I
A
E
S
I
n
t
e
r
n
at
io
n
al
Jou
r
n
al
of
A
r
t
if
ic
ia
l
I
n
t
e
ll
ig
e
n
c
e
(
I
J
-
AI
)
V
ol
.
14
, N
o.
5
,
O
c
to
be
r
2025
, pp.
3897
~
3905
I
S
S
N
:
2252
-
8938
,
D
O
I
:
10.11591/
ij
a
i.
v
14
.i
5
.pp
3897
-
3905
3897
Jou
r
n
al
h
om
e
page
:
ht
tp
:
//
ij
ai
.
ia
e
s
c
or
e
.c
om
A
d
van
c
e
d
r
i
sk
ass
e
ss
m
e
n
t
u
s
i
n
g
m
ac
h
i
n
e
l
e
ar
n
i
n
g an
d
se
n
t
i
m
e
n
t
an
al
ysi
s o
n
l
og d
at
a
N
id
al
T
u
r
ab
1
, A
b
d
e
lr
ah
m
an
A
b
u
s
h
at
t
al
1
, Jam
al
A
l
-
N
ab
u
ls
i
2
, H
am
z
a A
b
u
O
w
id
a
2
1
D
e
pa
r
t
m
e
nt
of
N
e
t
w
or
ks
a
nd C
ybe
r
S
e
c
ur
i
t
y, F
a
c
ul
t
y of
I
nf
or
m
a
t
i
on
T
e
c
hnol
ogy, A
l
-
A
hl
i
yya
A
m
m
a
n U
ni
ve
r
s
i
t
y, A
m
m
a
n, J
or
da
n
2
D
e
pa
r
t
m
e
nt
of
M
e
di
c
a
l
E
ngi
ne
e
r
i
ng, F
a
c
ul
t
y of
E
ngi
ne
e
r
i
ng, A
l
-
A
hl
i
yya
A
m
m
a
n U
ni
ve
r
s
i
t
y, A
m
m
a
n, J
or
da
n
A
r
t
ic
le
I
n
f
o
A
B
S
T
R
A
C
T
A
r
ti
c
le
h
is
to
r
y
:
R
e
c
e
iv
e
d
S
e
p 8
,
2024
R
e
vi
s
e
d
J
ul
5
,
2025
A
c
c
e
pt
e
d
A
ug 6
,
2025
Standar
d
risk
assess
ment
appro
ache
s
are
sometime
s
time
-
consumi
ng
and
subjective.
In
order
to
overcome
th
ese
challenges
an
innovative
meth
od
will
be
presented
in
this
article
by
mixing
sentiment
analysis
and
m
achine
learning
(ML)
.
The
suggested
technique
improves
the
effecti
veness,
precision,
and
scope
of
risk
insights
when
it
comes
to
the
detect
ion
of
feelings
in
logs
via
the
use
of
automated
data
collection.
The
re
search
examines
several
different
ML
classifi
ers
and
makes
use
of
a
deep
le
arning
model
that
has
been
pre
-
trained
to
evaluate
risks
in
logs
that
are
multi
-
linguistic.
This
proves
the
adaptability
and
scalability
of
our
techniqu
e
when
used
in
a
multilanguage
setting.
This
combination
of
sentiment
analy
sis
and
ML
are
a
si
gnificant
advancement
in
comparison
to
tradition
al
appr
oaches
since
it
enables
real
-
time
processing
and
delivers
important
insights
i
nto
the
manageme
nt of orga
nizational r
isks.
K
e
y
w
o
r
d
s
:
K
-
ne
a
r
e
s
t
ne
ig
hbor
s
N
a
tu
r
a
l
la
ngua
ge
pr
oc
e
s
s
in
g
S
e
nt
im
e
nt
a
na
ly
s
is
R
is
k a
s
s
e
s
s
m
e
nt
S
uppor
t
m
a
c
hi
ne
s
This is an
open
acce
ss artic
le unde
r the
CC BY
-
SA
license.
C
or
r
e
s
pon
di
n
g A
u
th
or
:
N
id
a
l
T
ur
a
b
D
e
pa
r
tm
e
nt
of
N
e
twor
ks
a
nd C
ybe
r
S
e
c
ur
it
y, F
a
c
ul
ty
of
I
nf
or
m
a
ti
on T
e
c
hnol
ogy
Al
-
A
hl
iy
ya
A
m
m
a
n U
ni
ve
r
s
it
y
A
m
m
a
n 19328, J
or
da
n
E
m
a
il
:
n.t
ur
a
b@
a
m
m
a
nu.e
du.j
o
1.
I
N
T
R
O
D
U
C
T
I
O
N
D
a
ta
lo
gs
a
r
e
e
s
s
e
nt
ia
l
doc
um
e
nt
a
ti
on
of
c
om
put
e
r
s
ys
te
m
s
or
ne
twor
k
e
ve
nt
s
th
a
t
pr
ovi
de
a
n
a
udi
t
tr
a
il
f
or
a
s
s
e
s
s
in
g
a
nd
a
ddr
e
s
s
in
g
pr
obl
e
m
s
.
L
ogs
pl
a
y
a
vi
ta
l
r
ol
e
in
va
r
io
us
f
unc
ti
ons
s
uc
h
a
s
a
udi
t
a
nd
c
om
pl
ia
nc
e
to
e
ns
ur
e
c
om
pl
ia
n
c
e
w
it
h
r
e
gul
a
ti
ons
.
T
he
y
a
ls
o
a
id
in
tr
oubl
e
s
hoot
in
g
by
pr
ovi
di
ng
de
ta
il
e
d
in
f
or
m
a
ti
on
to
f
in
d
th
e
r
oot
c
a
us
e
s
of
f
a
il
ur
e
s
[
1]
.
I
n
a
ddi
ti
on,
l
ogs
a
r
e
us
e
d
f
or
s
e
c
ur
it
y
m
oni
to
r
in
g
to
de
te
c
t
s
us
pi
c
io
us
a
c
ti
vi
ti
e
s
a
nd
pot
e
nt
ia
l
br
e
a
c
he
s
[
2]
.
F
ur
th
e
r
m
or
e
,
lo
gs
a
r
e
va
lu
a
bl
e
f
or
pe
r
f
or
m
a
nc
e
a
na
ly
s
i
s
,
he
lp
in
g
to
unde
r
s
ta
nd
s
ys
te
m
pe
r
f
or
m
a
nc
e
a
nd
f
in
d
a
r
e
a
s
th
a
t
ne
e
d
im
pr
ove
m
e
nt
.
L
ogs
a
r
e
e
s
s
e
nt
ia
l
in
s
tr
um
e
nt
s
f
or
pr
e
s
e
r
vi
ng
in
te
gr
it
y
a
nd
pe
r
f
e
c
ti
ng
th
e
pe
r
f
or
m
a
nc
e
of
te
c
hnol
ogi
c
a
l
in
f
r
a
s
tr
uc
tu
r
e
s
due
to
th
e
ir
di
ve
r
s
e
na
tu
r
e
[
3]
.
W
it
hi
n
th
e
la
nds
c
a
pe
of
c
ybe
r
s
e
c
ur
it
y,
r
is
k
a
s
s
e
s
s
m
e
nt
is
qui
te
a
n
im
por
ta
nt
p
ie
c
e
to
e
ns
ur
e
c
r
uc
ia
l
in
f
or
m
a
ti
on
is
s
e
c
ur
e
d
a
nd
th
e
I
T
in
f
or
m
a
ti
on
s
ys
te
m
s
a
r
e
f
ul
ly
f
unc
ti
ona
l
a
nd
a
va
il
a
bl
e
[
4]
.
S
e
ve
r
a
l
c
onve
nt
io
na
l
r
is
k t
e
c
hni
que
s
a
r
e
r
e
la
te
d t
o c
ybe
r
s
e
c
ur
it
y.
E
xpe
r
t
-
ba
s
e
d r
is
k a
s
s
e
s
s
m
e
nt
(
E
B
R
A
)
us
e
s
e
xpe
r
ts
’
knowle
dge
t
o e
va
lu
a
te
, pr
io
r
it
iz
e
, a
nd de
f
in
e
t
he
r
is
ks
t
ha
t
th
e
s
ys
te
m
ha
s
. H
ow
e
ve
r
, i
t
is
s
us
c
e
pt
ib
le
t
o bi
a
s
e
s
a
nd
c
ont
r
a
di
c
ti
ons
[
5]
.
C
om
pl
ia
nc
e
-
ba
s
e
d
r
is
k
a
s
s
e
s
s
m
e
nt
(
C
B
R
A
)
de
f
in
e
s
th
e
r
is
k
s
by
gua
r
a
nt
e
e
in
g
c
om
pl
ia
nc
e
w
it
h
r
e
gul
a
ti
ons
li
ke
th
e
he
a
lt
h
in
s
ur
a
nc
e
por
ta
bi
li
ty
a
nd
a
c
c
ount
a
bi
li
ty
a
c
t
(
H
I
P
A
A
)
but
la
c
ks
f
le
xi
bi
li
ty
f
or
dyna
m
ic
th
r
e
a
ts
[
6]
.
R
e
d
t
e
a
m
/b
lu
e
te
a
m
e
xe
r
c
is
e
s
(
R
T
B
T
E
)
de
c
id
e
r
is
k
in
c
a
s
e
s
w
it
h
a
tt
a
c
k
e
r
s
(
r
e
d
te
a
m
)
ve
r
s
us
de
f
e
nde
r
s
(
bl
ue
te
a
m
)
[
7]
;
how
e
ve
r
,
it
is
ti
m
e
-
c
ons
um
in
g
f
or
da
ta
c
ol
le
c
ti
on
a
nd
pr
e
-
a
nd
pos
t
-
pa
r
ti
c
ip
a
ti
on s
ta
tu
s
c
ha
r
a
c
t
e
r
iz
a
ti
on [
7]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
.
14
, N
o.
5
,
O
c
to
be
r
2025
:
3897
-
3905
3898
A
r
ti
f
ic
ia
l
in
te
ll
ig
e
nc
e
(
A
I
)
ha
s
pr
ove
n
to
be
a
va
lu
a
bl
e
to
ol
i
n
e
va
lu
a
ti
ng
th
e
c
ybe
r
s
e
c
ur
it
y
r
is
ks
r
a
ngi
ng
f
r
om
A
I
-
ba
s
e
d
a
tt
a
c
ks
to
de
e
pf
a
ke
vi
de
o
s
[
8]
.
T
he
s
pe
e
c
h’
s
m
o
s
t
pot
e
nt
ia
l
a
ppl
ic
a
ti
on
is
u
s
e
of
na
tu
r
a
l
la
ngua
ge
pr
oc
e
s
s
in
g
(
N
L
P
)
,
w
hi
c
h
a
ll
ow
s
one
to
r
e
tr
ie
ve
im
por
ta
nt
da
ta
f
r
om
w
r
it
te
n
doc
um
e
nt
s
.
N
L
P
c
a
n
he
lp
to
in
c
r
e
a
s
e
th
e
e
f
f
ic
ie
nc
y
of
c
ybe
r
s
e
c
ur
it
y
r
is
k
a
s
s
e
s
s
m
e
nt
by
tr
a
ns
f
or
m
in
g
uns
tr
uc
tu
r
e
d
da
ta
in
to
s
tr
uc
tu
r
e
d
a
nd
us
e
f
ul
in
f
o
r
m
a
ti
on
[
9
]
.
E
ve
nt
lo
g
da
ta
is
of
te
n
r
ic
h
in
f
r
e
e
-
te
xt
in
f
or
m
a
ti
on,
m
a
ki
ng
it
pa
r
ti
c
ul
a
r
ly
us
e
f
ul
to
unde
r
s
ta
nd
r
is
k
a
s
s
e
s
s
m
e
nt
in
gr
e
a
t
de
ta
i
l,
a
nd
N
L
P
gr
e
a
tl
y
e
nha
nc
e
s
th
e
c
a
p
a
bi
li
ty
of
a
s
s
e
s
s
in
g
th
e
s
e
r
is
ks
[
9]
.
T
he
im
por
ta
nc
e
of
us
in
g
s
e
nt
im
e
nt
a
na
ly
s
is
is
us
in
g
th
e
c
om
put
e
r
a
lg
or
it
hm
s
t
o
m
e
a
s
ur
e
th
e
ps
yc
hol
ogi
c
a
l
s
tr
a
in
in
te
xt
[
10]
.
S
e
nt
im
e
nt
a
na
ly
s
is
e
va
lu
a
te
s
th
e
s
ubj
e
c
ti
ve
in
f
or
m
a
ti
on
in
c
lu
de
d
in
w
or
ds
,
phr
a
s
e
s
,
or
e
v
e
n
lo
nge
r
te
xt
pa
s
s
a
ge
s
to
f
in
d
if
th
e
a
tt
it
ude
e
xpr
e
s
s
e
d
is
good,
n
e
ga
ti
ve
,
or
ne
ut
r
a
l.
I
t
is
e
xt
e
ns
iv
e
ly
u
s
e
d
f
or
m
oni
to
r
in
g
c
om
pa
ny
r
e
put
a
ti
on,
c
li
e
nt
f
e
e
dba
c
k,
m
a
r
ke
t
r
e
s
e
a
r
c
h,
a
nd
e
nha
nc
in
g
c
us
to
m
e
r
s
e
r
vi
c
e
vi
a
e
m
ot
io
n
-
ba
s
e
d
r
e
s
pon
s
e
s
[
11]
.
T
hus
,
s
e
nt
im
e
nt
a
na
ly
s
is
c
om
pe
n
s
a
te
s
f
or
th
e
time
-
c
ons
um
in
g
na
tu
r
e
of
r
is
k
a
s
s
e
s
s
m
e
nt
by
a
ut
om
a
ti
ng
it
a
nd
pr
ovi
di
ng
r
e
a
l
-
ti
m
e
pr
oc
e
s
s
in
g
th
a
t
in
te
gr
a
ti
on i
m
pr
ove
s
c
or
por
a
te
r
is
k m
a
na
ge
m
e
nt
e
f
f
ic
ie
nc
y, a
c
c
ur
a
c
y, a
nd r
e
s
pons
e
.
2.
R
E
L
A
T
E
D
WORK
A
n
a
nom
a
ly
de
te
c
ti
on
s
ys
te
m
u
s
in
g
N
L
P
a
ppr
oa
c
he
s
f
or
lo
g
a
na
ly
s
is
,
te
r
m
f
r
e
que
nc
y
in
ve
r
s
e
doc
um
e
nt
f
r
e
que
nc
y
(
T
F
-
I
D
F
)
,
pol
a
r
it
y
s
c
or
e
,
a
nd
W
or
d2V
e
c
f
or
ve
c
to
r
iz
a
ti
on
[
12]
.
S
tu
di
a
w
a
n
e
t
al
.
[
13]
pr
opos
e
d
us
in
g
de
e
p
le
a
r
ni
ng
to
de
c
id
e
a
bnor
m
a
li
ti
e
s
in
ope
r
a
ti
ng
s
ys
te
m
lo
gs
us
in
g
s
e
nt
im
e
nt
a
na
ly
s
is
.
U
nba
la
nc
e
d
c
l
a
s
s
di
s
tr
ib
ut
io
n
m
a
y
be
a
=
ha
ndl
e
u
s
in
g
a
g
a
te
d
r
e
c
ur
r
e
nt
uni
t
(
G
R
U
)
la
ye
r
a
nd
T
om
e
k
c
onne
c
ti
on.
A
lo
ng
s
hor
t
-
te
r
m
m
e
m
or
y
(
L
S
T
M
)
ne
twor
k
w
a
s
us
e
d
to
pe
r
f
or
m
s
pa
ti
ot
e
m
por
a
l
s
e
nt
im
e
nt
a
na
ly
s
is
on dis
a
s
te
r
-
r
e
la
te
d t
w
e
e
ts
[
14]
.
T
he
pr
opos
e
d
r
is
k a
s
s
e
s
s
m
e
nt
s
e
nt
im
e
nt
a
na
ly
s
is
(
R
A
S
A
)
m
ode
l
ha
d
s
upe
r
io
r
pe
r
f
or
m
a
nc
e
c
om
pa
r
e
d
to
e
a
r
li
e
r
a
lg
or
it
hm
s
in
th
e
ta
s
k
of
s
e
nt
im
e
nt
c
a
te
gor
iz
a
ti
on.
H
a
n
e
t
al
.
[
15]
a
na
ly
z
e
d
a
nd
c
om
p
a
r
e
d
di
f
f
e
r
e
nt
m
a
c
hi
ne
le
a
r
ni
ng
(
M
L
)
c
la
s
s
if
ie
r
s
f
or
th
e
pur
pos
e
of
a
s
s
e
s
s
in
g
s
of
twa
r
e
r
is
k.
P
ot
e
nt
ia
l
pos
s
ib
il
it
ie
s
f
or
e
nha
nc
in
g
th
e
us
e
of
ML
m
o
de
ls
in
r
is
k
a
s
s
e
s
s
m
e
nt
w
e
r
e
a
l
s
o
di
s
c
us
s
e
d.
I
m
pr
ove
d
de
te
c
ti
on
a
nd
le
s
s
ti
m
e
-
c
ons
um
in
g
ope
r
a
ti
on
a
r
e
bot
h
pr
ovi
de
d
by
th
e
pr
opos
e
d
te
c
hnol
ogy.
A
lm
a
ha
di
n
e
t
al
.
[
16]
pr
opos
e
a
phys
ic
a
l
la
ye
r
s
e
c
ur
e
te
c
hni
que
us
in
g
pha
s
e
in
de
x
R
S
M
to
ove
r
c
om
e
e
a
ve
s
dr
oppe
r
s
e
m
pl
oyi
ng k
-
ne
a
r
e
s
t
ne
ig
hbor
s
(
K
N
N
)
s
upe
r
vi
s
e
d pa
tt
e
r
n r
e
c
ogni
ti
on.
T
he
us
e
of
s
uppor
t
ve
c
to
r
m
a
c
hi
ne
s
(
S
V
M
)
in
c
onj
unc
ti
on
w
i
th
r
e
c
ur
r
e
nt
ne
ur
a
l
ne
twor
ks
(
R
N
N
)
ha
s
be
e
n
s
how
n
to
im
pr
ove
th
e
a
c
c
ur
a
c
y
of
s
e
nt
im
e
nt
c
a
te
gor
i
z
a
ti
on,
r
e
a
c
hi
ng
a
m
a
xi
m
um
of
93.6%
ove
r
a
ll
[
17]
.
C
om
pa
r
e
d
to
c
onve
nt
io
na
l
a
ppr
oa
c
he
s
,
tr
a
ns
f
or
m
e
r
-
ba
s
e
d
m
ode
ls
,
s
uc
h
a
s
D
is
ti
lB
E
R
T
,
ha
ve
s
how
n
s
upe
r
io
r
pe
r
f
or
m
a
nc
e
,
obt
a
in
in
g
a
c
c
ur
a
c
y
r
a
te
s
of
96.10%
in
th
e
c
a
te
gor
iz
a
ti
on
of
e
m
ot
io
n
[
18]
.
N
L
P
is
us
e
d
to
im
pr
ove
ne
twor
k
s
e
c
ur
it
y
lo
g
a
na
ly
s
is
by
m
a
ki
ng c
om
pl
e
x
u
ns
tr
uc
tu
r
e
d
da
ta
in
to
us
a
bl
e
in
f
or
m
a
ti
on
us
in
g
th
e
L
S
T
M
R
N
N
m
ode
l,
w
hi
c
h
pr
ovi
de
s
a
n
F
1
s
c
or
e
of
a
bo
ut
90
[
19]
.
A
lm
odova
r
e
t
al
.
[
20]
in
tr
oduc
e
L
ogF
iT
,
a
s
e
lf
-
s
upe
r
vi
s
e
d
ML
m
ode
l
th
a
t
e
na
bl
e
s
th
e
de
t
e
c
ti
on
of
a
nom
a
li
e
s
in
lo
gs
.
T
he
m
ode
l
is
im
pl
e
m
e
nt
e
d
ba
s
e
d
on
bi
di
r
e
c
ti
ona
l
e
n
c
ode
r
r
e
pr
e
s
e
nt
a
ti
on
s
f
r
om
tr
a
ns
f
or
m
e
r
s
(
B
E
R
T
)
.
P
ha
m
a
nd
L
e
e
[
21]
pr
opos
e
T
r
a
nS
e
nt
L
og
m
e
th
ods
th
a
t
m
e
r
ge
tr
a
ns
f
or
m
e
r
s
w
it
h s
e
nt
im
e
nt
a
na
ly
s
is
to
pr
ovi
de
a
nom
a
ly
de
te
c
ti
on
of
t
he
e
ve
nt
l
ogs
. T
a
bl
e
1 pr
ovi
de
s
a
s
um
m
a
r
y of
t
he
c
ont
r
ib
ut
io
ns
a
nd l
im
it
a
ti
ons
of
e
a
r
li
e
r
w
or
ks
.
T
a
bl
e
1. S
um
m
a
r
y of
c
ont
r
ib
ut
io
ns
a
nd l
im
it
a
ti
ons
f
r
om
pr
e
vi
o
us
w
or
ks
R
e
f
e
r
e
nc
e
s
c
ont
r
i
but
i
on
C
ont
r
i
but
i
on
D
e
e
p l
e
a
r
ni
ng
-
ba
s
e
d G
R
U
n
e
t
w
or
k s
e
nt
i
m
e
nt
a
na
l
ys
i
s
of
O
S
i
s
pr
opos
e
d
[
10]
.
M
e
t
hods
i
nc
l
ude
m
a
nua
l
s
e
a
r
c
hi
ng
or
pr
e
de
t
e
r
m
i
ne
d r
ul
e
s
.
F
oc
us
on gl
oba
l
s
pa
t
i
ot
e
m
por
a
l
c
ha
r
a
c
t
e
r
i
s
t
i
c
s
t
o f
i
nd l
og a
nom
a
l
i
e
s
[
13]
.
C
ur
r
e
nt
t
e
c
hni
que
s
c
onc
e
nt
r
a
t
e
on
di
s
t
r
i
but
i
on,
s
ys
t
e
m
l
og
t
e
m
por
a
l
or
ge
ogr
a
phi
c
a
l
c
ha
r
a
c
t
e
r
i
s
t
i
c
s
S
e
nt
i
m
e
nt
a
nom
a
l
i
e
s
-
ba
s
e
d
s
e
m
i
s
upe
r
vi
s
e
d
[
14
]
.
G
l
oba
l
c
ha
r
a
c
t
e
r
i
s
t
i
c
s
c
a
nnot
be
e
xt
r
a
c
t
e
d
a
c
c
ur
a
t
e
l
y
f
or
a
nom
a
l
y
de
t
e
c
t
i
on us
i
ng c
ur
r
e
nt
a
ppr
oa
c
he
s
.
S
e
m
i
-
s
upe
r
vi
s
e
d
a
nom
a
l
y
de
t
e
c
t
i
on
a
nd
t
i
m
e
-
de
pe
nde
nt
c
onf
us
i
on
m
a
t
r
i
x
f
or
i
m
ba
l
a
nc
e
d da
t
a
s
e
t
a
s
s
e
s
s
m
e
nt
.
3.
M
A
C
H
I
N
E
L
E
A
R
N
I
N
G
C
L
A
S
S
I
F
I
C
A
T
I
O
N
O
ne
of
th
e
m
os
t
im
por
ta
nt
u
s
e
s
of
ML
i
s
pr
e
di
c
ti
ve
a
nd c
la
s
s
if
ic
a
ti
on
da
ta
.
ML
a
lg
or
it
hm
s
a
r
e
e
v
e
r
y
oc
c
ur
r
e
nc
e
in
e
ve
r
y
d
a
ta
s
e
t
w
it
h
th
e
s
a
m
e
a
tt
r
ib
ut
e
s
[
22]
.
ML
c
la
s
s
if
ic
a
ti
on
c
oul
d
b
e
c
a
te
gor
iz
e
d
in
to
va
r
io
us
te
c
hni
que
s
,
m
a
in
ly
s
upe
r
vi
s
e
d
a
nd
uns
up
e
r
vi
s
e
d
le
a
r
ni
ng
[
22]
,
[
23]
.
W
he
n
in
s
ta
nc
e
s
a
r
e
pr
ovi
de
d
w
it
h
known
la
be
ls
(
th
e
out
put
s
th
a
t
c
or
r
e
s
pond
to
th
e
m
)
,
th
e
le
a
r
ni
ng
pr
oc
e
s
s
is
r
e
f
e
r
r
e
d
to
be
s
upe
r
vi
s
e
d.
I
n
th
is
s
e
c
ti
on, we
w
il
l
br
ie
f
ly
i
ll
us
tr
a
te
s
om
e
of
t
he
ML
te
c
hni
que
s
t
ha
t
a
r
e
us
e
d i
n our
r
e
s
e
a
r
c
h.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
A
dv
anc
e
d r
is
k
a
s
s
e
s
s
m
e
nt
us
in
g m
ac
hi
ne
l
e
a
r
ni
ng and s
e
nt
im
e
nt
analy
s
is
on l
og dat
a
(
N
id
al
T
ur
ab
)
3899
3.1. L
ogi
s
t
ic
r
e
gr
e
s
s
io
n
A
lt
hough
lo
gi
s
ti
c
r
e
gr
e
s
s
io
n
w
a
s
m
a
de
to
w
or
k
s
ol
e
ly
f
o
r
bi
na
r
y
c
la
s
s
if
ic
a
ti
on,
it
c
a
n
a
ls
o
be
c
ha
nge
d
to
m
a
na
ge
m
ul
ti
c
la
s
s
c
la
s
s
if
ic
a
ti
ons
th
r
ough
th
e
us
e
of
m
ul
ti
nom
ia
l
lo
gi
s
ti
c
r
e
gr
e
s
s
io
n.
T
hi
s
ve
r
s
io
n
e
na
bl
e
s
th
e
m
ode
l
to
pr
e
di
c
t
pr
oba
bi
li
ti
e
s
of
i
ts
out
c
om
e
s
f
or
m
or
e
th
a
n
th
r
e
e
c
la
s
s
e
s
.
I
t
doe
s
s
o
by
us
in
g
th
e
s
of
tm
a
x f
unc
ti
on i
ns
te
a
d of
t
he
s
ig
m
oi
d f
unc
ti
on [
24]
, [
25]
.
3.2. K
-
n
e
ar
e
s
t
n
e
ig
h
b
or
s
T
he
K
N
N
a
lg
or
it
hm
is
pr
e
di
c
a
te
d
on
th
e
id
e
a
th
a
t
e
ve
r
y
in
s
ta
nc
e
th
a
t
is
in
c
lu
de
d
in
s
id
e
a
da
ta
s
e
t
w
oul
d,
in
m
os
t
c
a
s
e
s
,
be
f
ound
ne
a
r
ot
h
e
r
in
s
ta
nc
e
s
w
it
h
s
im
il
a
r
c
ha
r
a
c
te
r
is
ti
c
s
.
I
f
th
e
in
s
ta
nc
e
s
a
r
e
a
s
s
ig
ne
d
to
a
c
la
s
s
if
ic
a
ti
on
la
be
l,
th
e
la
be
l
va
lu
e
of
a
n
un
c
la
s
s
if
ie
d
in
s
ta
nc
e
c
a
n
be
de
c
id
e
d
by
e
x
a
m
in
in
g
th
e
c
la
s
s
if
ic
a
ti
on
of
th
e
in
s
ta
nc
e
s
c
lo
s
e
s
t
to
it
.
T
hi
s
de
c
is
io
n
is
b
a
s
e
d
on
th
e
c
la
s
s
if
ic
a
ti
ons
of
th
e
ne
ig
hbor
in
g
in
s
ta
nc
e
s
[
26]
.
3.3. S
u
p
p
or
t
ve
c
t
or
m
ac
h
in
e
s
T
he
S
V
M
s
a
r
e
a
r
e
c
e
nt
ly
de
ve
lo
pe
d
m
e
th
od
of
s
upe
r
vi
s
e
d
M
L
.
S
V
M
s
a
r
e
ba
s
e
d
on
th
e
c
onc
e
pt
of
a
”
m
a
r
gi
n”
,
w
hi
c
h
r
e
f
e
r
s
to
th
e
di
s
ta
nc
e
on
e
a
c
h
s
id
e
of
a
hype
r
pl
a
ne
th
a
t
di
vi
de
s
two
c
l
a
s
s
e
s
of
da
ta
.
I
t
h
a
s
be
e
n
pr
ove
d
th
a
t
m
a
xi
m
iz
in
g
th
e
m
a
r
gi
n,
w
hi
c
h
i
s
th
e
lo
ng
e
s
t
pos
s
ib
le
di
s
ta
nc
e
be
twe
e
n
th
e
s
e
pa
r
a
ti
ng
hype
r
pl
a
ne
a
nd t
he
i
ns
ta
nc
e
s
on e
it
he
r
s
id
e
, r
e
duc
e
s
t
h
e
pr
e
di
c
t
e
d ge
ne
r
a
li
z
a
ti
on e
r
r
or
[
27]
.
3.4. Ran
d
om
f
or
e
s
t
c
la
s
s
if
ie
r
A
M
L
te
c
hni
que
known
a
s
th
e
r
a
ndom
f
or
e
s
t
c
la
s
s
if
ie
r
ha
ndl
e
s
c
ons
tr
uc
ti
ng
m
a
ny
de
c
is
io
ns
.
I
t
th
e
n
c
om
bi
ni
ng
th
e
m
in
or
de
r
to
pr
ovi
de
a
f
or
e
c
a
s
t
th
a
t
is
bot
h
m
or
e
c
or
r
e
c
t
a
nd
m
or
e
s
ta
bl
e
.
T
h
e
pr
e
di
c
te
d
a
c
c
ur
a
c
y i
s
i
m
pr
ove
d by the
us
e
of
a
ve
r
a
gi
ng, whic
h a
l
s
o he
lp
s
to
r
e
duc
e
ove
r
f
it
ti
ng [
28]
.
4.
M
E
T
H
O
D
O
L
O
G
Y
I
n
t
h
i
s
s
e
c
t
i
o
n
,
t
h
e
s
ys
t
e
m
de
s
i
g
n
s
t
e
p
w
i
l
l
be
i
l
l
us
t
r
a
te
d
s
te
p
b
y
s
t
e
p.
A
s
s
ho
w
n
i
n
F
i
g
u
r
e
1
,
t
h
e
i
n
f
o
r
m
a
t
i
o
n
a
n
d
th
e
da
t
a
f
r
a
m
e
a
r
e
p
r
o
c
e
s
s
e
d
.
T
h
is
p
r
oc
e
s
s
in
g
is
d
o
ne
us
i
n
g
S
p
y
d
e
r
5
.
4
.
3
P
y
t
h
o
n
E
n
v
i
r
on
m
e
n
t.
F
ig
ur
e
1.
B
lo
c
k di
a
gr
a
m
f
or
a
ML
c
la
s
s
if
ic
a
ti
on pr
oc
e
s
s
4.1. Dat
a l
ab
e
li
n
g
F
or
s
e
nt
im
e
nt
a
na
ly
s
is
w
it
h
th
r
e
e
s
e
nt
im
e
nt
c
la
s
s
e
s
(
hi
gh,
m
ode
r
a
te
,
a
nd
lo
w
)
,
c
or
r
e
s
ponding
to
s
c
or
e
s
of
10,
5,
a
nd
2
r
e
s
p
e
c
ti
ve
ly
,
da
ta
la
b
e
li
ng
in
vol
ve
s
c
a
te
gor
iz
in
g
te
xt
de
s
c
r
ip
ti
ons
ba
s
e
d
on
s
e
nt
im
e
nt
in
te
ns
it
y.
I
n
th
is
pa
pe
r
,
m
a
nua
l
la
be
li
ng
w
a
s
a
dopt
e
d
w
he
r
e
t
he
de
s
c
r
ip
ti
on
of
th
e
lo
g
w
a
s
ge
ne
r
a
te
d
u
s
in
g
C
ha
tG
P
T
.
T
he
r
e
a
s
on
f
or
th
a
t
i
s
th
e
li
m
it
e
d
r
e
s
our
c
e
s
a
v
a
il
a
bl
e
f
or
lo
g
da
ta
s
e
t
s
. T
a
bl
e
2
pr
ovi
de
s
a
s
a
m
pl
e
of
th
e
da
ta
lo
g
s
f
r
om
th
e
d
a
ta
s
e
t,
in
c
lu
di
ng
a
de
s
c
r
ip
ti
on,
s
e
ve
r
it
y,
a
nd s
c
or
e
.
F
ig
ur
e
2
il
lu
s
tr
a
te
s
th
e
di
s
tr
ib
ut
io
n
of
s
e
nt
im
e
nt
c
la
s
s
e
s
in
your
da
ta
s
e
t.
S
pe
c
if
ic
a
ll
y,
it
s
how
s
th
a
t
40%
of
th
e
da
ta
s
e
t
is
c
la
s
s
if
ie
d
a
s
”
ve
r
y
hi
gh
s
e
nt
im
e
nt
,”
w
hi
le
30%
is
c
la
s
s
if
ie
d
a
s
”
m
ode
r
a
te
”
s
e
nt
im
e
nt
,
a
nd
a
not
he
r
30%
is
c
la
s
s
if
ie
d
a
s
”
lo
w
”
s
e
nt
im
e
nt
.
T
he
w
id
e
-
r
a
ngi
ng
di
s
tr
ib
ut
io
n
of
f
e
e
li
ngs
is
he
lp
f
ul
f
or
tr
a
in
in
g
s
e
nt
im
e
nt
a
na
ly
s
is
m
ode
ls
s
in
c
e
it
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
.
14
, N
o.
5
,
O
c
to
be
r
2025
:
3897
-
3905
3900
e
nha
nc
e
s
th
e
m
ode
l’
s
a
bi
li
ty
to
e
f
f
e
c
ti
ve
ly
c
a
te
gor
iz
e
s
e
nt
im
e
n
ts
in
unf
a
m
il
ia
r
te
xt
s
by
in
c
lu
di
ng
a
va
r
ie
ty
o
f
s
e
nt
im
e
nt
i
nt
e
ns
it
ie
s
.
T
a
bl
e
2. L
ogs
da
ta
s
a
m
pl
e
da
ta
s
e
t
H
e
i
ght
t
i
m
e
s
t
a
m
p
D
e
s
c
r
i
pt
i
on
S
e
ve
r
i
t
y
S
c
or
e
08/
06/
2024 13:
20
C
r
i
t
i
c
a
l
s
ys
t
e
m
ove
r
l
oa
d due
t
o D
D
oS
a
t
t
a
c
k. A
l
l
onl
i
ne
s
e
r
vi
c
e
s
a
r
e
dow
n.
V
e
r
y
h
i
gh
10
08/
06/
2024 14:
49
S
us
pi
c
i
ous
e
m
a
i
l
a
t
t
a
c
hm
e
nt
s
ope
n
e
d
i
n
t
he
ne
t
w
or
k,
m
ode
r
a
t
e
r
i
s
k
of
m
a
l
w
a
r
e
s
pr
e
a
d
M
ode
r
a
t
e
5
08/
06/
2024 16:
55
s
ubopt
i
m
a
l
pe
r
f
or
m
a
nc
e
w
a
s
not
i
c
e
d
i
n
c
a
f
e
t
e
r
i
a
P
O
S
s
ys
t
e
m
s
,
w
i
t
h
a
m
i
ni
m
a
l
r
i
s
k of
a
f
f
e
c
t
i
ng s
e
r
vi
c
e
s
pe
e
d
L
ow
2
F
ig
ur
e
2. D
a
ta
vi
s
ua
li
z
a
ti
on f
or
l
a
be
le
d da
ta
4.2. Dat
a p
r
e
p
r
oc
e
s
s
in
g
D
a
ta
pr
e
pr
oc
e
s
s
in
g
is
a
n
e
s
s
e
nt
ia
l
s
ta
ge
in
s
e
nt
im
e
nt
a
na
l
ys
is
th
a
t
f
oc
us
e
s
on
im
pr
ovi
ng
th
e
a
c
c
ur
a
c
y
a
nd
e
f
f
ic
ie
nc
y
of
th
e
m
ode
l.
T
he
pr
oc
e
s
s
be
gi
n
s
w
it
h
te
xt
c
le
a
ns
in
g,
s
pe
c
ia
l
c
h
a
r
a
c
te
r
s
,
a
nd
c
a
pi
ta
l
le
tt
e
r
s
di
s
a
pp
e
a
r
to
e
n
s
ur
e
c
on
s
is
te
n
c
y.
N
e
xt
,
to
ke
ni
z
a
ti
on
i
s
pe
r
f
or
m
e
d,
w
hi
c
h
in
vol
ve
s
di
vi
di
ng
th
e
te
xt
in
to
s
e
pa
r
a
te
w
or
ds
or
to
ke
ns
.
S
to
p
w
or
ds
,
w
hi
c
h
a
r
e
c
om
m
on
te
r
m
s
th
a
t
pr
ovi
de
li
tt
le
m
e
a
ni
ng,
a
r
e
la
te
r
e
li
m
in
a
te
d [
29]
.
4.3. Dat
a n
or
m
al
iz
at
io
n
an
d
s
t
an
d
ar
d
s
c
al
in
g
I
n
or
de
r
to
gua
r
a
nt
e
e
th
a
t
f
e
a
tu
r
e
s
a
r
e
on
a
c
om
pa
r
a
bl
e
s
c
a
le
,
nor
m
a
li
z
a
ti
on
a
nd
s
c
a
li
ng
a
r
e
c
r
uc
ia
l
pr
e
pr
oc
e
s
s
in
g
s
te
p
s
th
a
t
s
houl
d
be
ta
ke
n.
T
hi
s
m
a
ke
s
it
le
s
s
c
om
pl
ic
a
te
d
f
or
M
L
m
ode
ls
to
be
tr
a
in
e
d
a
nd
ge
ne
r
a
li
z
e
d
s
uc
c
e
s
s
f
ul
ly
.
w
he
r
e
s
ta
nda
r
d
xˆ
is
e
va
lu
a
te
d
a
s
in
(
1)
a
nd
w
he
r
e
µ
is
th
e
m
e
a
n
va
lu
e
a
nd
σ
is
th
e
s
ta
nda
r
d
de
vi
a
ti
on,
a
nd
it
c
oul
d
be
c
a
lc
ul
a
te
d
a
s
in
(
2)
.
A
ddi
ti
ona
ll
y,
th
e
r
e
s
houl
d
be
a
n
id
e
nt
ic
a
l
im
pl
e
m
e
nt
a
ti
on
of
th
e
nor
m
a
li
z
in
g
a
nd
s
c
a
li
ng
pr
oc
e
s
s
e
s
f
or
bot
h
th
e
tr
a
in
in
g
da
ta
s
e
t
a
nd
th
e
te
s
t
da
ta
s
e
t.
T
hi
s
pr
oc
e
s
s
s
houl
d be
g
e
ne
r
a
te
d us
in
g t
he
t
r
a
in
in
g da
ta
.
̌
=
−
µ
(
1)
=
√
1
−
1
∑
(
−
µ
)
2
=
1
(
2)
4.4. Dat
a s
p
li
t
t
in
g
I
t
is
e
s
s
e
nt
ia
l
to
di
vi
de
th
e
da
t
a
be
f
or
e
tr
a
in
in
g
a
nd
te
s
ti
ng
M
L
m
ode
ls
.
T
o
tr
a
in
th
e
m
od
e
l,
c
onf
ir
m
it
s
pe
r
f
or
m
a
nc
e
,
a
nd
te
s
t
it
s
ge
ne
r
a
li
z
a
ti
on
to
ne
w
da
ta
it
ha
s
not
s
e
e
n,
th
e
c
onve
nt
io
na
l
te
c
hni
que
in
c
lu
de
s
pa
r
ti
ti
oni
ng
th
e
da
ta
s
e
t
in
to
two
or
m
or
e
s
ub
s
e
ts
to
do
th
e
s
e
va
r
io
us
goa
ls
.
I
n
th
e
r
e
s
ul
t
s
e
c
ti
on,
th
e
e
f
f
e
c
t
of
th
e
s
pi
tt
in
g r
a
ti
o w
il
l
be
c
ons
id
e
r
e
d a
nd i
ll
us
tr
a
te
d.
4.5. F
r
e
q
u
e
n
c
y
-
in
ve
r
s
e
d
oc
u
m
e
n
t
f
r
e
q
u
e
n
c
y ve
c
t
or
iz
e
r
T
he
te
r
m
in
ol
ogy
T
F
-
I
D
F
c
a
lc
ul
a
te
s
th
e
r
e
le
va
n
c
e
of
a
phr
a
s
e
in
a
doc
um
e
nt
w
it
hi
n
a
c
ol
le
c
ti
on.
T
e
xt
a
n
a
ly
s
is
a
nd
N
L
P
us
e
ve
c
to
r
iz
e
r
s
to
tr
a
ns
l
a
te
te
xt
in
to
M
L
-
c
om
pa
ti
bl
e
num
e
r
ic
a
l
r
e
pr
e
s
e
nt
a
ti
ons
.
I
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
A
dv
anc
e
d r
is
k
a
s
s
e
s
s
m
e
nt
us
in
g m
ac
hi
ne
l
e
a
r
ni
ng and s
e
nt
im
e
nt
analy
s
is
on l
og dat
a
(
N
id
al
T
ur
ab
)
3901
m
e
a
s
ur
e
s
th
e
s
ig
ni
f
ic
a
nt
of
a
w
or
d
in
a
doc
um
e
nt
c
om
pa
r
e
d
to
ot
he
r
te
xt
s
[
28]
.
T
he
t
e
c
hni
que
u
s
e
s
te
r
m
f
r
e
que
nc
y
(
T
F
)
to
qua
nt
if
y
th
e
f
r
e
que
nc
y
of
a
phr
a
s
e
in
a
te
x
t
a
nd
in
ve
r
s
e
t
e
xt
f
r
e
que
nc
y
(
I
D
F
)
to
r
e
duc
e
te
r
m
s
th
a
t
a
ppe
a
r
of
te
n
in
m
ul
ti
pe
d
do
c
um
e
nt
s
.
T
F
-
I
D
F
m
ul
ti
pl
ie
s
th
e
s
e
two
m
e
tr
ic
s
to
hi
ghl
ig
ht
uni
que
a
nd
e
s
s
e
nt
ia
l
w
or
ds
i
n t
e
xt
s
. T
hi
s
i
s
us
e
d w
it
h doc
um
e
nt
c
la
s
s
if
ic
a
ti
on a
nd c
lu
s
te
r
in
g [
30]
.
4.6. M
od
e
l
t
r
ai
n
in
g
U
p
to
th
is
poi
nt
,
th
e
tr
a
in
in
g
s
e
t
w
il
l
be
f
e
d
to
th
e
s
e
le
c
te
d
M
L
c
la
s
s
if
ie
r
s
a
nd
c
hoos
in
g
a
s
ui
ta
bl
e
m
e
th
od, pr
e
pa
r
in
g t
he
da
ta
, i
ni
ti
a
li
z
in
g t
he
m
ode
l,
i
te
r
a
ti
ng
t
hr
o
ugh tr
a
in
in
g e
poc
hs
t
o upda
te
pa
r
a
m
e
te
r
s
, a
nd
a
s
s
e
s
s
in
g
th
e
m
ode
l’
s
pe
r
f
or
m
a
nc
e
on
va
li
da
ti
on
a
nd
te
s
t
s
e
ts
a
r
e
a
ll
s
te
ps
th
a
t
a
r
e
in
vol
ve
d
in
th
e
pr
oc
e
s
s
of
tr
a
in
in
g a
M
L
m
ode
l.
I
t
is
a
n e
s
s
e
nt
ia
l
s
ta
ge
i
n M
L
, a
nd i
t
c
a
ll
s
f
or
c
a
r
e
f
ul
a
tt
e
nt
io
n.
4.7
.
M
od
e
l
e
val
u
at
io
n
M
ode
l
e
va
lu
a
ti
on
is
a
n
e
s
s
e
nt
ia
l
s
te
p
th
a
t
m
us
t
be
ta
ke
n
in
or
de
r
to
gua
r
a
nt
e
e
th
a
t
th
e
tr
a
in
e
d
m
ode
l
s
a
ti
s
f
ie
s
th
e
ne
c
e
s
s
a
r
y
pe
r
f
or
m
a
nc
e
r
e
qui
r
e
m
e
nt
s
a
nd
is
pr
ope
r
f
or
th
e
a
ppl
ic
a
ti
on
f
or
w
hi
c
h
i
t
w
a
s
de
-
s
ig
ne
d.
S
e
le
c
ti
ng
r
e
le
va
nt
m
e
tr
ic
s
,
e
va
lu
a
ti
ng
pe
r
f
or
m
a
nc
e
on
t
e
s
t
da
ta
,
a
nd
it
e
r
a
ti
ve
ly
c
ha
ngi
ng
th
e
m
ode
l
a
s
r
e
qui
r
e
d
a
r
e
a
ll
a
c
ti
vi
t
ie
s
th
a
t
a
r
e
in
c
lu
de
d
in
th
is
pr
oc
e
s
s
.
T
he
r
e
a
r
e
va
r
io
us
m
e
tr
ic
s
us
ua
ll
y
us
e
d
to
e
va
lu
a
te
th
e
m
ode
l,
li
ke
a
c
c
ur
a
c
y
in
(
3)
,
p
r
e
c
is
io
n
in
(
4)
,
r
e
c
a
ll
in
(
5)
,
a
nd
th
e
F
1
s
c
or
e
(
6)
.
T
he
f
ol
lo
w
in
g
e
qua
ti
on
is
e
a
c
h one
of
t
he
m
[
30]
:
=
+
+
+
+
(
3)
=
+
(
4)
=
+
(
5)
1
=
2
×
×
+
(
6)
4.8
.
P
r
e
d
ic
t
io
n
H
a
vi
ng
f
in
is
he
d
th
e
tr
a
in
in
g
a
nd
e
va
lu
a
ti
on
of
your
m
o
de
l,
a
nd
be
in
g
de
li
ght
e
d
w
it
h
it
s
pe
r
f
or
m
a
nc
e
,
you
c
a
n
e
m
pl
oy
it
to
ge
ne
r
a
te
pr
e
di
c
ti
ons
on
ne
w
,
unobs
e
r
ve
d
da
ta
.
in
th
is
pa
pe
r
,
th
e
unc
e
r
ta
in
bi
na
r
y
s
ta
r
s
in
th
e
c
a
ta
lo
g
w
il
l
go
th
r
ough
th
e
s
a
m
e
pr
oc
e
s
s
th
a
t
th
e
tr
a
in
in
g
s
e
t
di
d
a
nd
f
in
a
ll
y,
pr
e
di
c
t
th
e
w
a
nt
e
d t
a
r
ge
t
da
ta
.
5.
R
E
S
U
L
T
S
T
he
pe
r
f
or
m
a
nc
e
m
a
tr
ix
th
a
t
is
c
ons
id
e
r
e
d
is
th
e
a
c
c
ur
a
c
y
m
e
tr
ic
w
hi
c
h
r
e
f
le
c
ts
th
e
a
c
c
ur
a
c
y
th
a
t
th
e
m
ode
l
pr
e
di
c
ts
th
e
c
or
r
e
c
t
ta
r
ge
t
va
lu
e
s
.
C
om
pa
r
is
ons
a
r
e
m
a
de
be
twe
e
n
th
e
pe
r
f
or
m
a
nc
e
m
e
tr
ic
s
of
f
ou
r
di
s
ti
nc
t
c
la
s
s
if
ie
r
s
,
na
m
e
ly
S
V
M
,
K
N
N
,
r
a
ndom
f
or
e
s
t,
a
nd
lo
gi
s
ti
c
r
e
gr
e
s
s
io
n.
T
he
s
e
c
om
pa
r
is
ons
a
r
e
m
a
de
a
c
r
os
s
f
our
m
a
jo
r
c
r
it
e
r
ia
:
r
e
c
a
ll
, a
c
c
ur
a
c
y, a
nd pr
e
c
i
s
io
n, a
s
w
e
ll
a
s
t
he
F
1 s
c
or
e
.
F
ig
ur
e
3
pr
ovi
de
s
th
e
pe
r
f
or
m
a
nc
e
a
s
s
e
s
s
m
e
nt
r
e
s
ul
ts
of
s
e
ve
r
a
l
c
la
s
s
if
ic
a
ti
on
m
ode
ls
u
s
in
g
va
r
io
us
m
e
tr
ic
s
.
F
ig
ur
e
3(
a
)
s
how
s
th
e
a
c
c
ur
a
c
y,
w
it
h
th
e
r
a
ndom
f
o
r
e
s
t
c
la
s
s
if
ie
r
a
c
hi
e
vi
ng
th
e
be
s
t
a
c
c
ur
a
c
y
of
90%
.
A
bout
60%
is
th
e
lo
w
e
s
t
le
v
e
l
a
c
hi
e
ve
d
by
S
V
M
.
F
ig
ur
e
3(
b)
s
how
s
how
e
a
c
h
m
ode
l
pr
e
c
is
io
n
d
e
c
id
e
s
pos
it
iv
e
in
s
ta
nc
e
s
w
hi
le
m
in
im
iz
in
g
f
a
ls
e
pos
it
iv
e
s
.
T
he
r
e
s
u
lt
s
s
how
th
a
t
r
a
ndom
f
or
e
s
t
ha
s
th
e
gr
e
a
te
s
t
pr
e
c
is
io
n
(
92%
)
w
hi
le
S
V
M
ha
s
th
e
lo
w
e
s
t,
a
r
ound
80%
.
F
ig
ur
e
3(
c
)
c
om
pa
r
e
s
m
ode
ls
us
in
g
r
e
c
a
ll
m
e
tr
ic
,
ba
la
nc
in
g
pos
it
iv
e
e
ve
nt
de
te
c
ti
on
w
it
h
m
is
s
in
g
a
nd
f
ound
one
s
.
A
c
c
or
di
ng
to
th
e
r
e
s
ul
ts
,
th
e
r
a
ndom
f
o
r
e
s
t
c
la
s
s
if
ie
r
ha
s
th
e
hi
ghe
s
t
r
e
c
a
ll
r
a
te
a
t
94%
.
A
ddi
ti
ona
ll
y,
S
V
M
yi
e
ld
s
th
e
lo
w
e
s
t
le
ve
l,
60%
.
T
he
ba
la
nc
e
d
c
la
s
s
if
ie
r
is
de
c
id
e
d
by
th
e
F
1
-
s
c
or
e
,
w
hi
c
h
c
om
bi
ne
s
pr
e
c
is
io
n
a
nd
r
e
c
a
ll
pe
r
f
or
m
a
nc
e
in
F
ig
ur
e
3(
d)
.
S
upe
r
io
r
pe
r
f
or
m
a
nc
e
i
s
a
c
hi
e
ve
d w
it
h r
a
ndom f
or
e
s
t
(
94.5%
)
.
T
he
c
onf
us
io
n
m
a
tr
ic
e
s
in
F
ig
ur
e
4
s
how
th
a
t
how
th
e
f
our
M
L
m
ode
ls
’
pe
r
f
or
m
a
nc
e
w
a
s
c
a
te
gor
iz
e
d
in
to
th
r
e
e
gr
oups
:
lo
w
(
0)
,
m
ode
r
a
te
(
1)
,
a
nd
hi
gh
(
2)
.
F
ig
ur
e
4(
a
)
pr
e
s
e
nt
s
th
e
c
onf
us
io
n
m
a
tr
ix
f
or
th
e
K
N
N
c
la
s
s
if
ie
r
,
w
he
r
e
it
m
is
s
e
d
c
la
s
s
if
yi
ng
th
e
lo
w
c
la
s
s
a
s
m
ode
r
a
te
.
O
n
th
e
ot
he
r
ha
nd,
F
ig
ur
e
4(
b)
s
how
s
th
e
c
onf
us
io
n
m
a
tr
ix
of
S
V
M
,
w
he
r
e
i
t
ha
s
a
poor
pe
r
f
or
m
a
nc
e
c
la
s
s
if
ic
a
ti
on
th
a
t
is
bi
a
s
e
d
to
”
hi
gh”
c
la
s
s
e
s
.
F
ig
ur
e
4(
c
)
di
s
pl
a
ys
th
e
c
onf
us
io
n
m
a
tr
ix
f
or
lo
gi
s
ti
c
r
e
gr
e
s
s
io
n,
w
h
e
r
e
it
pr
ovi
de
s
e
xc
e
ll
e
nt
c
la
s
s
if
ic
a
ti
on
w
it
h
one
m
is
s
in
g
(
lo
w
)
c
la
s
s
.
F
in
a
ll
y,
F
ig
ur
e
4(
d)
pr
ovi
de
s
th
e
r
a
ndom
f
or
e
s
t
c
la
s
s
if
ie
r
,
it
ha
s
th
e
be
s
t
c
la
s
s
if
ic
a
ti
on
f
or
a
ll
in
s
ta
nc
e
s
.
T
he
r
e
s
ul
ts
s
how
th
a
t
th
e
r
a
ndom
f
or
e
s
t
c
la
s
s
if
ie
r
pe
r
f
or
m
s
hi
gh
s
c
or
e
s
w
he
r
e
it
de
c
id
e
s
a
ll
th
e
in
s
ta
nc
e
s
c
or
r
e
c
tl
y.
T
hu
s
,
th
e
r
a
ndom
f
or
e
s
t
c
la
s
s
if
ie
r
is
th
e
m
os
t
s
ui
ta
bl
e
m
ode
l
f
or
s
e
nt
im
e
nt
a
na
ly
s
is
f
or
r
is
k a
s
s
e
s
s
m
e
nt
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
.
14
, N
o.
5
,
O
c
to
be
r
2025
:
3897
-
3905
3902
(
a
)
(
b)
(
c
)
(
d)
F
ig
ur
e
3. T
he
c
om
pa
r
is
on of
c
la
s
s
if
ic
a
ti
on t
e
c
hni
que
s
:
(
a
)
a
c
c
ur
a
c
y c
om
pa
r
is
on, (
b)
pr
e
c
is
io
n c
om
pa
r
is
on,
(
c
)
r
e
c
a
ll
c
om
pa
r
is
on, a
nd (
d)
F
1 s
c
or
e
c
om
pa
r
is
on
(
a
)
(
b)
(
c
)
(
d)
F
ig
ur
e
4. T
he
c
onf
us
io
n m
a
tr
ix
f
or
c
la
s
s
if
ic
a
ti
on t
e
c
hni
que
s
:
(
a
)
c
onf
us
io
n m
a
tr
ix
f
or
K
N
N
,
(
b)
c
onf
us
io
n m
a
tr
ix
f
or
S
V
M
, (
c
)
c
onf
us
io
n m
a
tr
ix
f
or
lo
g
is
ti
c
r
e
gr
e
s
s
io
n, a
nd
(
d)
c
onf
us
io
n m
a
tr
ix
f
or
r
a
ndom f
or
e
s
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
A
dv
anc
e
d r
is
k
a
s
s
e
s
s
m
e
nt
us
in
g m
ac
hi
ne
l
e
a
r
ni
ng and s
e
nt
im
e
nt
analy
s
is
on l
og dat
a
(
N
id
al
T
ur
ab
)
3903
6.
C
O
N
C
L
U
S
I
O
N
T
he
f
in
di
ngs
of
th
e
r
e
s
e
a
r
c
h
r
e
ve
a
le
d
th
a
t
th
e
c
om
bi
na
ti
on
of
s
e
nt
im
e
nt
a
na
ly
s
is
a
nd
ML
c
l
a
s
s
if
ie
r
s
ha
s
th
e
pot
e
nt
ia
l
to
c
ons
id
e
r
a
bl
y
im
pr
ove
th
e
e
f
f
e
c
ti
ve
ne
s
s
a
nd
pr
e
c
is
io
n
of
r
is
k
a
s
s
e
s
s
m
e
nt
s
u
s
in
g
bot
h
m
e
th
ods
.
T
h
e
s
y
s
te
m
w
a
s
a
bl
e
to
a
ut
om
a
te
th
e
id
e
nt
if
ic
a
ti
on
o
f
s
e
nt
im
e
nt
in
lo
gs
by
u
s
in
g
m
e
th
ods
s
uc
h
a
s
S
V
M
,
K
N
N
,
a
nd
r
a
ndom
f
or
e
s
t
c
la
s
s
if
ie
r
s
.
T
hi
s
a
ll
ow
s
f
or
a
m
or
e
in
-
de
pt
h
unde
r
s
ta
ndi
ng
of
po
s
s
ib
le
da
nge
r
s
.
A
f
ur
th
e
r
il
lu
s
tr
a
ti
on
of
th
e
f
le
xi
bi
li
ty
o
f
th
e
m
ode
l
a
c
r
os
s
a
va
r
ie
ty
of
la
ngua
ge
s
a
nd
da
ta
s
e
ts
is
pr
ovi
de
d
by
th
e
us
e
of
a
pr
e
-
tr
a
in
e
d
de
e
p
le
a
r
ni
ng
m
ode
l
f
or
t
he
a
na
ly
s
is
of
non
-
E
ngl
is
h
lo
gs
.
T
o
in
c
r
e
a
s
e
th
e
pr
e
di
c
te
d a
c
c
ur
a
c
y of
t
he
c
la
s
s
if
ic
a
ti
on a
lg
or
it
hm
s
.
F
U
N
D
I
N
G
I
N
F
O
R
M
A
T
I
O
N
A
ut
hor
s
s
ta
te
t
he
r
e
i
s
no f
undi
ng i
nvol
ve
d.
A
U
T
H
O
R
C
O
N
T
R
I
B
U
T
I
O
N
S
S
T
A
T
E
M
E
N
T
T
hi
s
jo
ur
na
l
us
e
s
th
e
C
ont
r
ib
ut
or
R
ol
e
s
T
a
xonomy
(
C
R
e
di
T
)
to
r
e
c
ogni
z
e
in
di
vi
dua
l
a
ut
hor
c
ont
r
ib
ut
io
ns
,
r
e
duc
e
a
ut
hor
s
hi
p
di
s
put
e
s
,
a
nd
f
a
c
il
it
a
te
c
ol
la
bor
a
ti
on.
T
he
f
ol
lo
w
in
g
ta
bl
e
pr
ovi
de
s
a
s
um
m
a
r
y
of
th
e
a
ut
hor
’
s
c
ont
r
ib
ut
io
ns
to
th
is
r
e
s
e
a
r
c
h
p
a
pe
r
.
E
a
c
h
a
ut
hor
’
s
r
e
s
pons
ib
il
it
ie
s
in
th
e
a
dva
nc
e
m
e
nt
of
t
hi
s
r
e
s
e
a
r
c
h a
r
e
l
is
te
d i
n de
ta
il
.
N
am
e
o
f
A
u
t
h
or
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
N
id
a
l
T
ur
a
b
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
A
b
d
e
l
r
a
hm
a
n
A
bu
s
h
a
t
t
a
l
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
J
a
m
a
l
A
l
-
N
a
bul
s
i
✓
✓
✓
✓
✓
✓
✓
✓
H
a
m
z
a
A
bu O
w
id
a
✓
✓
✓
✓
✓
C
:
C
onc
e
pt
ua
l
i
z
a
t
i
on
M
:
M
e
t
hodol
ogy
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
da
t
i
on
Fo
:
Fo
r
m
a
l
a
na
l
ys
i
s
I
:
I
nve
s
t
i
ga
t
i
on
R
:
R
e
s
our
c
e
s
D
:
D
a
t
a
C
ur
a
t
i
on
O
:
W
r
i
t
i
ng
-
O
r
i
gi
na
l
D
r
a
f
t
E
:
W
r
i
t
i
ng
-
R
e
vi
e
w
&
E
di
t
i
ng
Vi
:
Vi
s
ua
l
i
z
a
t
i
on
Su
:
Su
pe
r
vi
s
i
on
P
:
P
r
oj
e
c
t
a
dm
i
ni
s
t
r
a
t
i
on
Fu
:
Fu
ndi
ng a
c
qui
s
i
t
i
on
C
O
N
F
L
I
C
T
O
F
I
N
T
E
R
E
S
T
S
T
A
T
E
M
E
N
T
A
ut
hor
s
s
ta
te
no c
onf
li
c
t
of
i
nt
e
r
e
s
t.
D
A
T
A
A
V
A
I
L
A
B
I
L
I
T
Y
T
he
da
ta
th
a
t
s
uppor
ts
th
e
f
in
di
ngs
of
th
is
s
tu
dy
a
r
e
a
va
il
a
bl
e
f
r
om
th
e
c
or
r
e
s
ponding
a
ut
hor
,
[
NT
]
,
upon r
e
a
s
ona
bl
e
r
e
que
s
t.
R
E
F
E
R
E
N
C
E
S
[
1]
B
.
C
he
n
a
nd
Z
.
M
.
J
.
J
i
a
ng,
“
A
s
ur
ve
y
of
s
of
t
w
a
r
e
l
og
i
n
s
t
r
um
e
nt
a
t
i
on,”
A
C
M
C
om
put
i
ng
Sur
v
e
y
s
,
vol
.
54,
no.
4,
pp.
1
–
34,
M
a
y
2022, doi
:
10.1145/
3448976.
[
2]
R
.
Á
vi
l
a
,
R
.
K
hour
y,
R
.
K
hour
y,
a
nd
F
.
P
e
t
r
i
l
l
o,
“
U
s
e
of
s
e
c
ur
i
t
y
l
ogs
f
o
r
d
a
t
a
l
e
a
k
de
t
e
c
t
i
on:
a
s
ys
t
e
m
a
t
i
c
l
i
t
e
r
a
t
ur
e
r
e
vi
e
w
,”
Se
c
ur
i
t
y
and C
om
m
uni
c
at
i
on N
e
t
w
o
r
k
s
, vol
. 2021, pp. 1
–
29,
M
a
r
. 2021, doi
:
10.1155/
2021/
6615899.
[
3]
O
.
Y
a
pa
r
,
“
B
l
oc
kc
ha
i
n
-
ba
s
e
d
da
t
a
ba
s
e
m
a
n
a
ge
m
e
nt
f
or
na
t
i
ona
l
s
e
c
ur
i
t
y:
e
n
s
ur
i
ng
da
t
a
i
nt
e
gr
i
t
y
a
nd
pr
i
va
c
y,”
SSR
N
E
l
e
c
t
r
oni
c
J
our
nal
, 2024, doi
:
10.2139/
s
s
r
n.5032985.
[
4]
A
.
A
l
-
H
a
w
a
m
l
e
h,
“
C
ybe
r
r
e
s
i
l
i
e
nc
e
f
r
a
m
e
w
or
k:
s
t
r
e
ngt
he
ni
ng
de
f
e
ns
e
s
a
nd
e
nha
nc
i
ng
c
ont
i
nui
t
y
i
n
bus
i
ne
s
s
s
e
c
ur
i
t
y,”
I
nt
e
r
nat
i
onal
J
our
nal
of
C
om
put
i
ng and D
i
gi
t
al
Sy
s
t
e
m
s
, vol
. 15, no. 1, pp. 1315
–
1331, M
a
r
. 2024, doi
:
10.12785/
i
j
c
ds
/
150193.
[
5]
M
.
K
r
i
s
pe
r
,
J
.
D
oba
j
,
a
nd
G
.
M
a
c
he
r
,
“
A
s
s
e
s
s
i
ng
r
i
s
k
e
s
t
i
m
a
t
i
ons
f
or
c
ybe
r
-
s
e
c
ur
i
t
y
us
i
ng
e
xpe
r
t
j
udgm
e
nt
,”
C
om
m
uni
c
at
i
ons
i
n
C
om
put
e
r
and I
nf
or
m
at
i
on Sc
i
e
nc
e
, pp. 120
–
134, 2020, doi
:
10.1007/
978
-
3
-
030
-
56441
-
4_9.
[
6]
D
.
I
t
a
ni
,
R
.
I
t
a
ni
,
A
.
A
.
E
l
t
w
e
r
i
,
A
.
F
a
c
c
i
a
,
a
nd
L
.
W
a
nga
noo,
“
E
nha
nc
i
ng
c
ybe
r
s
e
c
ur
i
t
y
t
hr
ough
c
om
pl
i
a
nc
e
a
nd
a
udi
t
i
ng:
a
s
t
r
a
t
e
gi
c
a
ppr
oa
c
h
t
o
r
e
s
i
l
i
e
n
c
e
,”
i
n
2024
2nd
I
nt
e
r
nat
i
onal
C
onf
e
r
e
nc
e
on
C
y
be
r
R
e
s
i
l
i
e
nc
e
(
I
C
C
R
)
,
F
e
b.
2024,
pp.
1
–
10
,
doi
:
10.1109/
I
C
C
R
61006.2024.10532959.
[
7]
J
.
S
t
r
a
ub,
“
A
s
s
e
s
s
m
e
nt
of
c
ybe
r
s
e
c
ur
i
t
y
c
om
pe
t
i
t
i
on
t
e
a
m
s
a
s
e
xpe
r
i
e
nt
i
a
l
e
d
uc
a
t
i
on
e
xe
r
c
i
s
e
s
,”
A
SE
E
A
nnual
C
onf
e
r
e
nc
e
and
E
x
pos
i
t
i
on, C
onf
e
r
e
nc
e
P
r
o
c
e
e
di
ngs
, 2020, doi
:
10.18260/
1
-
2
--
34187.
[
8]
A
.
N
a
s
s
a
r
a
nd
M
.
K
a
m
a
l
,
“
M
a
c
hi
ne
l
e
a
r
ni
ng
a
nd
bi
g
da
t
a
a
na
l
yt
i
c
s
f
or
c
y
be
r
s
e
c
ur
i
t
y
t
hr
e
a
t
de
t
e
c
t
i
on:
a
hol
i
s
t
i
c
r
e
vi
e
w
of
t
e
c
hni
que
s
a
nd
c
a
s
e
s
t
udi
e
s
,”
J
our
nal
of
A
r
t
i
f
i
c
i
al
I
nt
e
l
l
i
ge
nc
e
and
M
ac
hi
ne
L
e
ar
ni
ng
i
n
M
anage
m
e
nt
,
vol
.
5,
no.
1,
pp.
51
–
63,
2021
.
[
9]
F
.
E
kunda
yo,
I
.
A
t
oye
b,
A
.
S
oye
l
e
,
a
nd
E
.
O
gunw
obi
,
“
P
r
e
di
c
t
i
ve
a
na
l
yt
i
c
s
f
o
r
c
ybe
r
t
hr
e
a
t
i
nt
e
l
l
i
ge
nc
e
i
n
f
i
nt
e
c
h
u
s
i
ng
bi
g
d
a
t
a
a
nd
m
a
c
hi
ne
l
e
a
r
ni
ng
,”
I
nt
e
r
nat
i
onal
J
our
nal
of
R
e
s
e
ar
c
h
P
ubl
i
c
at
i
on
and
R
e
v
i
e
w
s
,
vol
.
5,
no.
11,
pp.
5934
–
5948,
2024,
doi
:
10.55248/
ge
ngpi
.5.1124.3352.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
.
14
, N
o.
5
,
O
c
to
be
r
2025
:
3897
-
3905
3904
[
10]
S
.
M
.
M
oha
m
m
a
d,
“
S
e
nt
i
m
e
nt
a
na
l
ys
i
s
:
de
t
e
c
t
i
ng
va
l
e
nc
e
,
e
m
ot
i
ons
,
a
n
d
ot
he
r
a
f
f
e
c
t
ua
l
s
t
a
t
e
s
f
r
om
t
e
xt
,”
i
n
E
m
ot
i
on
M
e
as
ur
e
m
e
nt
, E
l
s
e
vi
e
r
, 2016, pp. 201
–
237
,
doi
:
10.1016/
B
978
-
0
-
08
-
100508
-
8.00009
-
6.
[
11]
W
.
M
e
dha
t
,
A
.
H
a
s
s
a
n,
a
nd
H
.
K
or
a
s
hy,
“
S
e
nt
i
m
e
nt
a
na
l
ys
i
s
a
l
gor
i
t
hm
s
a
nd
a
ppl
i
c
a
t
i
ons
:
a
s
ur
ve
y,”
A
i
n
Sham
s
E
ngi
ne
e
r
i
ng
J
our
nal
, vol
. 5, no. 4, pp. 1093
–
1113, D
e
c
. 2014, doi
:
10.1016/
j
.a
s
e
j
.2014.04.0
11.
[
12]
D
.
A
.
B
ha
n
a
ge
a
nd
A
.
V
.
P
a
w
a
r
,
“
I
m
pr
ovi
ng
c
l
a
s
s
i
f
i
c
a
t
i
on
-
ba
s
e
d
l
og
a
na
l
ys
i
s
us
i
ng
ve
c
t
or
i
z
a
t
i
on
t
e
c
hni
que
s
,”
L
e
c
t
ur
e
N
ot
e
s
i
n
N
e
t
w
or
k
s
and Sy
s
t
e
m
s
, vol
. 612, pp. 271
–
282, 2023, doi
:
10.1007/
978
-
981
-
19
-
9228
-
5_24.
[
13]
H
. S
t
udi
a
w
a
n, F
. S
ohe
l
, a
nd C
. P
a
yne
, “
A
nom
a
l
y
de
t
e
c
t
i
on i
n ope
r
a
t
i
ng s
y
s
t
e
m
l
ogs
w
i
t
h de
e
p l
e
a
r
ni
ng
-
ba
s
e
d s
e
nt
i
m
e
nt
a
na
l
ys
i
s
,”
I
E
E
E
T
r
ans
ac
t
i
ons
on
D
e
pe
ndabl
e
and
Se
c
ur
e
C
om
put
i
ng
,
vol
.
1
8,
no.
5,
pp.
2136
–
2148,
S
e
p.
2021,
doi
:
10.1109/
T
D
S
C
.2020.3037903.
[
14]
M
.
P
a
r
i
m
a
l
a
,
R
.
M
.
S
.
P
r
i
ya
,
M
.
P
.
K
.
R
e
ddy,
C
.
L
.
C
how
dha
r
y,
R
.
K
.
P
ol
u
r
u,
a
nd
S
.
K
ha
n,
“
S
pa
t
i
ot
e
m
por
a
l
‐
ba
s
e
d
s
e
nt
i
m
e
nt
a
na
l
ys
i
s
on
t
w
e
e
t
s
f
or
r
i
s
k
a
s
s
e
s
s
m
e
nt
of
e
ve
nt
us
i
ng
d
e
e
p
l
e
a
r
ni
ng
a
ppr
oa
c
h,”
Sof
t
w
ar
e
:
P
r
ac
t
i
c
e
and
E
x
pe
r
i
e
nc
e
,
vol
.
51,
no.
3
,
pp. 550
–
570, M
a
r
. 2021, doi
:
10.1002/
s
pe
.2851.
[
15]
P
.
H
a
n,
H
.
L
i
,
G
.
X
ue
,
a
nd
C
.
Z
ha
ng,
“
D
i
s
t
r
i
but
e
d
s
ys
t
e
m
a
nom
a
l
y
de
t
e
c
t
i
on
us
i
ng
de
e
p
l
e
a
r
ni
ng‐
ba
s
e
d
l
og
a
na
l
ys
i
s
,
”
C
om
put
at
i
onal
I
nt
e
l
l
i
ge
nc
e
, vol
. 39, no. 3, pp. 433
–
455, J
un. 2023, doi
:
10.111
1/
c
oi
n.12573.
[
16]
G
. A
l
M
a
ha
di
n,
M
. O
.
H
i
a
r
i
, A
. H
. H
u
s
s
e
i
n, N
. M
.
M
. T
ur
a
b, A
.
A
l
khr
e
s
he
h,
a
n
d M
. A
.
B
. A
l
-
T
a
r
a
w
ne
h, “
P
e
r
f
or
m
a
nc
e
e
va
l
ua
t
i
on
of
a
n
i
nt
e
l
l
i
ge
nt
a
nd
opt
i
m
i
z
e
d
m
a
c
hi
ne
l
e
a
r
ni
ng
f
r
a
m
e
w
or
k
f
or
a
t
t
a
c
k
de
t
e
c
t
i
on
,”
I
nt
e
r
nat
i
onal
J
our
nal
of
C
om
m
uni
c
at
i
on
N
e
t
w
or
k
s
and I
nf
or
m
at
i
on Se
c
u
r
i
t
y
, vol
. 14, no. 3, pp. 358
–
371, 2022.
[
17]
A
.
S
r
i
va
s
t
a
va
,
V
.
S
r
i
va
s
t
a
va
,
K
.
K
um
a
r
,
S
.
S
r
i
va
s
t
a
va
,
a
nd
N
.
G
a
r
g,
“
H
ybr
i
d
m
a
c
hi
ne
l
e
a
r
ni
ng
m
e
t
hod
f
or
s
e
nt
i
m
e
nt
a
na
l
ys
i
s
,”
i
n
3r
d
I
nt
e
r
nat
i
onal
C
onf
e
r
e
nc
e
on
I
nnov
at
i
v
e
M
e
c
hani
s
m
s
f
o
r
I
ndus
t
r
y
A
ppl
i
c
at
i
ons
,
I
C
I
M
I
A
2023
-
P
r
oc
e
e
di
ngs
,
D
e
c
.
2023,
pp
.
646
–
652
,
doi
:
10.1109/
I
C
I
M
I
A
60377.2023.10426420.
[
18]
S
.
B
a
ya
t
a
nd
G
.
I
ş
i
k,
“
E
va
l
ua
t
i
ng
t
he
e
f
f
e
c
t
i
ve
ne
s
s
of
di
f
f
e
r
e
nt
m
a
c
hi
ne
l
e
a
r
n
i
ng
a
ppr
oa
c
he
s
f
or
s
e
nt
i
m
e
nt
c
l
a
s
s
i
f
i
c
a
t
i
on
,”
I
ğdı
r
Ü
ni
v
e
r
s
i
t
e
s
i
F
e
n B
i
l
i
m
l
e
r
i
E
ns
t
i
t
üs
ü D
e
r
gi
s
i
, vol
. 13, no. 3, pp. 1496
–
1510, S
e
p
. 2023, doi
:
10.21597/
j
i
s
t
.1292050.
[
19]
J
.
W
u
a
nd
J
.
X
i
a
o,
“
A
ppl
i
c
a
t
i
on
o
f
na
t
ur
a
l
l
a
ngua
ge
pr
oc
e
s
s
i
ng
i
n
ne
t
w
or
k
s
e
c
ur
i
t
y
l
og
a
na
l
ys
i
s
,”
J
our
nal
of
C
om
put
e
r
T
e
c
hnol
ogy
and A
ppl
i
e
d M
at
he
m
at
i
c
s
, vol
. 1, no. 3, pp. 39
–
47, 2024, doi
:
10.5281/
z
e
nodo.13366745.
[
20]
C
.
A
l
m
odova
r
,
F
.
S
a
br
i
na
,
S
.
K
a
r
i
m
i
,
a
nd
S
.
A
z
a
d,
“
L
ogF
i
T
:
l
og
a
nom
a
l
y
d
e
t
e
c
t
i
on
us
i
ng
f
i
ne
-
t
une
d
l
a
ngua
ge
m
ode
l
s
,”
I
E
E
E
T
r
ans
ac
t
i
ons
on N
e
t
w
or
k
and Se
r
v
i
c
e
M
anage
m
e
nt
, vol
. 21, no. 2, pp. 1715
–
17
23, A
pr
. 2024, doi
:
10.1109/
T
N
S
M
.2024.3358730.
[
21]
T
. A
. P
ha
m
a
nd J
. H
. L
e
e
, “
T
r
a
ns
S
e
nt
L
og:
i
nt
e
r
pr
e
t
a
bl
e
a
nom
a
l
y de
t
e
c
t
i
on u
s
i
n
g t
r
a
ns
f
or
m
e
r
a
nd s
e
nt
i
m
e
nt
a
na
l
y
s
i
s
on i
ndi
vi
dua
l
l
og e
ve
nt
,”
I
E
E
E
A
c
c
e
s
s
, vol
. 11, pp. 96272
–
96282, 2023, doi
:
10.1109/
A
C
C
E
S
S
.2023.3311146.
[
22]
A
.
B
ooke
r
,
V
.
C
hi
u,
N
.
G
r
of
f
,
a
nd
V
.
J
.
R
i
c
ha
r
ds
on,
“
A
I
S
r
e
s
e
a
r
c
h
oppor
t
uni
t
i
e
s
ut
i
l
i
z
i
ng
m
a
c
hi
ne
l
e
a
r
ni
ng:
f
r
o
m
a
m
e
t
a
-
t
he
or
y
of
a
c
c
ount
i
ng
l
i
t
e
r
a
t
ur
e
,”
I
nt
e
r
nat
i
onal
J
ou
r
nal
of
A
c
c
ount
i
ng
I
nf
or
m
at
i
on
Sy
s
t
e
m
s
,
vol
.
52,
M
a
r
.
2024,
doi
:
10.1016/
j
.a
c
c
i
nf
.2023.100661.
[
23]
O
.
A
bua
l
gha
na
m
,
O
.
A
dw
a
n,
M
.
A
.
A
l
S
ha
r
i
a
h,
a
nd
M
.
Q
a
t
a
w
ne
h,
“
E
nha
n
c
i
ng
t
he
s
pe
e
d
of
t
he
l
e
a
r
ni
ng
ve
c
t
or
qua
nt
i
z
a
t
i
on
(
L
V
Q
)
a
l
gor
i
t
hm
by
a
ddi
ng
pa
r
t
i
a
l
di
s
t
a
nc
e
c
om
put
a
t
i
on
,”
C
y
be
r
ne
t
i
c
s
and
I
n
f
or
m
at
i
on
T
e
c
hnol
ogi
e
s
,
vol
.
22,
no.
2,
pp.
36
–
49,
2022, doi
:
10.2478/
c
a
i
t
-
2022
-
0015.
[
24]
M
.
A
.
A
l
s
h
a
r
a
i
a
h
e
t
al
.
,
“
N
e
ur
a
l
ne
t
w
or
k
pr
e
di
c
t
i
on
m
ode
l
t
o
e
xpl
or
e
c
o
m
pl
e
x
nonl
i
ne
a
r
be
ha
vi
or
i
n
dyna
m
i
c
bi
ol
ogi
c
a
l
ne
t
w
or
k
,”
I
nt
e
r
nat
i
onal
J
our
nal
of
I
nt
e
r
ac
t
i
v
e
M
obi
l
e
T
e
c
hnol
ogi
e
s
,
vol
.
16,
no.
12,
pp.
32
–
51,
2022,
doi
:
10.3991/
i
j
i
m
.v16i
12.30467.
[
25]
M
.
S
.
U
.
S
our
a
v
e
t
al
.
,
“
T
r
a
ns
f
or
m
e
r
-
ba
s
e
d
t
e
xt
c
l
a
s
s
i
f
i
c
a
t
i
on
on
uni
f
i
e
d
b
a
ngl
a
m
ul
t
i
-
c
l
a
s
s
e
m
ot
i
on
c
or
pus
,”
i
n
2024
25t
h
I
nt
e
r
nat
i
onal
A
r
ab
C
onf
e
r
e
nc
e
on
I
nf
or
m
at
i
on
T
e
c
hnol
ogy
,
A
C
I
T
2024
,
D
e
c
.
2024,
pp.
1
–
7
,
doi
:
10.1109/
A
C
I
T
62805.2024.10877210.
[
26]
R
.
K
.
H
a
l
de
r
,
M
.
N
.
U
dd
i
n,
M
.
A
.
U
d
di
n,
S
.
A
r
ya
l
,
a
n
d
A
.
K
h
r
a
i
s
a
t
,
“
E
n
ha
nc
i
n
g
k
-
n
e
a
r
e
s
t
ne
i
gh
b
o
r
a
l
g
o
r
i
t
hm
:
a
c
o
m
p
r
e
he
ns
i
v
e
r
e
v
i
e
w
a
n
d
pe
r
f
o
r
m
a
n
c
e
a
na
l
y
s
i
s
o
f
m
o
d
i
f
i
c
a
t
i
o
ns
,”
J
our
n
al
o
f
B
i
g
D
a
t
a
,
v
ol
.
1
1,
n
o
.
1,
A
ug
.
2
0
24
,
d
oi
:
1
0.
1
18
6
/
s
4
05
3
7
-
0
24
-
0
09
73
-
y.
[
27]
M
.
H
.
I
br
a
hi
m
,
E
.
A
.
B
a
dr
a
n,
a
nd
M
.
H
.
A
bde
l
-
R
a
hm
a
n,
“
D
e
t
e
c
t
,
c
l
a
s
s
i
f
y,
a
n
d
l
oc
a
t
e
f
a
ul
t
s
i
n
D
C
m
i
c
r
ogr
i
ds
ba
s
e
d
on
s
uppor
t
ve
c
t
or
m
a
c
hi
ne
s
a
nd
ba
gge
d
t
r
e
e
s
i
n
t
he
m
a
c
hi
ne
l
e
a
r
ni
ng
a
ppr
oa
c
h
,
”
I
E
E
E
A
c
c
e
s
s
,
vol
.
12,
pp.
139199
–
139224,
2024,
doi
:
10.1109/
A
C
C
E
S
S
.2024.3466652.
[
28]
F
.
B
i
n,
S
.
H
os
s
e
i
ni
,
J
.
C
he
n,
P
.
S
a
m
ui
,
H
.
F
a
t
t
a
hi
,
a
nd
D
.
J
a
he
d
A
r
m
a
gha
ni
,
“
P
r
opos
i
ng
opt
i
m
i
z
e
d
r
a
ndom
f
or
e
s
t
m
ode
l
s
f
or
pr
e
di
c
t
i
ng
c
om
pr
e
s
s
i
ve
s
t
r
e
ngt
h
of
ge
opol
ym
e
r
c
om
pos
i
t
e
s
,”
I
nf
r
as
t
r
uc
t
ur
e
s
,
vol
.
9,
no.
10,
2024,
doi
:
10.3390/
i
nf
r
a
s
t
r
uc
t
ur
e
s
9100181.
[
29]
M
.
C
.
H
i
noj
os
a
L
e
e
,
J
.
B
r
a
e
t
,
a
nd
J
.
S
pr
i
nga
e
l
,
“
P
e
r
f
or
m
a
nc
e
m
e
t
r
i
c
s
f
or
m
ul
t
i
l
a
be
l
e
m
ot
i
on
c
l
a
s
s
i
f
i
c
a
t
i
on:
c
om
pa
r
i
ng
m
i
c
r
o,
m
a
c
r
o, a
nd w
e
i
ght
e
d
F1
-
s
c
or
e
s
,”
A
ppl
i
e
d Sc
i
e
nc
e
s
,
vol
. 14, no. 21, 2024, doi
:
10.3390/
a
pp14219863.
[3
0
]
N
.
R
a
j
a
gukguk,
I
.
P
.
E
.
N
.
K
e
nc
a
na
,
a
nd
I
.
G
.
N
.
L
.
W
.
K
us
um
a
,
“
A
ppl
i
c
a
t
i
on
of
t
e
r
m
f
r
e
que
nc
y
-
i
nve
r
s
e
doc
um
e
nt
f
r
e
que
nc
y
i
n
t
he
N
a
i
ve
B
a
ye
s
a
l
gor
i
t
hm
f
or
C
ha
t
G
P
T
us
e
r
s
e
nt
i
m
e
nt
a
na
l
ys
i
s
,”
P
r
oc
e
e
di
ngs
of
t
he
F
i
r
s
t
I
nt
e
r
nat
i
onal
C
onf
e
r
e
n
c
e
on
A
ppl
i
e
d
M
at
he
m
at
i
c
s
, St
at
i
s
t
i
c
s
, and C
om
put
i
ng (
I
C
A
M
SA
C
2023)
,
pp. 29
–
40, 2024, doi
:
10.2991/
978
-
94
-
6463
-
413
-
6_4.
B
I
O
G
R
A
P
H
I
E
S
O
F
A
U
T
H
O
R
S
Nidal
Turab
is
Ph.D.
in
computer
science
Professor
at
the
Netwo
rks
and
Cyber
Secur
ity
Depar
tment,
Al
-
Ahliyya
Amman
University,
Jordan.
His
re
search
interests
include
WLAN
security,
computer
networks
security
and
cloud
computing
security,
eLearning,
and
internet of
things
. He can be contac
ted at email:
n
.turab@
ammanu.edu.jo.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
A
dv
anc
e
d r
is
k
a
s
s
e
s
s
m
e
nt
us
in
g m
ac
hi
ne
l
e
a
r
ni
ng and s
e
nt
im
e
nt
analy
s
is
on l
og dat
a
(
N
id
al
T
ur
ab
)
3905
Abdelrahman
Abushattal
is
a
IEEE
student
member
that
receiv
ed
a
B.Sc.
in
communi
cation
engineerin
g
from
Al
-
Hussein
Bin
Talal
University
in
2012
and
an
M.Sc.
from
Mutah
University
in
2016.
He
is
pursuing
a
Master’s
in
Cyberse
curit
y
at
Al
-
Ahliyya
Amma
n
University,
Jordan,
and
a
Ph.D.
in
Electrical
and
Electronics
En
gineering
at
Karadeniz
Technical
University,
Trabzon,
Turkey.
Res
ea
rc
h
int
er
es
ts
in
cl
ud
e
p
hys
ic
al
-
la
yer
sec
ur
ity
,
pow
er
-
do
mai
n
NOM
A,
or
th
ogo
na
l
tim
e
-
fre
qu
enc
y
space,
spatial
mo
dulation,
robotic
design,
AI,
fluid
control
networks,
natural
language
processing,
and
comp
uter
vision.
He
can
be
contacted
at email
:
ceabushatt
al@
gmail.co
m.
Jamal
Al
-
Nabulsi
is
Ph.D.
in
Biomedical
Engineering,
Professor
at
the
Medical
Engineering
Department,
Al
-
Ahliyya
Amman
University,
Jordan.
H
is
research
interests
are
biomedical
sensors,
digital
signal
processing,
and
image
processing.
He
can
be
contacted
at
email:
j.nabulsi@ammanu.edu.jo
.
Hamza
Abu
Owida
is
Ph.D.
in
Biomedical
Engineering,
Assistant
Professor
at
the
Medical
Engineering
Department,
Al
-
Ahliyya
Amman
Unive
rsity,
Jordan.
Research
interests
focused
on
biomedical
sensors,
nanotechn
ology,
and
tissue
engineer
ing.
He
can
be
contacted
at email
:
h.abuowida@
ammanu.edu.jo.
Evaluation Warning : The document was created with Spire.PDF for Python.