,
Vo
l
.
1
4
,
No
.
1
,
Ma
r
ch
20
2
5
:
47
-
57
54
Fig
u
r
e
6
.
C
o
n
f
u
s
io
n
m
atr
i
x
o
f
r
an
d
o
m
f
o
r
est
alg
o
r
ith
m
f
o
r
m
eta
ta
g
Fig
u
r
e
7
.
R
OC
-
AUC o
f
r
an
d
o
m
f
o
r
est alg
o
r
ith
m
f
o
r
m
eta
tag
Fig
u
r
e
8
.
C
o
n
f
u
s
io
n
m
atr
i
x
o
f
r
an
d
o
m
f
o
r
est
alg
o
r
ith
m
f
o
r
co
m
b
in
ed
tex
tu
a
l c
o
n
ten
ts
Fig
u
r
e
9
.
R
OC
-
AUC o
f
r
an
d
o
m
f
o
r
est alg
o
r
ith
m
f
o
r
co
m
b
in
ed
tex
tu
al
co
n
ten
ts
T
ab
le
3
r
ep
r
esen
ts
th
e
p
er
f
o
r
m
an
ce
m
etr
ics
o
f
v
ar
i
o
u
s
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
wh
e
n
ap
p
lied
to
th
e
tex
tu
al
co
n
ten
t
ex
tr
ac
ted
s
p
ec
if
ically
f
r
o
m
th
e
Par
ag
r
ap
h
tag
.
E
ac
h
alg
o
r
ith
m
'
s
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all,
F1
-
s
co
r
e,
an
d
s
p
ec
if
icity
ar
e
p
r
o
v
id
ed
.
L
o
o
k
in
g
at
th
e
r
esu
lts
,
it's
ev
id
en
t
th
at
all
a
lg
o
r
ith
m
s
p
e
r
f
o
r
m
r
ea
s
o
n
ab
ly
well
ac
r
o
s
s
th
e
b
o
ar
d
.
R
F
an
d
XGB
class
if
ie
r
s
s
tan
d
o
u
t
with
th
e
h
ig
h
es
t
ac
cu
r
ac
y
,
r
ec
all,
p
r
ec
is
io
n
an
d
F1
-
s
co
r
e
am
o
n
g
all
alg
o
r
ith
m
s
.
Sp
ec
if
ically
,
t
h
e
R
F
class
if
ier
ac
h
iev
es
an
im
p
r
ess
iv
e
ac
cu
r
ac
y
o
f
9
3
.
9
%
an
d
an
F1
-
s
co
r
e
o
f
9
9
.
5
%,
clo
s
ely
f
o
llo
wed
b
y
XG
B
with
an
ac
cu
r
ac
y
o
f
9
3
.
9
% a
n
d
an
F1
-
s
co
r
e
o
f
9
9
.
5
%
as
well.
T
h
ese
alg
o
r
ith
m
s
d
em
o
n
s
tr
ate
r
o
b
u
s
t
p
er
f
o
r
m
a
n
ce
in
ac
cu
r
ately
cl
ass
if
y
in
g
m
alicio
u
s
web
p
ag
es
b
ased
o
n
th
e
tex
tu
al
co
n
ten
t
f
r
o
m
Par
ag
r
ap
h
ta
g
s
.
Mo
r
eo
v
er
,
th
e
s
p
ec
if
icity
v
alu
es
in
d
icate
t
h
e
ab
ilit
y
o
f
th
e
class
if
ier
s
to
c
o
r
r
ec
tly
id
en
tify
b
e
n
ig
n
web
p
ag
es
as
b
e
n
ig
n
.
I
n
t
h
is
co
n
tex
t,
R
F
an
d
XGB
class
if
ier
s
m
ain
tain
h
ig
h
s
p
ec
if
icity
v
alu
es,
in
d
icatin
g
t
h
eir
ca
p
ab
ilit
y
to
ef
f
ec
tiv
ely
d
if
f
er
en
tiate
b
e
n
ig
n
web
p
ag
es f
r
o
m
m
alicio
u
s
o
n
e
s
,
co
n
tr
ib
u
tin
g
to
th
e
o
v
er
all
r
eliab
ilit
y
o
f
th
e
d
etec
tio
n
s
y
s
tem
.
T
ab
le
4
p
r
esen
ts
th
e
p
er
f
o
r
m
an
ce
m
etr
ics
o
f
v
ar
i
o
u
s
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
wh
en
ap
p
lied
t
o
th
e
tex
tu
al
c
o
n
ten
t
s
p
ec
if
ically
ex
tr
ac
ted
f
r
o
m
t
h
e
Div
ta
g
.
E
ac
h
alg
o
r
ith
m
'
s
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all,
F1
-
s
co
r
e,
an
d
s
p
ec
if
icity
ar
e
lis
ted
.
Ob
s
er
v
in
g
t
h
e
r
esu
lt
s
,
it's
n
o
tab
le
th
at
all
alg
o
r
i
th
m
s
d
em
o
n
s
tr
ate
co
m
m
en
d
a
b
le
p
er
f
o
r
m
an
ce
ac
r
o
s
s
th
e
m
etr
ics.
R
F
an
d
E
x
tr
em
e
Gr
ad
ien
t
B
o
o
s
tin
g
(
XGB
)
class
if
ier
s
ex
h
ib
it
n
o
tab
ly
h
ig
h
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all,
an
d
F1
-
s
co
r
e
am
o
n
g
th
e
alg
o
r
ith
m
s
ass
es
s
ed
.
S
p
ec
if
ically
,
th
e
R
F
class
if
ier
ac
h
iev
es
an
ac
cu
r
ac
y
o
f
9
3
.
7
%
an
d
an
F1
-
s
co
r
e
o
f
9
9
.
0
%,
wh
ile
XGB
clo
s
el
y
f
o
llo
ws
with
an
ac
cu
r
ac
y
o
f
9
3
.
5
%
a
n
d
a
n
F1
-
s
co
r
e
o
f
9
9
.
0
%
as
well.
T
h
es
e
alg
o
r
ith
m
s
d
em
o
n
s
tr
ate
r
o
b
u
s
t
p
er
f
o
r
m
an
ce
i
n
ac
cu
r
ately
class
if
y
in
g
m
alicio
u
s
web
p
ag
es
b
ased
o
n
th
e
tex
tu
al
co
n
ten
t
ex
tr
ac
te
d
f
r
o
m
D
iv
tag
s
.
Mo
r
eo
v
er
,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
AE
S
I
n
t
J
R
o
b
&
A
u
to
m
I
SS
N:
2722
-
2
5
8
6
I
n
teg
r
a
tio
n
o
f n
a
t
u
r
a
l la
n
g
u
a
g
e
p
r
o
ce
s
s
in
g
meth
o
d
s
a
n
d
m
a
ch
in
e
lea
r
n
in
g
mo
d
el
… (
S
h
a
h
ee
th
a
Lia
q
u
a
th
a
li
)
55
th
e
s
p
ec
if
icity
v
alu
es
in
d
icate
th
e
class
if
ier
s
'
ab
ilit
y
to
co
r
r
ec
tly
id
en
tify
b
en
ig
n
web
p
ag
e
s
as
b
en
ig
n
,
f
u
r
th
er
en
h
an
cin
g
th
e
r
eliab
ilit
y
o
f
t
h
e
d
etec
tio
n
s
y
s
tem
.
R
F
an
d
XGB
class
if
ier
s
m
ain
tain
h
ig
h
s
p
ec
if
icity
v
alu
es,
r
ef
lectin
g
th
eir
e
f
f
ec
tiv
en
ess
in
d
is
tin
g
u
is
h
in
g
b
en
ig
n
web
p
a
g
es f
r
o
m
m
alicio
u
s
o
n
es.
T
ab
le
5
p
r
esen
ts
th
e
ev
alu
ati
o
n
o
f
v
ar
io
u
s
m
ac
h
in
e
lear
n
i
n
g
s
y
s
tem
s
wh
en
ap
p
lied
to
th
e
tex
tu
al
co
n
ten
t
s
p
ec
if
ically
ex
tr
ac
ted
f
r
o
m
th
e
Me
ta
tag
.
E
ac
h
alg
o
r
ith
m
'
s
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
e
ca
ll,
F1
-
s
co
r
e,
an
d
s
p
ec
if
icity
ar
e
lis
ted
.
Up
o
n
e
x
am
in
atio
n
o
f
th
e
r
esu
lts
,
it'
s
ev
id
en
t
th
at
all
alg
o
r
ith
m
s
ex
h
ib
it
co
m
m
en
d
ab
le
p
er
f
o
r
m
an
ce
ac
r
o
s
s
th
e
m
etr
i
cs.
No
tab
ly
,
RF
an
d
ex
tr
em
e
g
r
ad
ien
t
b
o
o
s
tin
g
(
XGB)
clas
s
if
ier
s
d
em
o
n
s
tr
ate
h
ig
h
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
a
ll,
an
d
F1
-
s
co
r
e
am
o
n
g
th
e
as
s
ess
ed
alg
o
r
ith
m
s
.
Sp
ec
if
icall
y
,
th
e
R
F
cla
s
s
if
ier
ac
h
iev
es
an
ac
cu
r
ac
y
o
f
9
2
.
4
%
an
d
an
F1
-
s
co
r
e
o
f
9
8
.
7
%,
wh
ile
XG
B
clo
s
ely
f
o
llo
ws
with
an
ac
cu
r
ac
y
o
f
9
1
.
9
%
an
d
a
n
F1
-
s
co
r
e
o
f
9
8
.
5
%.
T
h
ese
alg
o
r
ith
m
s
s
h
o
wc
ase
r
o
b
u
s
t
p
er
f
o
r
m
an
ce
in
ac
cu
r
ately
class
if
y
in
g
m
alicio
u
s
web
p
ag
es
b
ased
o
n
th
e
tex
tu
al
co
n
ten
t
ex
tr
ac
t
ed
f
r
o
m
Me
ta
tag
s
.
Fu
r
th
er
m
o
r
e,
th
e
s
p
ec
if
icity
v
alu
es
p
r
o
v
id
e
in
s
ig
h
ts
in
to
th
e
class
if
ier
s
'
ab
ilit
y
to
co
r
r
e
ctly
id
en
tify
b
en
i
g
n
we
b
p
ag
e
s
as
b
en
ig
n
.
R
F
an
d
XGB
clas
s
if
ier
s
m
ain
tain
h
ig
h
s
p
ec
if
icity
v
alu
es,
in
d
icat
in
g
th
eir
e
f
f
ec
tiv
en
ess
in
d
is
tin
g
u
is
h
in
g
b
en
ig
n
web
p
ag
es f
r
o
m
m
alicio
u
s
o
n
e
s
,
wh
ich
co
n
tr
ib
u
tes to
th
e
o
v
e
r
all
r
eliab
ilit
y
o
f
th
e
d
etec
tio
n
s
y
s
tem
.
T
ab
le
6
p
r
esen
ts
th
e
e
v
alu
at
io
n
o
f
v
ar
i
o
u
s
m
ac
h
i
n
e
lear
n
in
g
s
y
s
tem
s
wh
en
u
tili
zin
g
co
m
b
in
e
d
tex
tu
al
co
n
ten
ts
ex
t
r
ac
ted
f
r
o
m
p
ar
a
g
r
ap
h
(
Par
a)
,
d
iv
is
io
n
(
Div
)
,
an
d
m
eta
tag
s
f
o
r
m
alicio
u
s
web
p
ag
e
d
etec
tio
n
.
T
h
e
al
g
o
r
ith
m
s
e
v
alu
ated
in
clu
d
e
lo
g
is
tic
r
eg
r
e
s
s
io
n
(
L
o
g
R
)
,
s
u
p
p
o
r
t
v
ec
to
r
class
if
ier
(
SVC
)
,
k
-
k
NN,
DT
,
R
F,
GB
,
an
d
XG
B
.
Acr
o
s
s
th
e
b
o
ar
d
,
all
al
g
o
r
i
th
m
s
d
em
o
n
s
tr
ate
s
tr
o
n
g
p
er
f
o
r
m
an
ce
in
ter
m
s
o
f
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all,
a
n
d
F1
-
s
co
r
e
,
in
d
icatin
g
th
eir
ef
f
ec
tiv
en
ess
in
d
etec
tin
g
m
alicio
u
s
web
p
ag
es
b
ased
o
n
c
o
m
b
in
e
d
tex
t
u
al
co
n
ten
t.
Par
ticu
lar
ly
n
o
tewo
r
th
y
is
th
e
p
er
f
o
r
m
an
ce
o
f
R
F
an
d
E
x
tr
em
e
Gr
ad
ie
n
t
B
o
o
s
tin
g
(
XGB)
class
if
ier
s
,
wh
ich
ac
h
iev
e
th
e
h
i
g
h
est
ac
cu
r
ac
y
o
f
9
4
.
7
%
an
d
9
4
.
6
%,
r
esp
ec
tiv
ely
,
alo
n
g
with
im
p
r
ess
iv
e
F1
-
s
co
r
es
o
f
9
4
.
8
%
an
d
9
4
.
7
%,
r
esp
ec
tiv
el
y
.
T
h
ese
class
if
ier
s
al
s
o
m
ain
tain
h
ig
h
p
r
ec
is
io
n
an
d
r
ec
all
v
al
u
es,
r
ef
lectin
g
t
h
eir
r
o
b
u
s
tn
ess
in
co
r
r
ec
tly
i
d
en
tify
in
g
b
o
th
m
alicio
u
s
an
d
b
en
ig
n
in
s
tan
ce
s
.
Fu
r
th
er
m
o
r
e
,
th
e
s
p
ec
if
icity
m
etr
ic
p
r
o
v
id
es
in
s
ig
h
ts
in
to
t
h
e
class
if
ier
s
'
ab
ilit
y
to
ac
cu
r
ately
id
en
tify
b
e
n
ig
n
in
s
tan
ce
s
as
b
en
ig
n
.
R
F
an
d
XGB
class
if
ier
s
ex
h
ib
it
th
e
h
ig
h
est
s
p
ec
if
icity
v
al
u
es
o
f
9
2
.
3
%
a
n
d
9
2
.
0
%,
r
esp
ec
tiv
ely
,
u
n
d
e
r
s
co
r
in
g
th
eir
ef
f
ec
tiv
en
ess
in
d
is
tin
g
u
is
h
in
g
b
etwe
en
b
en
ig
n
an
d
m
alicio
u
s
web
p
ag
es
b
ased
o
n
co
m
b
in
ed
te
x
tu
al
co
n
ten
t.
T
h
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
d
em
o
n
s
tr
ates
s
u
p
er
io
r
p
er
f
o
r
m
an
ce
co
m
p
a
r
ed
to
ex
is
tin
g
m
eth
o
d
s
f
o
r
co
n
ten
t
an
aly
s
is
o
f
web
p
ag
es.
T
h
r
o
u
g
h
a
co
m
p
ar
ativ
e
an
al
y
s
is
,
it
h
as
b
ee
n
estab
lis
h
ed
th
at
o
u
r
ap
p
r
o
ac
h
o
u
tp
er
f
o
r
m
s
tr
ad
itio
n
al
m
eth
o
d
s
in
ac
cu
r
ately
d
etec
tin
g
an
d
an
aly
zin
g
m
alicio
u
s
co
n
ten
t
em
b
ed
d
e
d
with
in
we
b
p
ag
es.
Ad
d
itio
n
ally
,
th
e
s
ca
lab
ilit
y
o
f
o
u
r
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
s
u
r
p
ass
es
th
at
o
f
ex
is
tin
g
m
eth
o
d
s
as sh
o
wn
in
Fig
u
r
e
1
0
.
Fig
u
r
e
1
0
.
Per
f
o
r
m
a
n
ce
co
m
p
ar
is
o
n
5.
CO
NCLU
SI
O
N
T
h
e
ev
er
-
ev
o
l
v
in
g
lan
d
s
ca
p
e
o
f
cy
b
er
t
h
r
ea
ts
u
n
d
er
s
co
r
es
th
e
cr
itical
n
ee
d
f
o
r
r
o
b
u
s
t
m
eth
o
d
s
t
o
d
etec
t
m
alicio
u
s
web
co
n
ten
t.
T
r
ad
itio
n
al
s
ig
n
at
u
r
e
-
b
ased
a
p
p
r
o
ac
h
es
o
f
ten
la
g
in
id
en
tif
y
in
g
s
o
p
h
is
ticated
m
alwa
r
e
an
d
p
h
is
h
in
g
attac
k
s
,
n
ec
ess
itatin
g
m
o
r
e
p
r
o
ac
ti
v
e
an
d
a
d
v
an
ce
d
d
etec
tio
n
te
ch
n
iq
u
es.
Dete
ctio
n
93%
94,70%
60
%
65
%
70
%
75
%
80
%
85
%
90
%
95
%
10
0%
Sale
e
m
et
al.
[
1
5]
Pro
p
o
s
e
d
A
p
p
ro
ach
Perf
o
rm
an
ce
Acc
u
r
acy
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
7
2
2
-
2
5
8
6
I
AE
S
I
n
t
J
R
o
b
&
A
u
to
m
,
Vo
l
.
1
4
,
No
.
1
,
Ma
r
ch
20
2
5
:
47
-
57
56
b
ased
o
n
web
co
n
te
n
t
is
p
a
r
ticu
lar
ly
v
ital
as
m
alicio
u
s
a
cto
r
s
ad
ep
tly
ca
m
o
u
f
lag
e
th
e
ir
ac
tiv
ities
with
in
s
ee
m
in
g
ly
leg
itima
te
web
p
ag
es.
B
y
d
elv
in
g
in
to
th
e
u
n
d
er
l
y
in
g
co
n
ten
t,
it
b
ec
o
m
es
f
ea
s
ib
le
to
u
n
v
eil
s
u
b
tle
p
atter
n
s
an
d
an
o
m
alies
in
d
i
ca
tiv
e
o
f
m
alicio
u
s
in
ten
t.
T
h
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
lev
er
ag
es
a
p
r
etr
ain
ed
W
o
r
d
2
Vec
m
o
d
el
in
co
n
ju
n
ctio
n
with
s
ev
en
d
iv
er
s
e
m
ac
h
in
e
lear
n
i
n
g
class
if
ier
s
to
b
o
ls
ter
m
alicio
u
s
web
p
ag
e
d
etec
tio
n
.
I
n
itially
,
web
co
n
ten
ts
ar
e
en
co
d
ed
in
t
o
d
en
s
e
v
ec
t
o
r
r
e
p
r
esen
tatio
n
s
u
s
in
g
W
o
r
d
2
Vec
em
b
ed
d
in
g
s
,
f
o
llo
wed
b
y
th
e
co
m
p
u
tatio
n
o
f
a
v
er
ag
e
e
m
b
e
d
d
in
g
s
f
o
r
ea
ch
d
o
cu
m
en
t.
Su
b
s
eq
u
en
t
tr
ain
in
g
o
f
class
if
ier
s
o
n
th
ese
em
b
ed
d
in
g
s
y
ield
s
p
r
o
m
is
in
g
r
esu
lts
,
with
th
e
r
an
d
o
m
f
o
r
est
class
if
ier
ac
h
iev
in
g
an
ac
cu
r
ac
y
o
f
9
4
.
8
%
an
d
an
F
1
-
s
co
r
e
o
f
9
4
.
9
%,
an
d
th
e
e
x
tr
em
e
g
r
ad
ien
t
b
o
o
s
tin
g
cla
s
s
if
ier
attain
in
g
an
ac
cu
r
ac
y
o
f
9
4
.
6
%
an
d
an
F
1
-
s
co
r
e
o
f
9
4
.
7
%.
T
h
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
ca
n
b
e
en
h
an
ce
d
b
y
in
co
r
p
o
r
atin
g
m
o
r
e
ad
v
an
ce
d
c
o
n
tex
t
-
awa
r
e
p
r
etr
ain
ed
m
o
d
els to
im
p
r
o
v
e
d
etec
tio
n
ac
cu
r
ac
y
.
RE
F
E
R
E
NC
E
S
[
1
]
J.
M
.
B
o
r
k
y
a
n
d
T.
H
.
B
r
a
d
l
e
y
,
“
P
r
o
t
e
c
t
i
n
g
i
n
f
o
r
m
a
t
i
o
n
w
i
t
h
c
y
b
e
r
sec
u
r
i
t
y
,
”
i
n
Ef
f
e
c
t
i
v
e
M
o
d
e
l
-
B
a
s
e
d
S
y
st
e
m
s
En
g
i
n
e
e
ri
n
g
,
S
p
r
i
n
g
e
r
I
n
t
e
r
n
a
t
i
o
n
a
l
P
u
b
l
i
sh
i
n
g
,
2
0
1
9
,
p
p
.
3
4
5
–
4
0
4
.
[
2
]
A
.
S
a
l
e
e
m R
a
j
a
,
S
.
P
e
e
r
b
a
s
h
a
b
,
Y
.
M
o
h
a
mm
e
d
I
q
b
a
l
,
B
.
S
u
n
d
a
r
v
a
d
i
v
a
z
h
a
g
a
n
,
a
n
d
M
.
M
o
h
a
m
e
d
S
u
r
p
u
t
h
e
e
n
,
“
S
t
r
u
c
t
u
r
a
l
a
n
a
l
y
si
s
o
f
U
R
L
f
o
r
mal
i
c
i
o
u
s
U
R
L
d
e
t
e
c
t
i
o
n
u
si
n
g
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
,
”
J
o
u
r
n
a
l
o
f
A
d
v
a
n
c
e
d
Ap
p
l
i
e
d
S
c
i
e
n
t
i
f
i
c
Re
s
e
a
r
c
h
,
v
o
l
.
5
,
n
o
.
4
,
p
p
.
28
–
4
1
,
J
u
l
.
2
0
2
3
,
d
o
i
:
1
0
.
4
6
9
4
7
/
j
o
a
a
s
r
5
4
2
0
2
3
6
7
9
.
[
3
]
C
.
D
o
X
u
a
n
,
H
.
D
.
N
g
u
y
e
n
,
a
n
d
T
.
V
.
N
i
k
o
l
a
e
v
i
c
h
,
“
M
a
l
i
c
i
o
u
s
U
R
L
d
e
t
e
c
t
i
o
n
b
a
se
d
o
n
mac
h
i
n
e
l
e
a
r
n
i
n
g
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
Ad
v
a
n
c
e
d
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
1
1
,
n
o
.
1
,
p
p
.
1
4
8
–
1
5
3
,
2
0
2
0
,
d
o
i
:
1
0
.
1
4
5
6
9
/
i
j
a
c
sa.
2
0
2
0
.
0
1
1
0
1
1
9
.
[
4
]
A
.
S
.
R
a
j
a
,
G
.
P
r
a
d
e
e
p
a
,
a
n
d
N
.
A
r
u
l
k
u
mar,
“
M
u
d
h
r
:
m
a
l
i
c
i
o
u
s
U
R
L
d
e
t
e
c
t
i
o
n
u
si
n
g
h
e
u
r
i
s
t
i
c
r
u
l
e
s
b
a
se
d
a
p
p
r
o
a
c
h
,
”
i
n
AI
P
C
o
n
f
e
re
n
c
e
Pr
o
c
e
e
d
i
n
g
s
,
2
0
2
2
,
v
o
l
.
2
3
9
3
,
d
o
i
:
1
0
.
1
0
6
3
/
5
.
0
0
7
4
0
7
7
.
[
5
]
A
.
S
a
l
e
e
m
R
a
j
a
,
R
.
V
i
n
o
d
i
n
i
,
a
n
d
A
.
K
a
v
i
t
h
a
,
“
Le
x
i
c
a
l
f
e
a
t
u
r
e
s
b
a
s
e
d
ma
l
i
c
i
o
u
s
U
R
L
d
e
t
e
c
t
i
o
n
u
si
n
g
mac
h
i
n
e
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s,
”
M
a
t
e
ri
a
l
s T
o
d
a
y
:
Pro
c
e
e
d
i
n
g
s
,
v
o
l
.
4
7
,
p
p
.
1
6
3
–
1
6
6
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
mat
p
r
.
2
0
2
1
.
0
4
.
0
4
1
.
[
6
]
M
i
m
i
,
“
W
o
r
d
2
V
e
c
f
o
r
w
o
r
d
e
mb
e
d
d
i
n
g
s
-
a
b
e
g
i
n
n
e
r
’
s
g
u
i
d
e
.
”
h
t
t
p
s
:
/
/
w
w
w
.
a
n
a
l
y
t
i
c
s
v
i
d
h
y
a
.
c
o
m/
b
l
o
g
/
2
0
2
1
/
0
7
/
w
o
r
d
2
v
e
c
-
f
o
r
-
w
o
r
d
-
e
m
b
e
d
d
i
n
g
s
-
a
-
b
e
g
i
n
n
e
r
s
-
g
u
i
d
e
/
(
a
c
c
e
s
se
d
A
p
r
.
0
2
,
2
0
2
4
)
.
[
7
]
S
.
R
.
A
.
S
a
ma
d
,
P
.
G
a
n
e
sa
n
,
J.
R
a
j
a
s
e
k
a
r
a
n
,
M
.
R
a
d
h
a
k
r
i
s
h
n
a
n
,
H
.
A
mm
a
i
p
p
a
n
,
a
n
d
V
.
R
a
m
a
m
u
r
t
h
y
,
“
S
mi
s
h
G
u
a
r
d
:
l
e
v
e
r
a
g
i
n
g
mac
h
i
n
e
l
e
a
r
n
i
n
g
a
n
d
n
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
r
o
c
e
ss
i
n
g
f
o
r
smis
h
i
n
g
d
e
t
e
c
t
i
o
n
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
A
d
v
a
n
c
e
d
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
1
4
,
n
o
.
1
1
,
p
p
.
5
8
6
–
5
9
3
,
2
0
2
3
,
d
o
i
:
1
0
.
1
4
5
6
9
/
I
JA
C
S
A
.
2
0
2
3
.
0
1
4
1
1
6
0
.
[
8
]
P
.
Ja
i
n
,
“
B
a
s
i
c
s
o
f
C
o
u
n
t
V
e
c
t
o
r
i
z
e
r
,
”
T
o
w
a
rd
s
D
a
t
a
S
c
i
e
n
c
e
,
2
0
2
1
.
h
t
t
p
s:
/
/
t
o
w
a
r
d
s
d
a
t
a
s
c
i
e
n
c
e
.
c
o
m/
b
a
s
i
c
s
-
of
-
c
o
u
n
t
v
e
c
t
o
r
i
z
e
r
-
e
2
6
6
7
7
9
0
0
f
9
c
.
[
9
]
K
e
n
a
n
E
k
i
c
i
,
“
Y
o
u
r
TFI
D
F
f
e
a
t
u
r
e
s
a
r
e
g
a
r
b
a
g
e
.
H
e
r
e
’
s
h
o
w
t
o
f
i
x
i
t
,
”
M
e
d
i
u
m
,
2
0
2
2
2
.
h
t
t
p
s
:
/
/
me
d
i
u
m
.
c
o
m/
@
k
e
n
a
n
e
k
i
c
i
/
y
o
u
r
-
t
f
i
d
f
-
f
e
a
t
u
r
e
s
-
a
r
e
-
g
a
r
b
a
g
e
-
h
e
r
e
s
-
h
o
w
-
to
-
f
i
x
-
it
-
c
a
5
4
8
8
8
3
d
8
a
0
(
a
c
c
e
ss
e
d
A
p
r
.
0
2
,
2
0
2
4
)
.
[
1
0
]
K
a
v
i
t
a
G
a
n
e
sa
n
,
“
H
a
sh
i
n
g
V
e
c
t
o
r
i
z
e
r
v
s.
C
o
u
n
t
V
e
c
t
o
r
i
z
e
r
.
”
h
t
t
p
s:
/
/
k
a
v
i
t
a
-
g
a
n
e
sa
n
.
c
o
m
/
h
a
s
h
i
n
g
v
e
c
t
o
r
i
z
e
r
-
vs
-
c
o
u
n
t
v
e
c
t
o
r
i
z
e
r
/
(
a
c
c
e
s
se
d
A
p
r
.
0
2
,
2
0
2
4
)
.
[
1
1
]
M
.
S
u
r
i
,
“
A
D
u
mm
y
’
s
G
u
i
d
e
t
o
W
o
r
d
2
V
e
c
,
”
Me
d
i
u
m
,
2
0
2
2
.
h
t
t
p
s
:
/
/
me
d
i
u
m.
c
o
m
/
@
ma
n
a
n
s
u
r
i
/
a
-
d
u
mm
y
s
-
g
u
i
d
e
-
to
-
w
o
r
d
2
v
e
c
-
4
5
6
4
4
4
f
3
c
6
7
3
(
a
c
c
e
sse
d
A
p
r
.
0
2
,
2
0
2
4
)
.
[
1
2
]
W
.
N
.
W
.
M
a
n
a
n
,
A
.
G
.
A
.
A
h
m
e
d
,
a
n
d
M
.
N
.
M
.
K
a
h
a
r
,
“
C
h
a
r
a
c
t
e
r
i
z
i
n
g
c
u
r
r
e
n
t
f
e
a
t
u
r
e
s
o
f
m
a
l
i
c
i
o
u
s
t
h
r
e
a
t
s
o
n
w
e
b
si
t
e
s
,
”
i
n
Ad
v
a
n
c
e
s i
n
I
n
t
e
l
l
i
g
e
n
t
S
y
st
e
m
s
a
n
d
C
o
m
p
u
t
i
n
g
,
v
o
l
.
8
6
6
,
S
p
r
i
n
g
e
r
I
n
t
e
r
n
a
t
i
o
n
a
l
P
u
b
l
i
sh
i
n
g
,
2
0
1
9
,
p
p
.
2
1
0
–
2
1
8
.
[
1
3
]
M
.
A
l
j
a
b
r
i
e
t
a
l
.
,
“
A
n
a
sse
ssm
e
n
t
o
f
l
e
x
i
c
a
l
,
n
e
t
w
o
r
k
,
a
n
d
c
o
n
t
e
n
t
-
b
a
se
d
f
e
a
t
u
r
e
s
f
o
r
d
e
t
e
c
t
i
n
g
m
a
l
i
c
i
o
u
s
U
R
L
s
u
si
n
g
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
a
n
d
d
e
e
p
l
e
a
r
n
i
n
g
m
o
d
e
l
s
,
”
C
o
m
p
u
t
a
t
i
o
n
a
l
I
n
t
e
l
l
i
g
e
n
c
e
a
n
d
N
e
u
ro
s
c
i
e
n
c
e
,
v
o
l
.
2
0
2
2
,
p
p
.
1
–
1
4
,
A
u
g
.
2
0
2
2
,
d
o
i
:
1
0
.
1
1
5
5
/
2
0
2
2
/
3
2
4
1
2
1
6
.
[
1
4
]
A
.
S
i
r
a
g
e
l
d
i
n
,
B
.
B
.
B
a
h
a
r
u
d
i
n
,
a
n
d
L
.
T.
J
u
n
g
,
“
M
a
l
i
c
i
o
u
s w
e
b
p
a
g
e
d
e
t
e
c
t
i
o
n
:
a
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
,
”
i
n
L
e
c
t
u
re
N
o
t
e
s i
n
El
e
c
t
r
i
c
a
l
En
g
i
n
e
e
r
i
n
g
,
v
o
l
.
2
7
9
,
S
p
r
i
n
g
e
r
B
e
r
l
i
n
H
e
i
d
e
l
b
e
r
g
,
2
0
1
4
,
p
p
.
2
1
7
–
2
2
4
.
[
1
5
]
A
.
S
a
l
e
e
m
R
a
j
a
,
B
.
S
u
n
d
a
r
v
a
d
i
v
a
z
h
a
g
a
n
,
R
.
V
i
j
a
y
a
r
a
n
g
a
n
,
a
n
d
S
.
V
e
e
r
a
man
i
,
“
M
a
l
i
c
i
o
u
s
w
e
b
p
a
g
e
c
l
a
ss
i
f
i
c
a
t
i
o
n
b
a
s
e
d
o
n
w
e
b
c
o
n
t
e
n
t
f
e
a
t
u
r
e
s
u
si
n
g
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
a
n
d
d
e
e
p
l
e
a
r
n
i
n
g
,
”
i
n
2
0
2
2
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
G
r
e
e
n
E
n
e
r
g
y
,
C
o
m
p
u
t
i
n
g
a
n
d
S
u
s
t
a
i
n
a
b
l
e
T
e
c
h
n
o
l
o
g
y
,
G
EC
O
S
T
2
0
2
2
,
O
c
t
.
2
0
2
2
,
p
p
.
3
1
4
–
3
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
G
EC
O
S
T
5
5
6
9
4
.
2
0
2
2
.
1
0
0
1
0
3
8
6
.
[
1
6
]
A
.
D
e
s
a
i
,
J
.
Jat
a
k
i
a
,
R
.
N
a
i
k
,
a
n
d
N
.
R
a
u
l
,
“
M
a
l
i
c
i
o
u
s
w
e
b
c
o
n
t
e
n
t
d
e
t
e
c
t
i
o
n
u
si
n
g
mac
h
i
n
e
l
e
a
n
i
n
g
,
”
i
n
RTEI
C
T
2
0
1
7
-
2
n
d
I
EEE
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
R
e
c
e
n
t
T
ren
d
s
i
n
El
e
c
t
r
o
n
i
c
s,
I
n
f
o
rm
a
t
i
o
n
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
T
e
c
h
n
o
l
o
g
y
,
Pro
c
e
e
d
i
n
g
s
,
M
a
y
2
0
1
7
,
p
p
.
1
4
3
2
–
1
4
3
6
,
d
o
i
:
1
0
.
1
1
0
9
/
R
TEI
C
T
.
2
0
1
7
.
8
2
5
6
8
3
4
.
[
1
7
]
G
.
P
r
a
d
e
e
p
a
a
n
d
R
.
D
e
v
i
,
“
Li
g
h
t
w
e
i
g
h
t
a
p
p
r
o
a
c
h
f
o
r
m
a
l
i
c
i
o
u
s
d
o
mai
n
d
e
t
e
c
t
i
o
n
u
s
i
n
g
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
,
”
S
c
i
e
n
t
i
f
i
c
a
n
d
T
e
c
h
n
i
c
a
l
J
o
u
r
n
a
l
o
f
I
n
f
o
rm
a
t
i
o
n
T
e
c
h
n
o
l
o
g
i
e
s
,
Me
c
h
a
n
i
c
s
a
n
d
O
p
t
i
c
s
,
v
o
l
.
2
2
,
n
o
.
2
,
p
p
.
2
6
2
–
2
6
8
,
A
p
r
.
2
0
2
2
,
d
o
i
:
1
0
.
1
7
5
8
6
/
2
2
2
6
-
1
4
9
4
-
2
0
2
2
-
22
-
2
-
2
6
2
-
2
6
8
.
[
1
8
]
A
.
S
.
S
a
l
e
e
m
R
a
j
a
,
G
.
P
r
a
d
e
e
p
a
,
S
.
M
a
h
a
l
a
k
sh
mi
,
a
n
d
M
.
S
.
Ja
y
a
k
u
mar,
“
N
a
t
u
r
a
l
l
a
n
g
u
a
g
e
b
a
se
d
mal
i
c
i
o
u
s
d
o
m
a
i
n
d
e
t
e
c
t
i
o
n
u
si
n
g
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
a
n
d
d
e
e
p
l
e
a
r
n
i
n
g
,
”
S
c
i
e
n
t
i
f
i
c
a
n
d
T
e
c
h
n
i
c
a
l
J
o
u
r
n
a
l
o
f
I
n
f
o
rm
a
t
i
o
n
T
e
c
h
n
o
l
o
g
i
e
s,
M
e
c
h
a
n
i
c
s a
n
d
O
p
t
i
c
s
,
v
o
l
.
2
3
,
n
o
.
2
,
p
p
.
3
0
4
–
3
1
2
,
A
p
r
.
2
0
2
3
,
d
o
i
:
1
0
.
1
7
5
8
6
/
2
2
2
6
-
1
4
9
4
-
2
0
2
3
-
23
-
2
-
3
0
4
-
3
1
2
.
[
1
9
]
S
.
R
.
A
b
d
u
l
S
a
ma
d
e
t
a
l
.
,
“
A
n
a
l
y
s
i
s
o
f
t
h
e
p
e
r
f
o
r
ma
n
c
e
i
m
p
a
c
t
o
f
f
i
n
e
-
t
u
n
e
d
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
m
o
d
e
l
f
o
r
p
h
i
sh
i
n
g
U
R
L
d
e
t
e
c
t
i
o
n
,
”
El
e
c
t
ro
n
i
c
s (
S
w
i
t
z
e
rl
a
n
d
)
,
v
o
l
.
1
2
,
n
o
.
7
,
M
a
r
.
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
e
l
e
c
t
r
o
n
i
c
s
1
2
0
7
1
6
4
2
.
[
2
0
]
M
.
S
i
d
d
h
a
r
t
h
a
,
“
M
a
l
i
c
i
o
u
s
U
R
L
s
d
a
t
a
se
t
,
”
K
a
g
g
l
e
,
2
0
2
1
.
h
t
t
p
s
:
/
/
w
w
w
.
k
a
g
g
l
e
.
c
o
m
/
d
a
t
a
se
t
s/
s
i
d
3
2
1
a
x
n
/
m
a
l
i
c
i
o
u
s
-
u
r
l
s
-
d
a
t
a
se
t
?
r
e
so
u
r
c
e
=
d
o
w
n
l
o
a
d
(
a
c
c
e
sse
d
A
p
r
.
0
2
,
2
0
2
4
)
.
[
2
1
]
“
U
R
L
d
a
t
a
s
e
t
(
I
S
C
X
-
U
R
L2
0
1
6
)
,
”
C
a
n
a
d
i
a
n
I
n
st
i
t
u
t
e
f
o
r
C
y
b
e
rsec
u
r
i
t
y
,
2
0
1
6
.
h
t
t
p
s:
/
/
w
w
w
.
u
n
b
.
c
a
/
c
i
c
/
d
a
t
a
set
s/
u
r
l
-
2
0
1
6
.
h
t
m
l
(
a
c
c
e
s
se
d
A
p
r
.
0
2
,
2
0
2
4
)
.
[
2
2
]
“
‘
P
h
i
s
h
T
a
n
k
,
’
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s:
/
/
w
w
w
.
p
h
i
s
h
t
a
n
k
.
c
o
m
/
d
e
v
e
l
o
p
e
r
_
i
n
f
o
.
p
h
p
.
[
A
c
c
e
ss
e
d
:
2
A
p
r
2
0
2
4
]
.
”
[
2
3
]
A
.
B
.
F
.
K
h
a
n
,
K
.
K
a
m
a
l
a
k
a
n
n
a
n
,
a
n
d
N
.
S
.
S
.
A
h
m
e
d
,
“
I
n
t
e
g
r
a
t
i
n
g
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
a
n
d
st
o
c
h
a
s
t
i
c
p
a
t
t
e
r
n
a
n
a
l
y
si
s
f
o
r
t
h
e
f
o
r
e
c
a
st
i
n
g
o
f
t
i
me
-
s
e
r
i
e
s
d
a
t
a
,
”
S
N
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
4
,
n
o
.
5
,
J
u
n
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
7
/
s
4
2
9
7
9
-
023
-
0
1
9
8
1
-
0.
[
2
4
]
T.
L
i
n
,
Y
.
W
a
n
g
,
X
.
L
i
u
,
a
n
d
X
.
Q
i
u
,
“
A
su
r
v
e
y
o
f
t
r
a
n
sf
o
r
m
e
r
s,”
AI
O
p
e
n
,
v
o
l
.
3
,
p
p
.
1
1
1
–
1
3
2
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
a
i
o
p
e
n
.
2
0
2
2
.
1
0
.
0
0
1
.
[
2
5
]
T.
W
u
,
M
.
W
a
n
g
,
Y
.
X
i
,
a
n
d
Z
.
Z
h
a
o
,
“
M
a
l
i
c
i
o
u
s
U
R
L
d
e
t
e
c
t
i
o
n
m
o
d
e
l
b
a
sed
o
n
b
i
d
i
r
e
c
t
i
o
n
a
l
g
a
t
e
d
r
e
c
u
r
r
e
n
t
u
n
i
t
a
n
d
a
t
t
e
n
t
i
o
n
mec
h
a
n
i
sm
,
”
Ap
p
l
i
e
d
S
c
i
e
n
c
e
s (
S
w
i
t
z
e
rl
a
n
d
)
,
v
o
l
.
1
2
,
n
o
.
2
3
,
D
e
c
.
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
a
p
p
1
2
2
3
1
2
3
6
7
.
Evaluation Warning : The document was created with Spire.PDF for Python.