I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
p
ute
r
E
ng
in
ee
ring
(
I
J
E
CE
)
Vo
l.
10
,
No
.
5
,
Octo
b
er
2
0
2
0
,
p
p
.
5
5
2
6
~
5
5
3
4
I
SS
N:
2088
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v
10
i
5
.
pp
5
5
2
6
-
5
5
3
4
5526
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m/in
d
ex
.
p
h
p
/I
JE
C
E
Senti
m
en
tal
c
la
ss
i
ficatio
n
a
na
ly
sis
of
p
o
la
rity
m
ul
ti
-
v
i
e
w
t
ex
tual
d
a
ta
u
sing
d
a
ta
m
ining
t
ech
niques
Ali H
a
m
ee
d
Y
a
s
s
ir
1
,
Al
i A
.
M
o
ha
m
m
e
d
2
,
Adel
Ab
du
l
-
J
a
bb
a
r
Al
k
ha
zr
a
j
i
3
,
M
us
t
a
f
a
E
m
a
d
H
a
m
ee
d
4
,
M
o
ha
m
m
e
d Sa
a
d T
a
lib
5
,
M
o
ha
na
d F
a
eq
Ali
6
1
C
o
ll
e
g
e
o
f
Co
m
p
u
ter S
c
ien
c
e
a
n
d
I
n
f
o
rm
a
ti
o
n
T
e
c
h
n
o
lo
g
y
,
Un
iv
e
rsity
o
f
S
u
m
e
r
,
Ira
q
2,
6
F
a
c
u
lt
y
o
f
In
f
o
r
m
a
ti
o
n
a
n
d
C
o
m
m
u
n
ica
ti
o
n
T
e
c
h
n
o
lo
g
y
,
Un
iv
e
r
siti
T
e
k
n
ik
a
l
M
a
la
y
sia
M
e
lak
a
(U
T
e
M
)
,
M
a
la
y
sia
3
Co
m
p
u
ter
S
c
ien
c
e
De
p
a
rtm
e
n
t,
Co
ll
e
g
e
o
f
S
c
ie
n
c
e
,
Un
iv
e
rsit
y
o
f
Di
y
a
la,
Ira
q
4
F
a
c
u
lt
y
o
f
El
e
c
tro
n
ic an
d
C
o
m
p
u
ter
En
g
in
e
e
ri
n
g
,
Un
iv
e
rsiti
T
e
k
n
ik
a
l
M
a
la
y
sia
M
e
la
k
a
(U
T
e
M
),
M
a
la
y
sia
5
Co
ll
e
g
e
o
f
A
d
m
in
istratio
n
a
n
d
E
c
o
n
o
m
ics
,
Un
iv
e
rsit
y
o
f
Ba
b
y
lo
n
,
Ira
q
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
No
v
28
,
2
0
1
9
R
ev
i
s
ed
A
p
r
2
5
,
2
0
2
0
A
cc
ep
ted
Ma
y
7
,
2
0
2
0
T
h
e
d
a
ta
a
n
d
in
f
o
r
m
a
ti
o
n
a
v
a
il
a
b
le
in
m
o
st
c
o
m
m
u
n
it
y
e
n
v
iro
n
m
e
n
ts
is
c
o
m
p
lex
in
n
a
tu
re
.
S
e
n
ti
m
e
n
tal
d
a
ta res
o
u
rc
e
s
m
a
y
p
o
ss
ib
ly
c
o
n
sist o
f
tex
tu
a
l
d
a
ta
c
o
ll
e
c
ted
f
ro
m
m
u
lt
ip
l
e
in
f
o
r
m
a
ti
o
n
so
u
rc
e
s
w
it
h
d
iff
e
re
n
t
re
p
re
se
n
tatio
n
s
a
n
d
u
su
a
ll
y
h
a
n
d
led
b
y
d
iffer
e
n
t
a
n
a
l
y
ti
c
a
l
m
o
d
e
ls.
T
h
e
se
t
y
p
e
s
o
f
d
a
ta
re
so
u
rc
e
c
h
a
ra
c
teristics
c
a
n
f
o
r
m
m
u
lt
i
-
v
ie
w
p
o
larity
tex
tu
a
l
d
a
ta.
Ho
w
e
v
e
r,
k
n
o
w
led
g
e
c
re
a
ti
o
n
f
ro
m
th
is
t
y
p
e
o
f
s
e
n
ti
m
e
n
tal
tex
tu
a
l
d
a
ta
re
q
u
ires
c
o
n
si
d
e
r
a
b
le
a
n
a
ly
ti
c
a
l
e
ff
o
rts
a
n
d
c
a
p
a
b
il
it
ies
.
In
p
a
rti
c
u
lar
,
d
a
ta
m
in
in
g
p
ra
c
ti
c
e
s
c
a
n
p
ro
v
id
e
e
x
c
e
p
ti
o
n
a
l
re
su
lt
s
in
h
a
n
d
l
in
g
tex
tu
a
l
d
a
ta f
o
rm
a
ts.
Be
sid
e
s,
in
t
h
e
c
a
se
o
f
th
e
tex
tu
a
l
d
a
ta ex
ists as
m
u
lt
i
-
v
ie
w
o
r
u
n
stru
c
t
u
re
d
d
a
ta
f
o
r
m
a
ts,
th
e
h
y
b
rid
a
n
d
in
teg
ra
ted
a
n
a
ly
si
s
e
ff
o
rts
o
f
tex
t
d
a
ta
m
in
in
g
a
lg
o
rit
h
m
s
a
re
v
it
a
l
to
g
e
t
h
e
l
p
f
u
l
re
su
lt
s
.
T
h
e
o
b
jec
ti
v
e
o
f
th
is
re
se
a
rc
h
is
to
e
n
h
a
n
c
e
th
e
k
n
o
w
led
g
e
d
isc
o
v
e
r
y
f
ro
m
se
n
ti
m
e
n
tal
m
u
lt
i
-
v
ie
w
te
x
tu
a
l
d
a
ta
w
h
ich
c
a
n
b
e
c
o
n
sid
e
re
d
a
s
u
n
stru
c
tu
re
d
d
a
ta
f
o
r
m
a
t
to
c
las
si
fy
th
e
p
o
larity
in
f
o
r
m
a
ti
o
n
d
o
c
u
m
e
n
ts
in
th
e
fo
rm
o
f
t
w
o
d
if
fe
re
n
t
c
a
te
g
o
ries
o
r
t
y
p
e
s
o
f
u
se
f
u
l
in
f
o
r
m
a
ti
o
n
.
A
p
ro
p
o
se
d
f
ra
m
e
w
o
rk
w
it
h
in
teg
ra
ted
d
a
ta
m
in
in
g
a
lg
o
rit
h
m
s
h
a
s
b
e
e
n
d
isc
u
ss
e
d
i
n
t
h
is
p
a
p
e
r,
w
h
ich
is
a
c
h
ie
v
e
d
th
ro
u
g
h
th
e
a
p
p
li
c
a
ti
o
n
o
f
X
-
m
e
a
n
s
a
l
g
o
rit
h
m
f
o
r
c
lu
ste
rin
g
a
n
d
Ho
tS
p
o
t
a
lg
o
ri
th
m
o
f
a
ss
o
c
iatio
n
ru
les
.
T
h
e
a
n
a
ly
sis
re
su
lt
s
h
a
v
e
sh
o
w
n
i
m
p
ro
v
e
d
a
c
c
u
ra
c
ie
s
o
f
c
las
sify
in
g
th
e
se
n
ti
m
e
n
tal
m
u
lt
i
-
v
ie
w
tex
tu
a
l
d
a
ta
in
to
tw
o
c
a
teg
o
ries
th
ro
u
g
h
t
h
e
a
p
p
li
c
a
ti
o
n
o
f
th
e
p
ro
p
o
se
d
f
ra
m
e
w
o
rk
o
n
o
n
li
n
e
p
o
larity
u
se
r
-
re
v
ie
w
s d
a
tas
e
t
u
p
o
n
a
g
iv
e
n
t
o
p
ics
.
K
ey
w
o
r
d
s
:
Data
m
i
n
i
n
g
Data
s
cien
ce
P
o
lar
ity
t
e
x
t
d
ata
Sen
ti
m
e
n
t a
n
a
l
y
s
is
T
ex
t
d
ata
m
i
n
i
n
g
Co
p
y
rig
h
t
©
2
0
2
0
In
stit
u
te o
f
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
rin
g
a
n
d
S
c
ien
c
e
.
Al
l
rig
h
ts re
se
rv
e
d
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
A
li
A
b
d
u
l
-
J
ab
b
ar
Mo
h
a
m
m
ed
,
Facu
lt
y
o
f
I
n
f
o
r
m
a
tio
n
a
n
d
C
o
m
m
u
n
icatio
n
T
ec
h
n
o
lo
g
y
,
Un
i
v
er
s
iti T
ek
n
ik
al
Ma
la
y
s
ia
Me
lak
a
(U
T
e
M
)
,
Han
g
T
u
ah
J
a
y
a,
7
6
1
0
0
Du
r
ian
T
u
n
g
g
al,
Me
la
k
a,
Ma
la
y
s
ia.
E
m
ail: a
liab
d
u
lj
ab
ar
2
2
@
y
ah
o
o
.
co
m
1.
I
NT
RO
D
UCT
I
O
N
I
n
th
e
d
i
g
ital
a
g
e
o
f
t
h
e
r
ap
i
d
s
h
i
f
t
to
in
f
o
r
m
a
tio
n
er
a,
e
n
o
r
m
o
u
s
a
m
o
u
n
ts
o
f
tex
t
u
al
d
ata
h
av
e
ac
h
iev
ed
b
y
th
e
p
ar
ticip
ati
n
g
i
n
v
ir
t
u
al
co
m
m
u
n
i
ties
o
n
b
eh
al
f
o
f
s
h
ar
i
n
g
d
at
a
an
d
in
f
o
r
m
atio
n
,
s
o
cial
co
m
m
u
n
icatio
n
s
a
n
d
d
ail
y
tr
an
s
ac
tio
n
s
[
1
]
.
Ma
n
u
al
l
y
p
r
o
ce
s
s
in
g
o
f
t
h
i
s
h
u
g
e
a
m
o
u
n
t
o
f
u
n
s
tr
u
ct
u
r
ed
tex
t
u
al
d
ata
an
d
in
f
o
r
m
at
io
n
i
s
v
er
y
co
s
t
l
y
i
n
ter
m
o
f
ti
m
e
-
co
n
s
u
m
i
n
g
an
d
ex
p
e
n
s
i
v
e
p
r
o
ce
s
s
.
A
cc
o
r
d
in
g
l
y
,
tak
i
n
g
ad
v
a
n
ta
g
es
o
f
t
h
i
s
t
y
p
e
o
f
d
ata
an
d
in
f
o
r
m
atio
n
i
s
v
er
y
h
ar
d
f
o
r
t
h
e
r
ea
s
o
n
o
f
its
av
ailab
i
lit
y
i
n
v
ar
io
u
s
f
o
r
m
s
s
u
c
h
as
u
n
s
tr
u
ctu
r
ed
o
r
s
e
m
i
-
s
tr
u
ct
u
r
ed
tex
t
d
ata
f
o
r
m
s
[2
,
3]
.
T
h
e
m
aj
o
r
it
y
o
f
in
d
ex
ed
r
eso
u
r
ce
s
o
f
th
is
te
x
t
u
al
d
ata
ar
e
m
ai
n
l
y
ex
t
r
ac
ted
f
r
o
m
o
n
li
n
e
r
ep
o
r
ts
an
d
d
o
cu
m
e
n
ts
,
e
-
m
ails
,
o
n
lin
e
te
x
t
u
al
b
lo
g
s
,
m
e
m
o
s
,
d
ig
i
tal
letter
s
,
etc
[
3
]
.
T
ex
tu
al
d
ata
r
eso
u
r
ce
s
m
a
y
p
o
s
s
ib
l
y
co
n
s
is
t
o
f
d
ata
co
lle
cted
f
r
o
m
m
u
ltip
l
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8708
S
en
timen
ta
l c
la
s
s
ifica
tio
n
a
n
a
lysi
s
o
f p
o
la
r
ity
mu
lti
-
view tex
tu
a
l d
a
ta
u
s
in
g
d
a
ta
.
..
(
A
li Ha
mee
d
Ya
s
s
ir
)
5527
in
f
o
r
m
atio
n
co
n
tr
ib
u
to
r
s
w
it
h
d
i
f
f
er
e
n
t
r
ep
r
esen
tatio
n
s
a
n
d
u
s
u
a
ll
y
h
a
n
d
led
b
y
h
eter
o
g
en
eo
u
s
a
n
al
y
tical
m
o
d
el
s
.
T
h
is
d
i
v
er
s
it
y
in
th
e
t
ex
tu
a
l d
ata
s
o
u
r
ce
s
a
n
d
p
r
o
ce
d
u
r
es
f
o
r
m
s
th
e
s
e
n
ti
m
e
n
t
m
u
l
ti
-
v
ie
w
tex
t
u
al
d
ata.
T
h
is
t
y
p
e
o
f
d
ata
co
m
p
r
i
s
e
u
n
iq
u
e
i
n
f
o
r
m
atio
n
f
ea
t
u
r
es
th
at
in
v
o
lv
e
in
f
o
r
m
a
tio
n
h
a
v
e
t
h
e
s
a
m
e
i
n
s
ta
n
ce
b
u
t
w
it
h
v
ar
io
u
s
r
ep
r
esen
tatio
n
s
o
r
h
av
e
d
if
f
er
e
n
t
i
n
s
ta
n
ce
s
with
s
a
m
e
r
ep
r
ese
n
tatio
n
[
4
]
.
T
h
u
s
,
th
e
e
f
f
icie
n
t
u
n
d
er
s
ta
n
d
in
g
o
f
s
en
ti
m
e
n
t
m
u
lti
-
v
ie
w
te
x
tu
a
l
d
ata
is
b
ec
o
m
i
n
g
a
n
ex
tr
e
m
el
y
cr
iti
ca
l
m
o
v
e
m
e
n
t
b
y
t
h
e
w
a
y
o
f
u
s
in
g
d
ata
m
in
i
n
g
m
eth
o
d
s
.
T
h
e
m
ai
n
p
r
o
m
is
e
o
f
d
ata
m
i
n
in
g
to
o
ls
a
n
d
tech
n
iq
u
e
s
t
o
tak
e
ad
v
an
tag
e
o
f
m
u
lti
-
v
ie
w
tex
tu
al
d
ata
i
s
th
e
p
o
ten
tial
s
p
ee
d
u
p
o
f
b
ig
d
ata
er
a
[
5
]
.
Var
io
u
s
m
et
h
o
d
s
an
d
s
tr
ateg
ie
s
o
f
d
ata
an
al
y
s
is
ar
e
p
r
o
p
o
s
ed
f
o
r
d
eli
v
er
in
g
o
n
l
in
e
a
n
d
/o
r
in
te
g
r
ate
d
s
o
lu
tio
n
s
to
s
o
l
v
e
co
m
p
le
x
it
ies
ass
o
ciate
d
w
i
th
th
e
m
a
n
a
g
e
m
en
t
o
f
v
ar
io
u
s
d
ata
f
o
r
m
s
i
n
s
e
n
ti
m
e
n
t
m
u
lt
i
-
v
i
e
w
te
x
t
u
al
d
ata.
T
h
e
m
i
x
ed
m
eth
o
d
s
an
d
lik
e
w
is
e
th
e
h
y
b
r
id
m
et
h
o
d
s
o
f
te
x
t
d
a
ta
m
in
i
n
g
tec
h
n
o
lo
g
ies
m
a
y
b
e
class
if
ied
a
s
t
h
e
m
o
d
er
n
p
r
a
ctice
f
o
r
a
n
al
y
zi
n
g
s
en
ti
m
e
n
t
m
u
lti
-
v
ie
w
tex
t
u
al
d
ata
[
6
]
.
T
ex
tu
al
Data
Min
in
g
i
s
“
r
o
u
g
h
l
y
eq
u
i
v
alen
t
to
te
x
t
a
n
al
y
tics
,
r
ef
er
s
to
t
h
e
p
r
o
ce
s
s
o
f
d
er
iv
in
g
h
ig
h
-
q
u
al
it
y
in
f
o
r
m
a
tio
n
f
r
o
m
te
x
t”
[
7
]
.
Mo
r
eo
v
er
,
it
is
as
w
ell
co
n
s
id
er
ed
as
a
p
at
h
o
f
d
ata
m
in
i
n
g
f
ield
w
h
e
n
d
ata
m
i
n
i
n
g
al
g
o
r
ith
m
s
an
d
m
et
h
o
d
s
ar
e
ap
p
lied
f
o
r
th
e
d
is
co
v
er
y
o
f
k
n
o
w
led
g
e
in
t
h
e
tex
t
u
al
d
ata
f
o
r
m
s
.
Ma
n
y
o
f
tex
t
d
ata
m
i
n
in
g
tec
h
n
iq
u
e
s
ca
n
b
e
i
n
d
iv
id
u
all
y
ap
p
lied
o
r
in
co
r
p
o
r
atio
n
w
i
t
h
o
t
h
er
tech
n
iq
u
es
f
o
r
d
is
co
v
er
i
n
g
s
e
n
s
ib
le
an
d
v
alu
ab
le
k
n
o
w
led
g
e.
A
l
th
o
u
g
h
,
te
x
t d
ata
m
in
i
n
g
h
a
s
ac
h
ie
v
ed
v
ar
io
u
s
s
u
cc
e
s
s
f
u
l
ap
p
licatio
n
s
in
m
a
n
y
ar
ea
s
,
s
o
m
e
d
ee
p
s
e
n
ti
m
e
n
t
tex
t
a
n
al
y
s
i
s
tas
k
s
s
ti
ll
r
e
m
a
in
as
a
c
h
alle
n
g
e
a
n
d
n
o
t
h
an
d
led
y
et.
T
h
e
r
ea
l
d
i
s
tin
ctio
n
o
f
tec
h
n
o
lo
g
ical
ef
f
o
r
ts
w
it
h
tex
t
d
ata
m
in
i
n
g
f
r
o
m
t
h
e
g
e
n
er
al
co
n
ce
p
t
o
f
d
ata
m
i
n
i
n
g
is
th
e
p
o
s
s
ib
il
ities
to
d
ea
l
w
it
h
t
h
e
s
p
ec
ial
ch
ar
ac
ter
is
tic
s
o
f
i
m
p
licit
in
f
o
r
m
at
io
n
i
n
tex
tu
a
l
d
ata
th
at
s
ee
m
s
to
b
e
p
r
im
ar
il
y
u
n
s
tr
u
ctu
r
ed
.
Ho
w
e
v
er
,
th
e
ty
p
ica
l
p
r
ac
tice
o
f
s
u
ch
tech
n
o
lo
g
ical
p
o
s
s
ib
ilit
ie
s
is
y
et
in
t
h
e
ea
r
l
y
s
tag
e
s
to
r
ev
ea
l
t
h
e
n
ec
es
s
ar
y
k
n
o
w
led
g
e
t
h
at
ex
tr
ac
ted
f
r
o
m
t
h
is
u
n
s
tr
u
ctu
r
ed
tex
t
u
al
d
ata
[
8
]
.
T
h
er
ef
o
r
e,
th
e
k
e
y
as
p
ec
t
m
o
ti
v
ated
th
i
s
p
ap
er
is
th
at
m
u
l
ti
-
v
ie
w
te
x
t
u
al
d
ata
ca
n
b
e
ch
ar
ac
ter
ized
as
a
v
er
y
co
m
p
lex
i
m
p
licit
d
ata
s
tr
u
ct
u
r
e.
I
n
th
i
s
co
n
tex
t,
t
h
e
m
ai
n
p
r
o
b
lem
to
d
ea
l
w
it
h
th
i
s
t
y
p
e
o
f
d
ata
is
to
d
esig
n
h
y
b
r
id
tec
h
n
iq
u
es
o
f
d
ata
m
in
i
n
g
alg
o
r
it
h
m
s
t
h
a
t
ab
le
to
h
an
d
le
i
m
p
lici
t
d
ata
an
d
in
f
o
r
m
atio
n
.
A
d
d
itio
n
al
co
m
b
i
n
ed
ef
f
o
r
ts
o
f
k
n
o
w
led
g
e
d
is
co
v
er
y
ap
p
licatio
n
s
h
a
v
e
th
e
p
o
s
s
ib
ilit
ie
s
to
r
ev
ea
l
as
s
o
ciatin
g
co
m
m
o
n
i
n
f
o
r
m
atio
n
p
atter
n
s
in
t
h
e
m
u
lt
ip
le
v
ie
w
s
o
r
m
o
d
a
liti
es
o
f
tex
t
u
al
d
ata
f
o
r
m
at
s
.
A
t
P
r
esen
t,
th
er
e
is
f
e
w
er
ad
d
r
ess
i
n
g
m
et
h
o
d
s
o
f
d
ata
m
i
n
in
g
a
n
d
m
ac
h
i
n
e
lear
n
in
g
f
o
r
lear
n
i
n
g
f
r
o
m
t
h
ese
s
e
m
i
-
s
tr
u
ct
u
r
ed
o
r
u
n
s
tr
u
ct
u
r
ed
d
ata
t
y
p
es i
n
th
e
tex
t
u
al
d
ata
f
o
r
m
at
s
[
9
]
.
M.
J
ak
u
b
ik
[
1
0
]
h
as
s
u
m
m
ar
i
ze
d
k
n
o
w
led
g
e
cr
ea
tio
n
m
a
n
a
g
e
m
e
n
t
w
it
h
in
co
r
p
o
r
atio
n
o
f
g
ath
er
i
n
g
an
d
s
h
ar
in
g
i
n
f
o
r
m
atio
n
,
an
d
ef
f
ec
ti
v
el
y
d
ev
elo
p
in
g
t
h
i
s
i
n
f
o
r
m
atio
n
a
n
d
t
h
en
s
h
i
f
ti
n
g
th
e
k
n
o
w
led
g
e
to
d
em
a
n
d
er
s
.
T
h
is
r
e
s
ea
r
ch
i
s
g
o
in
g
to
in
v
est
ig
ate
th
e
tr
a
ck
o
u
t
lin
ed
b
y
M.
J
ak
u
b
ik
,
[
1
0
]
f
o
r
m
a
n
a
g
in
g
th
e
cr
ea
tio
n
o
f
k
n
o
w
led
g
e
b
u
t
th
e
n
w
ill
co
n
s
tr
u
ct
t
h
is
r
esear
ch
b
y
o
p
er
atin
g
t
h
e
to
o
ls
a
n
d
tech
n
iq
u
es
o
f
d
ata
m
i
n
in
g
i
n
m
u
lti
-
v
ie
w
te
x
tu
a
l
d
ata.
C
o
n
s
eq
u
en
tl
y
,
t
h
e
p
r
im
ar
y
p
u
r
p
o
s
e
o
f
co
n
d
u
ct
in
g
th
is
r
esea
r
ch
i
s
in
th
e
d
ir
ec
tio
n
o
f
f
i
n
d
o
u
t
i
n
wh
ich
w
a
y
ca
n
au
to
m
atica
l
l
y
cl
ass
i
f
y
k
n
o
w
led
g
e
a
n
d
m
o
r
e
ac
cu
r
atel
y
id
e
n
ti
f
ied
class
es
f
r
o
m
u
n
s
tr
u
c
tu
r
ed
tex
t
u
al
d
ata
d
o
cu
m
e
n
ts
.
C
o
n
s
eq
u
en
tl
y
,
i
t
w
ill
o
f
f
er
f
u
r
t
h
er
k
n
o
w
led
g
e
th
at
ca
n
b
e
r
ea
lized
,
im
p
r
o
v
ed
,
an
d
d
eliv
e
r
ed
to
th
e
d
em
a
n
d
er
s
.
2.
RE
L
AT
E
D
WO
RK
I
n
th
i
s
s
ec
tio
n
,
a
b
r
ief
d
escr
ip
tio
n
is
p
r
o
v
id
ed
to
ass
es
s
t
h
e
ex
is
t
in
g
ap
p
r
o
ac
h
es
o
f
te
x
t
u
al
s
en
ti
m
e
n
t
an
al
y
s
is
a
n
d
th
e
d
is
ti
n
ctio
n
b
et
w
ee
n
th
e
p
r
o
p
o
s
ed
f
r
am
e
w
o
r
k
a
n
d
r
elate
d
p
r
ev
io
u
s
w
o
r
k
s
.
T
h
e
tex
tu
al
s
en
ti
m
e
n
t
a
n
al
y
s
is
is
v
er
y
well
-
k
n
o
w
n
as
p
ar
t
o
f
d
ata
m
i
n
in
g
ap
p
r
o
ac
h
es,
w
h
ich
ca
n
ess
e
n
tiall
y
clas
s
i
f
y
th
e
s
e
n
ti
m
en
ta
l
m
u
lti
-
v
ie
w
te
x
tu
a
l
d
ata
in
to
t
w
o
ca
teg
o
r
ie
s
as
p
o
s
itiv
e
o
r
n
eg
ati
v
e
[
1
1
]
.
I
n
th
is
r
esear
ch
,
th
e
p
r
o
p
o
s
ed
f
r
a
m
e
w
o
r
k
i
s
ap
p
lied
o
n
o
n
lin
e
p
o
lar
it
y
u
s
er
-
r
ev
ie
w
s
to
b
e
clas
s
i
f
ied
as
‘
g
o
o
d
’
o
r
‘
b
ad
’
s
en
ti
m
e
n
ts
.
I
n
ac
tu
a
l
f
ac
t,
e
x
t
r
ac
tin
g
k
n
o
w
led
g
e
f
r
o
m
in
f
o
r
m
atio
n
a
v
ailab
le
i
n
tex
t
u
al
d
ata
is
in
cr
ea
s
in
g
l
y
g
ain
in
g
m
o
r
e
co
n
s
id
er
atio
n
.
T
h
is
co
n
s
id
er
atio
n
b
y
r
elate
d
p
r
ev
io
u
s
w
o
r
k
s
h
a
v
e
d
e
v
e
lo
p
ed
a
n
u
m
b
er
o
f
s
en
ti
m
e
n
t
a
n
al
y
s
i
s
m
et
h
o
d
s
to
m
ea
s
u
r
e
t
h
e
o
p
i
n
io
n
s
o
f
u
s
er
s
ab
o
u
t
p
r
o
d
u
cts
o
r
s
er
v
ice
s
.
As
e
x
a
m
p
le
o
f
ea
r
l
y
d
ev
elo
p
ed
m
et
h
o
d
s
,
R
ed
d
ick
et
al.
[
1
2
]
in
v
esti
g
at
ed
f
ea
t
u
r
es
o
f
te
x
t
i
n
w
h
ich
s
o
cial
m
e
d
ia
b
y
s
o
m
e
p
u
b
li
c
s
er
v
ices
d
eli
v
er
y
.
L
i
k
e
w
i
s
e,
Mü
ller
et
al.
[
1
3
]
h
ad
ap
p
lied
a
v
ar
iet
y
o
f
te
x
t
a
n
al
y
tic
ap
p
r
o
ac
h
es
to
u
n
d
er
s
ta
n
d
o
b
s
tacle
s
r
elate
d
to
cu
s
to
m
e
r
s
er
v
ice.
So
m
e
d
ata
m
i
n
i
n
g
ap
p
r
o
ac
h
es
h
av
e
i
n
v
o
lv
ed
in
th
e
k
n
o
w
led
g
e
d
is
co
v
er
y
f
r
o
m
s
e
n
ti
m
e
n
tal
m
u
lti
-
v
ie
w
tex
tu
al
d
ata,
s
u
c
h
as
f
ea
t
u
r
es
e
x
tr
ac
tio
n
o
f
o
n
lin
e
u
s
er
g
e
n
er
ated
co
n
ten
t
s
[
1
4
]
,
to
p
r
ed
ictio
n
a
n
d
clas
s
i
f
icatio
n
[
1
5
]
.
I
n
t
h
e
r
elate
d
r
esear
ch
o
f
u
s
er
’
s
r
e
v
ie
w
s
,
a
s
en
t
i
m
e
n
t
class
i
f
icatio
n
o
f
m
o
v
ie
r
ev
ie
ws
h
as
b
ee
n
h
a
n
d
led
b
y
Ah
u
j
a
et
al.
[
1
6
]
u
s
in
g
d
u
al
tr
ai
n
i
n
g
an
d
d
u
al
p
r
ed
ictio
n
w
h
ile
ad
d
r
ess
i
n
g
p
o
lar
it
y
s
h
if
t.
C
o
n
s
eq
u
e
n
tl
y
,
Z
o
la
e
t
al.
[
1
7
]
h
av
e
p
r
o
p
o
s
e
a
n
o
v
el
s
e
n
ti
m
en
t
clas
s
i
f
icatio
n
,
in
w
h
ic
h
a
th
r
ee
-
s
tep
m
e
th
o
d
o
lo
g
y
is
ex
p
lo
r
ed
b
ased
o
n
b
alan
ce
d
tr
ain
i
n
g
,
te
x
t
p
r
ep
r
o
c
ess
i
n
g
a
n
d
m
ac
h
in
e
lear
n
in
g
u
s
i
n
g
t
w
o
la
n
g
u
a
g
es
: E
n
g
l
is
h
an
d
I
talia
n
.
I
n
th
i
s
co
n
te
x
t,
th
e
d
ata
m
i
n
in
g
al
g
o
r
ith
m
s
ar
e
e
n
g
a
g
ed
s
tr
o
n
g
l
y
f
o
r
k
n
o
w
led
g
e
cr
ea
tio
n
f
r
o
m
s
en
ti
m
e
n
tal
tex
t
u
al
d
ata.
As
ex
a
m
p
le
s
o
f
u
s
i
n
g
d
ata
m
i
n
in
g
al
g
o
r
it
h
m
s
,
W
an
g
et
a
l.
[
1
8
]
in
tr
o
d
u
ce
d
a
co
n
v
o
lu
t
io
n
al
r
ec
u
r
r
e
n
t
n
e
u
r
al
n
et
w
o
r
k
f
o
r
class
i
f
y
i
n
g
tex
t
d
o
cu
m
en
t
s
,
u
s
in
g
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
i
n
e
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
10
,
No
.
5
,
Octo
b
e
r
2
0
2
0
:
5
5
2
6
-
5534
5528
(
SVM)
to
in
f
l
u
e
n
ce
o
f
w
o
r
d
n
o
r
m
aliza
tio
n
in
te
x
t
clas
s
i
f
icatio
n
[
1
9
,
2
0
]
,
u
s
in
g
n
ai
v
e
b
ay
e
s
f
o
r
o
p
ti
m
al
f
ea
t
u
r
e
s
elec
tio
n
i
n
te
x
t
ca
t
eg
o
r
izatio
n
b
y
T
an
g
et
al.
[
2
1
]
,
an
d
u
s
in
g
n
aiv
e
b
a
y
e
s
ian
a
lg
o
r
it
h
m
b
y
T
an
g
et
al.
[
2
2
]
f
o
r
te
x
t
u
n
s
tr
u
ctu
r
ed
cla
s
s
i
f
icat
io
n
i
n
t
h
e
S
p
ar
k
co
m
p
u
ti
n
g
e
n
v
ir
o
n
m
e
n
t.
E
v
e
n
t
h
o
u
g
h
s
o
m
e
o
f
th
o
s
e
s
o
l
u
tio
n
s
h
ap
p
en
to
b
e
h
ig
h
l
y
e
f
f
ec
tiv
e,
t
h
e
ch
all
en
g
e
s
co
n
ti
n
u
e
to
b
e
r
elev
an
t
to
th
e
v
ar
iatio
n
o
f
m
u
lti
-
v
ie
w
p
o
lar
it
y
te
x
t
u
al
d
ata
th
at
r
eq
u
ir
ed
p
r
ep
r
o
ce
s
s
in
g
,
an
d
d
ec
id
in
g
o
n
th
e
p
r
o
p
e
r
class
if
icatio
n
p
r
o
ce
d
u
r
e
[
2
3
]
.
T
h
er
ef
o
r
e,
th
is
r
esear
c
h
i
s
f
o
cu
s
i
n
g
o
n
th
e
d
if
f
ic
u
lt
y
o
f
p
r
o
v
id
in
g
i
m
p
r
o
v
ed
ac
cu
r
ac
ies
o
f
class
i
f
y
in
g
th
e
s
e
n
ti
m
e
n
tal
m
u
lti
-
v
ie
w
te
x
t
u
al
d
ata
in
t
o
t
w
o
ca
teg
o
r
ies
th
r
o
u
g
h
t
h
e
ap
p
licati
o
n
o
f
th
e
p
r
o
p
o
s
ed
f
r
am
e
w
o
r
k
o
n
o
n
lin
e
p
o
lar
it
y
u
s
er
-
r
e
v
ie
w
s
d
ata
s
et
u
p
o
n
a
g
iv
e
n
to
p
ics.
3.
RE
S
E
ARCH
M
E
T
H
O
D
T
h
e
m
et
h
o
d
s
tr
ateg
y
t
h
at
h
a
s
b
ee
n
d
esig
n
ed
f
o
r
th
i
s
p
ap
er
i
s
d
iv
id
ed
in
to
co
n
ce
p
tu
al
d
ev
elo
p
m
e
n
t
s
tag
e
o
f
t
h
e
p
r
o
p
o
s
ed
f
r
a
m
e
w
o
r
k
w
h
ic
h
r
elie
s
o
n
t
h
e
c
o
m
b
i
n
ed
to
o
ls
a
n
d
tech
n
iq
u
es
o
f
d
ata
m
i
n
in
g
.
T
h
is
co
m
b
in
atio
n
o
f
d
ata
m
i
n
in
g
alg
o
r
it
h
m
s
ca
n
s
er
v
e
th
e
p
u
r
p
o
s
e
o
f
h
a
n
d
li
n
g
u
n
s
tr
u
ctu
r
ed
tex
tu
al
d
ata
s
ets.
T
h
e
co
m
p
lete
p
r
o
ce
d
u
r
e
in
v
o
lv
ed
w
it
h
id
en
t
if
y
i
n
g
co
m
p
r
e
h
en
d
in
g
k
n
o
w
led
g
e
f
r
o
m
u
n
s
tr
u
ctu
r
ed
tex
tu
al
d
ata
s
o
u
r
ce
s
is
s
i
m
p
l
y
i
m
p
le
m
e
n
te
d
u
s
i
n
g
t
h
e
d
ir
ec
tio
n
o
u
tli
n
ed
in
a
m
atter
o
f
t
h
is
s
ec
t
io
n
.
T
h
e
f
o
llo
w
i
n
g
ac
tio
n
s
h
ad
b
ee
n
ex
p
er
ien
ce
d
to
i
m
p
r
o
v
e
an
d
r
ef
i
n
e
th
e
d
ev
e
lo
p
m
e
n
t o
f
t
h
e
f
r
a
m
e
w
o
r
k
d
esi
g
n
:
-
I
d
en
tif
icat
io
n
o
f
v
ar
io
u
s
s
o
u
r
c
e
in
f
o
r
m
atio
n
r
e
g
ar
d
in
g
te
x
t
u
a
l a
n
al
y
s
is
d
ata
m
i
n
in
g
.
-
E
v
alu
a
tio
n
an
d
an
al
y
zi
n
g
a
v
ailab
le
tech
n
iq
u
es
an
d
ap
p
r
o
ac
h
es
o
f
d
ata
m
i
n
in
g
f
o
r
an
al
y
s
i
n
g
tex
t
u
al
d
atab
ases
.
-
Dete
r
m
i
n
i
n
g
b
en
e
f
icial
d
ata
m
i
n
in
g
al
g
o
r
ith
m
s
to
d
esig
n
h
y
b
r
id
ized
f
r
a
m
e
w
o
r
k
f
o
r
an
al
y
s
i
n
g
te
x
t
u
al
d
atab
ases
.
T
h
e
n
ex
t
s
ta
g
e
i
s
th
e
d
e
s
i
g
n
an
d
im
p
l
em
en
ta
t
i
o
n
o
f
th
e
f
r
am
ew
o
r
k
t
o
o
b
t
a
in
t
h
e
r
e
s
u
l
ts
an
d
v
e
r
if
y
th
e
o
b
je
c
t
i
v
es
o
f
t
h
is
r
es
e
a
r
ch
.
3
.
1
.
Co
ncept
ua
l
dev
elo
p
m
ent
o
f
a
lg
o
rit
h
m
s
3
.
1
.
1
.
X
-
m
ea
ns
Cl
us
t
er
ing
Al
g
o
rit
h
m
T
h
e
X
-
Me
an
s
cl
u
s
ter
i
n
g
alg
o
r
ith
m
p
r
o
p
o
s
ed
b
y
P
elleg
an
d
Mo
o
r
e
[
2
4
]
is
a
K
-
Me
an
s
ex
te
n
d
ed
b
y
an
Stru
ct
u
r
e
I
m
p
r
o
v
ed
p
ar
t.
T
h
e
ce
n
tr
o
id
in
th
is
p
ar
t o
f
th
e
al
g
o
r
ith
m
tr
ie
s
to
b
e
d
iv
id
ed
au
to
m
atica
ll
y
i
n
its
ar
ea
.
T
h
e
j
o
in
d
ec
is
io
n
a
m
o
n
g
th
e
ch
ild
r
en
o
f
ea
c
h
ce
n
tr
o
id
w
it
h
t
h
e
ce
n
tr
o
id
it
s
elf
is
f
in
a
liz
ed
co
m
p
ar
in
g
w
it
h
th
e
B
a
y
esia
n
I
n
f
o
r
m
at
io
n
C
r
it
er
io
n
(
B
I
C
)
v
alu
es.
Ne
v
er
th
el
ess
,
th
e
ai
m
o
f
X
-
m
ea
n
s
al
g
o
r
ith
m
is
to
d
eliv
er
a
f
as
t a
n
d
ef
f
ec
ti
v
e
w
a
y
to
cl
u
s
ter
u
n
s
tr
u
ctu
r
ed
d
ata.
T
h
e
X
-
Me
a
n
s
cl
u
s
ter
i
n
g
al
g
o
r
ith
m
allo
w
s
to
s
p
ec
if
y
th
e
r
a
n
g
e
n
u
m
b
er
o
f
cl
u
s
ter
s
.
I
t
s
ta
r
ts
to
r
u
n
w
it
h
th
e
m
in
i
m
u
m
n
u
m
b
er
o
f
clu
s
ter
s
(
K)
o
f
t
h
e
g
i
v
en
r
an
g
e
an
d
a
s
s
i
g
n
s
a
B
I
C
s
co
r
e
a
n
d
it
co
n
tin
u
e
s
to
ad
d
ce
n
tr
o
id
s
w
h
er
e
t
h
e
y
ar
e
n
ee
d
ed
u
n
til
t
h
e
m
ax
i
m
u
m
n
u
m
b
er
o
f
clu
s
ter
s
(
K
m
a
x
)
r
an
g
e
is
r
ea
ch
ed
.
T
h
r
o
u
g
h
o
u
t
th
e
m
en
tio
n
ed
p
r
o
ce
s
s
,
t
h
e
ce
n
tr
o
id
s
et
w
il
l
b
e
r
ec
o
r
d
ed
w
h
en
it
i
s
ac
h
ie
v
in
g
t
h
e
b
est
s
co
r
e,
an
d
th
i
s
b
ec
o
m
e
s
th
e
f
i
n
al
ce
n
tr
o
id
s
et
o
u
tp
u
t.
E
ac
h
m
o
d
e
l
h
a
s
(
K)
to
(
K
m
a
x
)
+1
n
u
m
b
er
s
o
f
ce
n
tr
o
id
s
a
n
d
a
B
I
C
s
co
r
e
w
h
ich
w
il
l
d
ec
id
es
t
h
e
f
it
n
u
m
b
er
o
f
cl
u
s
ter
s
.
T
h
u
s
,
it
i
s
a
u
to
m
at
icall
y
s
elec
ti
n
g
t
h
e
o
p
ti
m
a
l
r
esu
lta
n
t
o
f
cl
u
s
ter
s
n
u
m
b
er
.
T
h
e
s
tep
s
f
lo
w
o
f
t
h
e
X
-
Me
a
n
s
cl
u
s
ter
i
n
g
alg
o
r
it
h
m
is
g
i
v
en
i
n
t
h
e
Fi
g
u
r
e
1
,
as d
ef
i
n
ed
in
[
2
4
]
.
T
h
e
X
-
m
ea
n
s
C
lu
s
ter
i
n
g
f
u
n
c
tio
n
s
is
a
s
s
o
ciate
d
w
it
h
s
tr
e
n
g
th
s
f
o
r
f
u
r
t
h
er
i
m
p
r
o
v
e
t
h
e
k
n
o
w
led
g
e
d
is
co
v
er
y
w
it
h
in
m
u
lti
-
v
ie
w
tex
tu
al
d
ataset
i
n
ter
m
s
o
f
t
h
e
p
r
o
p
o
s
ed
f
r
a
m
e
w
o
r
k
i
m
p
le
m
e
n
tat
io
n
.
T
h
e
clu
s
ter
i
n
g
ta
s
k
p
r
o
ce
s
s
is
r
ep
ea
ted
w
it
h
u
p
d
atin
g
t
h
e
b
e
s
t
B
I
C
s
co
r
e
till
th
e
m
a
x
i
m
u
m
n
u
m
b
er
o
f
cl
u
s
ter
s
(
K
m
ax
)
r
a
n
g
e
is
r
ea
c
h
e
d
w
h
i
ch
is
p
r
o
v
id
es
a
t
y
p
ical
n
u
m
b
er
o
f
clu
s
ter
s
.
A
ls
o
,
t
h
i
s
w
il
l
af
f
ec
ts
p
o
s
itiv
e
l
y
th
e
r
eq
u
ir
ed
p
r
o
ce
s
s
in
g
ti
m
e.
Fu
r
t
h
er
m
o
r
e,
th
e
u
s
e
o
f
co
n
cu
r
r
en
c
y
s
p
ee
d
s
u
p
t
h
e
cl
u
s
ter
in
g
p
r
o
ce
s
s
an
d
th
e
u
s
e
o
f
t
h
e
B
I
C
g
i
v
es a
m
at
h
e
m
a
ticall
y
co
m
p
lete
m
ea
s
u
r
e
o
f
q
u
al
it
y
.
3
.
1
.
2
.
H
o
t
Sp
o
t
a
lg
o
rit
hm
o
f
a
s
s
o
cia
t
io
n r
ules
Ho
tSp
o
t
alg
o
r
ith
m
o
f
ass
o
ciati
o
n
r
u
les
f
in
d
s
o
u
t
a
s
et
o
f
r
u
le
s
p
r
esen
ted
in
s
tr
u
ct
u
r
e
s
i
m
ila
r
to
a
tr
ee
,
w
h
ic
h
i
s
m
ax
i
m
ize
o
r
m
i
n
i
m
i
ze
a
tar
g
et
v
al
u
e
o
f
i
n
ter
est
[
2
5
,
26]
.
T
h
e
Ho
tSp
o
t
a
lg
o
r
ith
m
w
it
h
a
n
o
m
in
a
l
tar
g
et
lo
o
k
s
f
o
r
s
eg
m
e
n
ts
o
f
t
h
e
in
f
o
r
m
at
io
n
w
h
er
e
th
er
e
is
a
h
ig
h
ch
a
n
ce
o
f
a
m
in
o
r
it
y
v
alu
e
o
cc
u
r
r
in
g
b
y
g
iv
e
n
th
e
co
n
s
tr
ain
t
o
f
a
m
i
n
i
m
u
m
s
u
p
p
o
r
t,
an
d
w
it
h
a
n
u
m
er
ic
tar
g
et
co
u
ld
b
e
attr
ac
te
d
in
d
i
s
co
v
er
y
o
f
s
eg
m
e
n
ts
w
h
er
e
th
e
y
a
r
e
h
i
g
h
er
th
a
n
a
v
er
ag
e
in
th
e
w
h
o
le
d
ataset.
Fo
r
ex
a
m
p
le,
i
n
th
e
s
en
ti
m
en
t
m
u
lti
-
v
ie
w
tex
t
u
al
d
ata
s
ce
n
ar
io
,
it
atte
m
p
ts
to
f
in
d
i
n
w
h
ic
h
p
o
s
it
iv
e
o
r
n
eg
a
tiv
e
tex
tu
al
r
ev
ie
w
g
r
o
u
p
s
ar
e
at
t
h
e
h
ig
h
e
s
t
e
m
o
tio
n
al
clai
m
r
atio
,
o
r
i
n
w
h
ic
h
p
o
s
iti
v
e
o
r
n
e
g
ati
v
e
t
ex
tu
a
l
r
e
v
ie
w
g
r
o
u
p
s
h
a
v
e
t
h
e
h
i
g
h
e
s
t
a
v
er
ag
e
s
eq
u
en
ce
o
f
s
e
n
s
i
tiv
e
is
s
u
e
s
in
s
en
ti
m
e
n
tal
r
ev
ie
w
s
d
atas
et.
T
h
is
al
g
o
r
ith
m
is
clo
s
e
t
o
th
e
P
R
I
M
b
u
m
p
h
u
n
ti
n
g
a
lg
o
r
it
h
m
d
escr
ib
ed
b
y
Frie
d
m
a
n
an
d
Fi
s
h
er
,
[
2
7
]
.
T
h
e
Sin
g
le
-
k
e
y
I
n
f
o
r
m
a
tio
n
s
eq
u
en
ce
s
w
er
e
r
ec
e
n
tl
y
li
n
k
e
d
w
ith
id
e
n
ti
f
ied
clu
s
ter
s
u
s
in
g
X
-
m
ea
n
s
clu
s
ter
i
n
g
alg
o
r
it
h
m
,
w
h
ich
is
u
s
ed
in
t
h
i
s
r
esear
ch
f
o
r
s
ep
ar
atio
n
to
a
s
e
n
ti
m
en
ta
l
r
ev
ie
w
ca
te
g
o
r
ies.
T
h
e
g
o
al
h
er
e
is
to
id
en
tify
ch
ar
ac
ter
is
tic
s
o
f
ea
ch
tex
t
u
a
l
d
o
cu
m
en
t
as
p
o
s
iti
v
e
o
r
n
eg
ati
v
e
u
s
er
-
r
ev
ie
w
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8708
S
en
timen
ta
l c
la
s
s
ifica
tio
n
a
n
a
lysi
s
o
f p
o
la
r
ity
mu
lti
-
view tex
tu
a
l d
a
ta
u
s
in
g
d
a
ta
.
..
(
A
li Ha
mee
d
Ya
s
s
ir
)
5529
ca
teg
o
r
ies,
w
h
er
e
p
er
f
o
r
m
i
n
g
Ho
tSp
o
t
ass
o
ciatio
n
r
u
le
s
l
ater
f
o
r
m
u
l
ti
-
v
ie
w
te
x
t
u
al
d
ata
is
s
i
g
n
i
f
ica
n
tl
y
h
elp
f
u
l
i
n
ter
m
s
o
f
m
ax
i
m
i
zin
g
th
e
tar
g
et
v
al
u
e
o
f
i
n
t
er
est
in
ea
c
h
cl
u
s
ter
to
o
b
t
ain
t
h
e
m
ax
i
m
u
m
co
n
ce
n
tr
atio
n
i
m
p
r
o
v
e
m
e
n
t
b
y
r
ec
o
g
n
izin
g
t
h
e
f
r
eq
u
e
n
t
i
n
d
iv
id
u
al
ter
m
s
in
ea
c
h
cl
u
s
ter
an
d
co
m
b
in
e
s
t
h
e
m
to
a
lar
g
er
ter
m
s
et
s
as
lo
n
g
as
th
o
s
e
ter
m
s
ets
ap
p
ea
r
ap
p
r
o
p
r
iately
f
r
eq
u
en
ted
i
n
th
e
d
atab
ase.
T
h
e
w
h
o
le
alg
o
r
ith
m
ca
n
b
e
d
iv
id
ed
in
to
f
o
u
r
s
tep
s
a
s
th
e
F
ig
u
r
e
2
.
T
h
e
au
to
m
atic
d
is
co
v
er
y
o
f
f
r
eq
u
en
t
ter
m
s
ets
o
r
m
u
lti
-
k
e
y
f
r
eq
u
e
n
t
i
n
f
o
r
m
atio
n
r
u
le
s
(
MK
FIR)
is
a
v
er
y
p
r
o
m
is
i
n
g
ar
ea
o
f
d
ata
m
i
n
in
g
ap
p
licatio
n
s
in
t
h
e
f
iel
d
o
f
m
u
l
ti
-
v
ie
w
te
x
t
u
al
d
ata.
A
f
r
eq
u
e
n
t
ter
m
s
e
t
is
d
escr
ib
ed
as
b
ein
g
a
co
llec
tio
n
o
f
k
e
y
p
h
r
ase
s
w
h
ich
u
s
u
all
y
ta
k
es
p
lace
i
n
s
id
e
t
h
e
te
x
t
d
o
cu
m
en
t
s
et
ın
ex
ce
s
s
o
f
th
e
m
i
n
i
m
u
m
s
u
p
p
o
r
t ti
m
e
s
.
Fig
u
r
e
1
.
Flo
w
o
f
i
n
f
o
r
m
atio
n
in
x
-
m
ea
n
s
c
l
u
s
ter
i
n
g
a
lg
o
r
ith
m
Fig
u
r
e
2
.
Flo
w
d
iag
r
a
m
f
o
r
h
ot
s
p
o
t
a
lg
o
r
ith
m
f
o
r
MK
FIR
f
o
r
m
atio
n
3
.
1
.
3
.
Na
iv
e
ba
y
es
a
nd
decisi
o
n t
re
es
(
J
4
8
)
a
lg
o
rit
h
m
s
f
o
r
cla
s
s
if
ica
t
io
n
I
n
th
i
s
s
ta
g
e,
t
w
o
d
if
f
er
en
t c
la
s
s
i
f
ier
s
,
i
n
p
ar
ticu
lar
d
ec
is
io
n
tr
ee
s
(
J
4
8
)
an
d
n
aiv
e
b
a
y
es
cla
s
s
i
f
icatio
n
alg
o
r
ith
m
s
ar
e
co
m
p
ar
ed
an
d
ap
p
lied
in
th
is
r
e
s
ea
r
ch
.
T
h
e
p
u
r
p
o
s
e
o
f
a
n
al
y
z
in
g
ea
c
h
o
f
t
h
ese
d
i
s
ti
n
ct
class
i
f
icatio
n
al
g
o
r
ith
m
s
lies
i
n
t
h
e
f
ac
t
t
h
a
t
ea
ch
u
s
es
d
i
f
f
e
r
en
t
s
elec
tio
n
cr
iter
ia
s
to
t
h
e
cr
itical
in
f
o
r
m
a
tio
n
v
ar
iab
les.
T
h
e
J
4
8
class
if
icati
o
n
alg
o
r
ith
m
is
b
ased
o
n
d
ec
i
s
io
n
tr
ee
.
Naïv
e
b
ay
e
s
class
if
i
ca
tio
n
alg
o
r
ith
m
i
s
b
ased
o
n
p
r
o
b
ab
ilit
y
[
2
8
]
.
Su
cc
ess
i
v
el
y
,
th
e
y
e
x
te
n
d
r
ig
h
t
f
r
o
m
th
e
s
i
m
p
le
d
is
ta
n
ce
m
ea
s
u
r
e
to
o
th
er
p
at
h
o
f
p
r
o
b
a
b
ilis
tic
d
is
ta
n
ce
m
ea
s
u
r
es
to
s
ea
r
ch
th
e
s
i
m
ilar
it
y
cr
it
er
ia
a
m
o
n
g
a
s
et
o
f
d
o
cu
m
e
n
t
s
an
d
t
h
e
n
clas
s
i
f
y
t
h
em
d
i
r
e
c
tly
b
a
s
e
d
o
n
th
e
i
r
p
a
r
t
i
cu
l
a
r
c
a
t
eg
o
r
y
.
I
n
th
is
m
an
n
e
r
,
t
h
e
m
a
in
o
b
je
c
t
iv
e
o
f
s
u
ch
ex
p
e
r
im
en
t
s
i
s
to
v
e
r
if
y
th
e
p
r
o
p
o
s
e
d
f
r
am
e
w
o
r
k
o
n
s
u
ch
b
as
is
a
s
m
ax
im
i
z
es
th
e
t
a
r
g
e
t
v
a
lu
e
o
f
in
te
r
e
s
t
b
y
d
is
c
o
v
e
r
in
g
m
u
l
ti
-
k
e
y
f
r
eq
u
en
t
in
f
o
r
m
a
tio
n
r
u
les
,
wh
ich
d
eli
v
er
s
co
n
s
id
er
ab
l
y
b
et
ter
ac
cu
r
ac
y
w
h
en
co
m
p
ar
ed
to
th
e
s
i
m
p
le
ter
m
-
b
ased
class
i
f
icatio
n
tech
n
iq
u
es.
T
h
e
Dec
is
io
n
T
r
ee
s
an
d
Naiv
e
B
a
y
e
s
clas
s
i
f
icatio
n
al
g
o
r
ith
m
s
h
as
b
ee
n
ex
a
m
in
ed
w
it
h
r
eg
ar
d
s
to
clas
s
if
y
in
g
te
x
t
u
al
d
ata
in
ac
co
r
d
an
ce
w
it
h
t
w
o
d
if
f
er
en
t c
lass
e
s
.
4.
T
H
E
P
RO
P
O
SE
D
F
RAM
E
WO
RK
I
n
th
is
s
ec
tio
n
th
e
m
aj
o
r
f
u
n
ct
io
n
s
o
f
k
n
o
w
led
g
e
d
is
co
v
er
y
d
ef
in
ed
ab
o
v
e
ar
e
in
teg
r
ated
to
d
ev
elo
p
th
e
p
r
o
p
o
s
ed
f
r
a
m
e
w
o
r
k
f
o
r
class
i
f
y
i
n
g
t
h
e
s
e
n
ti
m
en
ta
l
m
u
lti
-
v
ie
w
te
x
tu
a
l
d
ata.
T
h
is
ca
n
b
e
d
o
n
e
th
r
o
u
g
h
m
a
x
i
m
izi
n
g
t
h
e
tar
g
et
v
al
u
e
o
f
in
ter
est
b
y
d
is
co
v
er
in
g
m
u
lti
-
k
e
y
f
r
eq
u
e
n
t
in
f
o
r
m
atio
n
r
u
les
.
T
h
e
in
itial
s
ta
g
es
o
f
k
n
o
w
le
d
g
e
d
is
co
v
er
y
ar
e
r
ea
lized
b
y
ap
p
l
y
in
g
t
h
e
X
-
m
ea
n
s
cl
u
s
ter
in
g
al
g
o
r
ith
m
t
o
id
en
ti
f
y
th
e
Si
n
g
le
-
k
e
y
I
n
f
o
r
m
atio
n
s
eq
u
e
n
ce
s
.
Af
ter
w
ar
d
,
th
e
id
e
n
ti
f
ied
Si
n
g
le
-
k
e
y
I
n
f
o
r
m
at
io
n
s
eq
u
en
ce
s
u
tili
ze
d
b
y
t
h
e
Ho
tSp
o
t
al
g
o
r
ith
m
o
f
ass
o
ciat
io
n
r
u
les
to
g
e
n
er
ate
m
u
lti
-
k
e
y
f
r
eq
u
e
n
t
in
f
o
r
m
atio
n
r
u
les
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
10
,
No
.
5
,
Octo
b
e
r
2
0
2
0
:
5
5
2
6
-
5534
5530
T
h
e
in
teg
r
ated
ap
p
licatio
n
o
f
t
h
ese
d
ata
m
i
n
i
n
g
al
g
o
r
ith
m
s
w
h
ic
h
g
e
n
er
ate
t
h
e
m
u
lti
-
k
e
y
f
r
eq
u
en
t i
n
f
o
r
m
atio
n
r
u
les
ca
n
b
e
u
tili
ze
d
to
s
h
ap
e
a
k
n
o
w
led
g
e
-
b
ased
r
elatio
n
a
l
d
atab
ase
f
o
r
i
m
p
r
o
v
ed
an
al
y
z
i
n
g
a
n
d
h
a
n
d
lin
g
o
f
th
e
s
e
n
ti
m
e
n
t
m
u
lti
-
v
ie
w
te
x
t
u
al
d
ata.
T
h
e
b
en
ef
it
w
it
h
d
is
c
o
v
er
in
g
th
e
k
e
y
i
n
f
o
r
m
atio
n
in
ter
m
s
o
f
m
u
lti
-
k
e
y
f
r
eq
u
en
t
i
n
f
o
r
m
atio
n
r
u
les
s
er
v
es
to
a
u
to
m
ate
th
e
p
r
o
ce
s
s
o
f
ca
te
g
o
r
izin
g
t
h
e
s
e
n
ti
m
en
tal
m
u
lti
-
v
ie
w
te
x
tu
a
l
d
o
cu
m
en
ts
w
it
h
i
n
t
w
o
p
r
ed
ef
i
n
ed
ca
teg
o
r
ies (
i.e
.
p
o
s
itiv
e
o
r
n
eg
ati
v
e
i
n
f
o
r
m
atio
n
cla
s
s
e
s
)
.
Fu
r
t
h
er
m
o
r
e,
t
h
e
p
r
o
p
o
s
ed
f
r
a
m
e
w
o
r
k
is
o
u
tli
n
ed
to
a
n
al
y
ze
s
e
n
ti
m
en
t
m
u
lti
-
v
ie
w
te
x
t
u
al
d
ataset,
w
h
ic
h
lies
to
f
o
u
r
m
ai
n
d
ata
p
r
o
ce
s
s
in
g
s
ta
g
e
s
,
as s
h
o
w
n
i
n
F
ig
u
r
e
3
.
-
Data
p
r
ep
ar
atio
n
an
d
p
r
ep
r
o
ce
s
s
i
n
g
s
tag
e.
-
Data
clu
s
ter
in
g
s
ta
g
e.
-
Ma
x
i
m
izi
n
g
tar
g
et
v
a
lu
e
o
f
i
n
t
er
est s
tag
e.
-
C
las
s
i
f
icatio
n
s
tag
e.
T
h
e
k
n
o
w
led
g
e
d
is
co
v
er
y
a
n
d
tex
t
class
if
ica
tio
n
f
r
o
m
t
h
ese
f
o
u
r
m
ai
n
p
ar
ts
is
in
f
l
u
en
ce
d
b
y
th
e
d
is
co
v
er
y
o
f
m
u
lti
-
k
e
y
f
r
eq
u
en
t
i
n
f
o
r
m
atio
n
r
u
le
s
.
W
h
en
th
e
in
f
o
r
m
at
io
n
i
s
alr
ea
d
y
h
as
b
ee
n
s
tr
u
c
tu
r
ed
b
y
th
e
d
ata
p
r
ep
ar
atio
n
an
d
p
r
ep
r
o
ce
s
s
in
g
s
tag
e,
it
is
th
e
n
p
ass
ed
to
th
e
d
ata
cl
u
s
ter
i
n
g
s
tag
e
f
o
r
ar
r
an
g
i
n
g
tex
t
u
al
d
ata
in
to
a
n
u
m
b
er
o
f
clu
s
ter
s
an
d
id
en
ti
f
y
th
e
Sin
g
le
-
k
e
y
I
n
f
o
r
m
atio
n
s
eq
u
en
ce
s
.
I
d
en
tify
i
n
g
th
e
s
eq
u
e
n
ce
o
f
k
e
y
ter
m
s
h
e
lp
s
en
s
u
r
i
n
g
s
o
m
e
k
e
y
in
f
o
r
m
atio
n
r
u
le
s
i
n
t
h
e
m
u
lti
-
v
ie
w
tex
t
u
al
d
o
cu
m
e
n
ts
.
I
n
th
e
co
n
te
x
t
o
f
cu
r
r
en
t
r
es
ea
r
ch
,
ea
ch
d
o
cu
m
e
n
t
av
aila
b
le
in
th
e
s
en
ti
m
en
ta
l
m
u
lti
-
v
ie
w
tex
tu
al
d
atase
t
co
m
p
r
is
e
s
a
co
m
b
in
a
tio
n
o
f
t
er
m
s
to
e
x
p
licate
i
m
p
r
es
s
io
n
.
Ma
x
i
m
izi
n
g
tar
g
et
v
alu
e
o
f
i
n
ter
est
b
y
u
tili
z
in
g
th
e
m
u
lt
i
-
k
e
y
f
r
eq
u
e
n
t
i
n
f
o
r
m
atio
n
r
u
les
e
n
ab
les
s
u
m
m
ar
izin
g
th
e
k
e
y
i
n
f
o
r
m
atio
n
w
i
th
i
n
ea
ch
p
ar
ticu
lar
d
o
cu
m
en
t f
o
r
f
u
r
th
er
p
r
o
ce
s
s
i
n
g
o
r
class
if
ica
tio
n
.
Fig
u
r
e
3
.
Mu
lti
-
v
ie
w
t
ex
tu
al
c
lass
i
f
icatio
n
s
y
s
te
m
T
h
e
illu
s
tr
ated
f
r
a
m
e
w
o
r
k
ab
o
v
e
w
it
h
i
n
te
g
r
atio
n
o
f
d
if
f
er
en
t
e
f
f
o
r
ts
o
f
d
ata
m
i
n
i
n
g
alg
o
r
ith
m
s
i
s
p
r
o
p
o
s
ed
to
an
al
y
ze
f
r
ee
f
o
r
m
atted
f
o
r
m
o
f
m
u
lti
-
v
ie
w
te
x
tu
a
l
d
ata.
T
h
e
p
r
o
p
o
s
ed
f
r
am
e
w
o
r
k
i
s
ac
h
ie
v
ed
th
r
o
u
g
h
th
e
ap
p
licatio
n
o
f
X
-
m
ea
n
s
al
g
o
r
ith
m
f
o
r
clu
s
t
er
in
g
an
d
Ho
t
Sp
o
t
alg
o
r
ith
m
o
f
a
s
s
o
ciatio
n
r
u
les
to
d
is
co
v
er
u
s
e
f
u
l k
n
o
w
led
g
e
i
n
t
er
m
s
o
f
m
u
lti
-
k
e
y
f
r
eq
u
en
t in
f
o
r
m
at
io
n
r
u
le
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8708
S
en
timen
ta
l c
la
s
s
ifica
tio
n
a
n
a
lysi
s
o
f p
o
la
r
ity
mu
lti
-
view tex
tu
a
l d
a
ta
u
s
in
g
d
a
ta
.
..
(
A
li Ha
mee
d
Ya
s
s
ir
)
5531
5.
DATA AN
AL
Y
SI
S
T
h
e
an
al
y
s
i
s
o
f
clas
s
if
y
i
n
g
th
e
s
en
ti
m
en
ta
l
m
u
lti
-
v
ie
w
te
x
tu
a
l
d
ata
in
to
t
w
o
ca
te
g
o
r
ies
th
r
o
u
g
h
th
e
ap
p
licatio
n
o
f
t
h
e
p
r
o
p
o
s
ed
f
r
a
m
e
w
o
r
k
i
s
d
is
cu
s
s
ed
in
t
h
e
s
ce
n
ar
io
o
f
o
n
li
n
e
p
o
lar
it
y
u
s
er
-
r
ev
ie
w
s
d
ataset.
4
.
1
.
Da
t
a
co
llect
io
n
I
n
o
r
d
er
to
p
er
f
o
r
m
t
h
e
e
x
p
er
i
m
e
n
tatio
n
o
f
th
is
r
e
s
ea
r
ch
s
t
u
d
y
,
t
h
e
p
er
f
o
r
m
a
n
ce
s
o
f
p
ar
t
icu
lar
te
s
t
an
d
an
al
y
s
is
w
it
h
th
e
p
r
o
p
o
s
ed
f
r
a
m
e
w
o
r
k
is
co
n
d
u
ct
ed
u
s
in
g
s
en
ti
m
en
t
p
o
lar
it
y
r
ev
ie
w
s
d
ataset.
T
h
e
s
en
ti
m
en
t
p
o
lar
it
y
r
ev
ie
w
s
d
ata
p
r
o
v
id
ed
b
y
[
2
9
]
is
o
n
e
o
f
th
e
co
m
m
o
n
l
y
u
s
ed
s
en
ti
m
e
n
t
a
n
al
y
s
i
s
d
atasets
.
T
h
e
s
en
t
i
m
e
n
t
p
o
lar
i
t
y
r
ev
ie
w
s
d
ataset
in
v
o
lv
e
s
2
0
0
0
o
n
lin
e
u
s
er
-
cr
ea
ted
r
ev
ie
w
s
o
n
a
g
iv
e
n
to
p
ic
ar
ch
iv
ed
f
r
o
m
t
h
e
cs.c
o
r
n
ell
w
eb
p
o
r
tal
an
d
it
is
av
ailab
le
in
(
h
ttp
://
www
.
c
s
.
co
r
n
ell.
ed
u
/p
eo
p
le/p
ab
o
/
m
o
v
ie
-
re
v
ie
w
-
d
ata)
.
4
.
2
.
F
un
ct
io
na
l
re
qu
ire
m
e
nts
W
ek
a
3
.
7
.
1
2
s
o
f
t
w
ar
e
is
u
s
ed
to
s
u
p
p
o
r
t
th
e
ac
tiv
ities
d
ef
in
ed
d
u
r
in
g
th
e
e
x
p
er
i
m
e
n
tatio
n
o
f
th
e
p
r
o
p
o
s
ed
f
r
a
m
e
w
o
r
k
,
s
t
ar
tin
g
w
it
h
t
h
e
p
r
e
-
p
r
o
ce
s
s
i
n
g
s
ta
g
e
to
w
ar
d
t
h
e
ap
p
lica
tio
n
o
f
k
n
o
w
led
g
e
d
is
co
v
er
y
le
v
el,
w
h
ic
h
i
s
u
s
ed
to
class
i
f
y
t
h
e
s
e
n
ti
m
e
n
t p
o
lar
it
y
r
ev
ie
w
s
d
ataset.
4
.
3
.
I
m
ple
m
ent
a
t
io
n o
f
f
ra
m
ew
o
rk
f
u
nct
io
na
litie
s
T
h
e
v
alid
atio
n
o
f
th
e
p
r
o
p
o
s
ed
f
r
a
m
e
w
o
r
k
is
p
er
f
o
r
m
ed
i
n
th
i
s
s
ec
tio
n
i
n
o
r
d
er
to
en
s
u
r
e
b
etter
p
er
f
o
r
m
a
n
ce
t
h
a
n
t
h
e
s
i
m
p
le
ter
m
-
b
ased
clas
s
if
icatio
n
m
o
d
els,
th
r
o
u
g
h
id
e
n
ti
f
y
in
g
k
e
y
k
n
o
w
led
g
e
ar
ea
s
w
it
h
i
n
s
e
n
ti
m
e
n
t p
o
lar
it
y
r
ev
i
e
w
s
d
ataset.
a.
P
r
ep
ar
atio
n
an
d
P
r
e
-
p
r
o
ce
s
s
in
g
o
f
T
ex
t D
o
cu
m
en
t
s
T
h
e
d
ata
p
r
ep
ar
atio
n
an
d
p
r
e
-
p
r
o
ce
s
s
i
n
g
ta
s
k
s
ar
e
b
ein
g
u
s
ed
to
r
em
o
v
e
u
n
-
n
ec
es
s
ar
y
t
er
m
s
f
r
o
m
u
n
s
tr
u
ct
u
r
ed
te
x
t
d
ata
f
o
r
m
a
ts
a
n
d
p
r
ep
ar
e
th
e
tex
t
u
al
d
ata
f
o
r
f
u
r
th
er
an
al
y
s
is
o
f
v
ar
io
u
s
d
ata
m
i
n
i
n
g
alg
o
r
ith
m
s
f
o
r
s
u
b
s
eq
u
en
t
a
n
al
y
s
i
s
o
f
t
h
e
te
x
t.
T
h
e
d
ata
p
r
ep
ar
atio
n
an
d
p
r
e
-
p
r
o
ce
s
s
i
n
g
task
s
ar
e
i
n
cl
u
d
in
g
th
e
f
o
llo
w
i
n
g
s
tep
s
:
-
R
e
m
o
v
e
t
h
e
s
to
p
w
o
r
d
s
o
r
t
er
m
s
th
at
ca
n
b
e
les
s
e
f
f
ec
ti
v
e
i
n
k
n
o
w
led
g
e
d
is
co
v
er
y
f
r
o
m
s
e
n
ti
m
en
tal
m
u
lti
-
v
ie
w
tex
tu
al
d
ata.
-
P
er
f
o
r
m
w
o
r
d
s
te
m
m
i
n
g
p
r
o
ce
s
s
b
y
w
a
y
o
f
r
ed
u
ci
n
g
d
er
iv
e
d
w
o
r
d
s
o
r
ter
m
s
to
its
ac
tu
al
s
te
m
o
r
th
e
r
o
o
t
f
o
r
m
.
-
Stru
ct
u
r
e
th
e
i
n
f
o
r
m
atio
n
i
n
to
v
ec
to
r
f
o
r
m
i
n
w
h
ich
e
v
er
y
ter
m
is
co
n
s
id
er
ed
a
w
o
r
d
v
ec
to
r
.
-
R
ep
r
esen
t
te
x
t
in
a
f
o
r
m
o
f
t
er
m
f
r
eq
u
en
c
y
(
T
F)
m
atr
ix
b
y
co
u
n
ti
n
g
th
e
w
o
r
d
s
a
n
d
t
h
eir
co
r
r
esp
o
n
d
in
g
f
r
e
q
u
en
c
ies,
as s
h
o
w
n
in
F
ig
u
r
e
4
(
a
)
.
-
Dete
r
m
i
n
e
th
e
s
i
m
ilar
itie
s
b
y
ca
lcu
lati
n
g
th
e
E
u
clid
ea
n
Di
s
t
an
ce
m
et
h
o
d
.
-
W
eig
h
t
ter
m
s
i
n
t
h
e
s
et
o
f
d
o
cu
m
e
n
ts
t
h
r
o
u
g
h
I
n
v
er
s
e
d
o
cu
m
e
n
t
f
r
eq
u
e
n
c
y
(
I
DF)
a
n
d
ter
m
f
r
eq
u
e
n
c
y
(
T
F)
b
y
u
s
i
n
g
i
ts
m
atr
ices (
T
F*
I
DF)
.
-
T
h
e
o
u
tp
u
t
is
th
e
n
s
av
ed
i
n
at
tr
ib
u
te
r
elatio
n
f
ile
f
o
r
m
at
(
.
ar
f
f
)
f
ile
th
a
t
w
ill
b
e
u
s
ed
later
f
o
r
f
u
r
th
er
d
at
a
m
i
n
in
g
a
n
al
y
tics
.
b.
A
p
p
licatio
n
o
f
X
-
Me
an
s
clu
s
t
er
in
g
al
g
o
r
ith
m
T
h
e
X
-
m
ea
n
s
al
g
o
r
ith
m
is
ap
p
lied
o
n
t
h
e
(
.
ar
f
f
)
f
ile
f
o
r
m
a
t
w
h
ich
w
as
o
b
tain
ed
as
a
r
e
s
u
lt
f
r
o
m
th
e
p
r
ep
ar
atio
n
an
d
p
r
e
-
p
r
o
c
ess
i
n
g
s
tag
e.
T
h
is
alg
o
r
it
h
m
is
v
er
y
s
u
p
p
o
r
tiv
e
to
ca
lcu
late
au
to
m
at
icall
y
th
e
ap
p
r
o
p
r
iate
n
u
m
b
er
o
f
g
en
er
ated
clu
s
ter
s
th
at
ca
p
t
u
r
i
n
g
r
elate
d
in
f
o
r
m
a
tio
n
b
y
u
s
in
g
t
h
e
Si
n
g
le
-
k
e
y
I
n
f
o
r
m
a
tio
n
s
eq
u
e
n
ce
s
.
T
h
e
i
m
p
le
m
en
ta
tio
n
ex
p
e
r
i
m
en
t
o
f
th
i
s
r
esear
ch
h
as
b
ee
n
s
h
o
w
n
th
at
X
-
m
ea
n
s
alg
o
r
ith
m
ad
o
p
ts
th
e
to
tal
cl
u
s
ter
s
f
o
r
th
e
u
s
ed
s
en
ti
m
e
n
t
m
u
l
ti
-
v
ie
w
te
x
t
u
al
d
ataset
to
b
e
tw
o
cl
u
s
ter
s
a
s
sh
o
w
n
in
t
h
e
F
ig
u
r
e
4
(
b
)
.
I
t
g
av
e
s
u
b
s
tan
ce
to
b
e
v
er
y
ad
v
an
ta
g
eo
u
s
clu
s
ter
in
g
al
g
o
r
ith
m
in
p
r
es
er
v
in
g
v
alu
ab
le
i
n
f
o
r
m
atio
n
s
tr
u
ct
u
r
es
t
h
at
id
en
ti
f
ied
i
n
th
e
s
e
n
ti
m
e
n
t
m
u
lti
-
v
ie
w
tex
t
u
al
d
ataset.
T
h
u
s
,
t
h
e
X
-
m
ea
n
s
alg
o
r
ith
m
h
as
b
ee
n
u
tili
ze
d
as
an
e
f
f
icie
n
t
to
o
l
to
ca
p
tu
r
e
v
al
u
ab
le
k
e
y
I
n
f
o
r
m
at
io
n
w
it
h
i
n
e
v
er
y
s
in
g
le
clu
s
ter
,
as
s
h
o
w
n
in
t
h
e
Fi
g
u
r
e
4
(
c
)
.
c.
A
p
p
licatio
n
o
f
Ho
tSp
o
t a
s
s
o
ci
atio
n
r
u
les al
g
o
r
it
h
m
T
h
e
d
is
co
v
er
y
o
f
m
ax
i
m
u
m
t
ar
g
et
v
al
u
e
o
f
i
n
ter
est
b
y
d
eter
m
in
in
g
m
u
l
ti
-
k
e
y
f
r
eq
u
en
t
i
n
f
o
r
m
atio
n
r
u
les
i
n
m
u
lti
-
v
ie
w
tex
t
u
al
d
atasets
h
elp
s
w
it
h
f
i
n
d
in
g
ter
m
s
th
a
t
ap
p
ea
r
m
o
s
t
f
r
eq
u
e
n
t
l
y
w
i
th
ea
ch
o
t
h
er
.
Kn
o
w
led
g
e
d
is
co
v
er
ed
w
i
th
t
h
e
h
elp
o
f
r
u
les
f
o
r
m
a
x
i
m
u
m
tar
g
et
v
al
u
e
o
f
i
n
ter
es
t
is
e
m
p
lo
y
ed
f
o
r
m
ap
p
in
g
th
e
p
o
s
iti
v
e
o
r
n
eg
ati
v
e
i
n
f
o
r
m
atio
n
d
o
cu
m
e
n
t c
ate
g
o
r
ies,
as sh
o
w
n
i
n
th
e
F
ig
u
r
e
4
(
d
)
.
d.
A
p
p
licatio
n
o
f
clas
s
if
icatio
n
a
lg
o
r
ith
m
s
T
h
e
o
u
tp
u
t o
f
(
.
ar
f
f
)
f
ile
b
a
s
ed
o
n
Sin
g
le
-
k
e
y
I
n
f
o
r
m
atio
n
s
e
q
u
en
ce
s
o
b
tai
n
ed
b
y
X
-
m
ea
n
s
clu
s
ter
i
n
g
an
d
m
ax
i
m
u
m
tar
g
et
v
al
u
e
o
f
in
ter
est
ac
h
iev
ed
b
y
Ho
t
Sp
o
t
ass
o
ciatio
n
r
u
le
s
,
w
h
ic
h
i
s
t
h
e
n
lo
ad
ed
i
n
to
W
ek
a
3
.
7
.
1
2
.
T
w
o
d
if
f
er
en
t
clas
s
i
f
ie
r
alg
o
r
ith
m
s
w
er
e
test
ed
to
class
i
f
y
t
h
e
s
e
n
ti
m
en
t
m
u
lti
-
v
ie
w
te
x
tu
a
l
d
ata
in
to
p
o
s
itiv
e
o
r
n
eg
ati
v
e
in
f
o
r
m
a
tio
n
d
o
cu
m
e
n
t
ca
teg
o
r
ies
an
d
th
e
r
esu
lt
ac
cu
r
ac
ies
t
h
at
ac
h
iev
ed
b
y
b
o
t
h
class
i
f
icatio
n
alg
o
r
it
h
m
s
ar
e
d
escr
ib
ed
an
d
d
is
cu
s
s
ed
i
n
n
e
x
t
(
Sectio
n
5
)
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
10
,
No
.
5
,
Octo
b
e
r
2
0
2
0
:
5
5
2
6
-
5534
5532
(
a)
(
b
)
(
c)
(
d
)
Fig
u
r
e
4
.
I
m
p
le
m
e
n
tatio
n
o
f
f
r
a
m
e
w
o
r
k
f
u
n
ctio
n
alitie
s
,
(
a)
m
atr
i
x
r
ep
r
esen
tatio
n
o
f
ter
m
s
an
d
th
eir
r
esp
ec
tiv
e
f
r
eq
u
en
c
y
,
(
b
)
n
u
m
b
er
o
f
clu
s
t
er
s
ca
p
tu
r
ed
b
y
x
-
m
ea
n
s
al
g
o
r
ith
m
ap
p
licatio
n
,
(
c)
k
e
y
i
n
f
o
r
m
atio
n
s
eq
u
e
n
ce
s
w
it
h
i
n
ea
ch
cl
u
s
ter
,
(
d
)
d
is
co
v
er
y
r
es
u
lt o
f
m
a
x
i
m
u
m
tar
g
et
v
alu
e
s
o
f
i
n
ter
est
6.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
O
F
CL
ASS
I
F
I
CA
T
I
O
N
ACCU
RAC
I
E
S
T
h
e
class
if
icat
io
n
r
es
u
lt
ac
c
u
r
ac
ies
ar
e
m
ea
s
u
r
ed
b
y
w
a
y
o
f
clas
s
if
y
i
n
g
s
en
ti
m
en
ta
l
m
u
lti
-
v
ie
w
tex
t
u
al
d
ata
a
s
p
o
s
iti
v
e
o
r
n
eg
ativ
e
i
n
f
o
r
m
atio
n
d
o
cu
m
e
n
t
c
ateg
o
r
ies.
T
h
e
cla
s
s
i
f
icatio
n
r
esu
lt
s
s
h
o
w
th
e
r
ate
o
f
co
r
r
ec
tly
cla
s
s
i
f
y
i
n
g
t
h
e
p
o
lar
ity
d
ata
w
h
e
n
b
o
th
o
f
s
i
m
p
le
-
ter
m
b
ased
clas
s
i
f
i
ca
tio
n
m
et
h
o
d
an
d
th
e
p
r
o
p
o
s
ed
m
eth
o
d
ar
e
s
ep
a
r
atel
y
u
s
ed
f
o
r
class
i
f
y
i
n
g
s
e
n
ti
m
e
n
t
m
u
lti
-
v
ie
w
te
x
t
u
al
d
ata
o
f
p
o
lar
it
y
r
e
v
ie
w
s
d
ataset
in
to
t
h
eir
r
esp
ec
tiv
e
ca
teg
o
r
ies.
T
h
e
ca
lcu
lated
co
r
r
ec
tl
y
clas
s
i
f
y
in
g
r
ates a
r
e
p
r
o
v
id
ed
in
T
ab
le
1
.
T
h
e
co
r
r
ec
tly
clas
s
i
f
icatio
n
r
ates
h
a
v
e
b
ee
n
m
ea
s
u
r
ed
alo
n
g
s
id
e
w
it
h
ev
er
y
s
i
n
g
le
cla
s
s
i
f
ier
f
o
r
p
r
ed
ictin
g
th
e
co
r
r
ec
t
ca
teg
o
r
y
o
f
ea
ch
p
o
lar
it
y
in
f
o
r
m
a
tio
n
d
o
cu
m
e
n
t.
C
o
n
s
eq
u
e
n
tl
y
,
th
e
h
i
g
h
est
r
ate
o
f
co
r
r
ec
tly
c
lass
if
ica
tio
n
h
as
t
h
e
ca
p
ac
i
t
y
w
it
h
m
o
r
e
ac
cu
r
ac
y
as
s
u
r
ed
i
n
t
h
e
p
r
o
ce
s
s
o
f
d
ec
is
io
n
m
ak
in
g
.
Dif
f
er
en
t
s
etti
n
g
s
o
f
p
ar
a
m
et
er
s
h
a
v
e
b
ee
n
te
s
ted
b
y
u
s
in
g
W
ek
a
3
.
7
.
1
2
s
o
f
t
w
ar
e,
i
n
w
h
ic
h
d
if
f
er
e
n
t
tes
t
m
o
d
e
s
lik
e
cr
o
s
s
-
v
a
lid
atio
n
w
i
th
1
0
f
o
ld
s
an
d
u
s
in
g
tr
ain
in
g
d
ataset
ar
e
test
ed
.
T
h
e
c
lass
if
i
er
s
in
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
u
s
i
n
g
r
u
le
s
f
o
r
m
ax
i
m
u
m
tar
g
et
v
al
u
e
o
f
i
n
ter
est
ar
e
all
r
esu
lt
i
m
p
r
o
v
ed
ac
cu
r
ac
y
th
a
n
t
h
e
s
i
m
p
l
e
ter
m
-
b
a
s
ed
class
i
f
icatio
n
m
e
th
o
d
,
w
h
ic
h
is
u
s
i
n
g
(
ter
m
f
r
eq
u
en
c
y
m
atr
i
x
)
o
n
l
y
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8708
S
en
timen
ta
l c
la
s
s
ifica
tio
n
a
n
a
lysi
s
o
f p
o
la
r
ity
mu
lti
-
view tex
tu
a
l d
a
ta
u
s
in
g
d
a
ta
.
..
(
A
li Ha
mee
d
Ya
s
s
ir
)
5533
T
ab
le
1
.
R
esu
lts
o
f
co
r
r
ec
tl
y
class
i
f
icatio
n
r
ates b
y
t
es
t
m
o
d
e
C
l
a
ssi
f
i
c
a
t
i
o
n
M
e
t
h
o
d
C
o
r
r
e
c
t
C
l
a
ssi
f
i
c
a
t
i
o
n
R
a
t
e
:
u
si
n
g
c
r
o
ss
-
v
a
l
i
d
a
t
i
o
n
w
i
t
h
1
0
f
o
l
d
s
C
o
r
r
e
c
t
C
l
a
ssi
f
i
c
a
t
i
o
n
R
a
t
e
:
u
si
n
g
t
r
a
i
n
i
n
g
d
a
t
a
se
t
T
h
e
p
r
o
p
o
se
d
c
l
a
ssi
f
i
c
a
t
i
o
n
f
r
a
me
w
o
r
k
S
i
mp
l
e
t
e
r
m
-
b
a
se
d
c
l
a
ssi
f
i
c
a
t
i
o
n
mo
d
e
l
T
h
e
p
r
o
p
o
se
d
c
l
a
ssi
f
i
c
a
t
i
o
n
f
r
a
me
w
o
r
k
S
i
mp
l
e
t
e
r
m
-
b
a
se
d
c
l
a
ssi
f
i
c
a
t
i
o
n
mo
d
e
l
D
e
c
i
si
o
n
T
r
e
e
s (J4
8
)
7
0
.
3
%
6
6
.
7
5
%
9
8
.
5
%
9
6
.
8
5
%
N
a
i
v
e
B
a
y
e
s
7
9
.
8
5
%
7
6
.
7
%
8
5
.
5
5
%
8
3
.
2
%
7.
CO
NCLU
SI
O
N
T
h
is
p
ap
er
h
as
b
ee
n
d
is
c
u
s
s
ed
th
e
in
teg
r
at
io
n
o
f
d
i
f
f
er
en
t
d
ata
m
i
n
i
n
g
al
g
o
r
it
h
m
s
f
o
r
s
en
ti
m
e
n
t
class
i
f
icatio
n
o
f
m
u
l
ti
-
v
ie
w
p
o
lar
ity
te
x
t
u
al
d
ata.
A
d
ata
m
in
in
g
f
r
a
m
e
w
o
r
k
i
s
p
r
o
p
o
s
ed
f
o
r
g
en
er
ati
n
g
Sin
g
le
-
k
e
y
I
n
f
o
r
m
atio
n
s
eq
u
e
n
ce
s
o
f
k
n
o
w
led
g
e
b
y
X
-
m
ea
n
s
clu
s
ter
in
g
a
n
d
m
a
x
i
m
u
m
tar
g
et
v
alu
e
o
f
i
n
ter
est
b
y
d
is
co
v
er
i
n
g
Mu
l
ti
-
Ke
y
Fre
q
u
en
t
I
n
f
o
r
m
atio
n
R
u
les
in
ea
ch
cl
u
s
ter
b
y
u
s
i
n
g
Ho
tSp
o
t
as
s
o
ciatio
n
r
u
les
w
i
ll
p
er
f
o
r
m
clas
s
i
f
icatio
n
w
it
h
i
m
p
r
o
v
ed
ac
cu
r
ac
ies
o
f
t
h
e
cla
s
s
if
ier
s
.
T
h
e
a
n
al
y
s
i
s
r
es
u
lt
s
h
a
v
e
s
h
o
w
n
i
m
p
r
o
v
ed
ac
cu
r
ac
ies
o
f
c
lass
if
y
i
n
g
th
e
s
en
ti
m
e
n
tal
m
u
lti
-
v
ie
w
te
x
t
u
al
d
ata
i
n
to
p
o
s
itiv
e
o
r
n
e
g
ati
v
e
in
f
o
r
m
a
tio
n
d
o
cu
m
en
t
ca
teg
o
r
ies.
RE
F
E
R
E
NC
E
S
[1
]
Öz
e
rk
Ya
v
u
z
,
A
d
e
m
Ka
r
a
h
o
c
a
,
Dilek
Ka
ra
h
o
c
a
,
“
A
Da
ta
M
in
in
g
A
p
p
ro
a
c
h
F
o
r
De
sire
A
n
d
In
ten
ti
o
n
T
o
P
a
rti
c
i
p
a
te
In
V
i
rtu
a
l
Co
m
m
u
n
it
ies
,
”
In
t
e
rn
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
El
e
c
tr
ica
l
a
n
d
Co
mp
u
t
er
En
g
in
e
e
rin
g
(
IJ
ECE
)
,
v
o
l.
9
,
n
o
.
5
,
p
p
.
3
7
1
4
-
3
7
1
9
,
Oc
t
.
2
0
1
9
.
[2
]
T
.
Cru
z
,
e
t
a
l.
,
“
Ex
p
lo
r
in
g
d
a
ta
a
n
a
ly
ti
c
s
o
f
d
a
ta
v
a
riet
y
,
”
in
W
o
rld
C
o
n
fer
e
n
c
e
o
n
In
f
o
ra
ti
o
n
S
y
ste
ms
a
n
d
T
e
c
h
n
o
l
o
g
ies
,
p
p
.
9
2
0
-
9
3
0
,
2
0
1
8
.
[3
]
M
.
S
a
m
m
o
u
r,
e
t
a
l.
,
“
DN
S
Tu
n
n
e
l
in
g
:
a
Re
v
ie
w
o
n
F
e
a
tu
re
s
,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
E
n
g
i
n
e
e
rin
g
a
n
d
Te
c
h
n
o
l
o
g
y
,
v
o
l.
7
,
n
o
.
3
.
2
0
,
p
p
.
1
-
5
,
2
0
1
8
.
[4
]
B.
M
c
w
il
li
a
m
s
a
n
d
G
.
M
o
n
tan
a
,
“
M
u
lt
i
-
v
iew
p
re
d
ictiv
e
p
a
rti
ti
o
n
i
n
g
in
h
ig
h
d
im
e
n
sio
n
s,”
S
ta
t
isti
c
a
l
An
a
l
y
sis
a
n
d
Da
ta
M
in
in
g
:
T
h
e
AS
A
Da
t
a
S
c
ie
n
c
e
J
o
u
rn
a
l
,
v
o
l
.
5
,
n
o
.
4
,
p
p
.
1
-
3
1
,
2
0
1
2
.
[5
]
T
.
Ba
lt
ru
sa
it
is,
e
t
a
l.
,
“
M
u
lt
im
o
d
a
l
M
a
c
h
in
e
L
e
a
rn
in
g
:
A
S
u
rv
e
y
a
n
d
T
a
x
o
n
o
m
y
,
”
IEE
E
T
ra
n
sa
c
ti
o
n
s
o
n
Pa
tt
e
r
n
An
a
lys
is
a
n
d
M
a
c
h
i
n
e
In
tell
ig
e
n
c
e
,
v
o
l.
4
1
,
n
o
.
2
.
p
p
.
4
2
3
-
4
4
3
,
2
0
1
9
.
[6
]
G
.
M
in
e
r,
e
t
a
l.
,
“
P
ra
c
ti
c
a
l
T
e
x
t
M
in
i
n
g
a
n
d
S
t
a
ti
stica
l
A
n
a
ly
sis
f
o
r
No
n
-
st
ru
c
tu
re
d
T
e
x
t
Da
ta
A
p
p
li
c
a
ti
o
n
s
,”
Aca
d
e
mic
Pre
ss
,
2
0
1
2
.
[7
]
J.
L
.
S
o
lk
a
,
“
T
e
x
t
d
a
ta
m
in
in
g
:
T
h
e
o
ry
a
n
d
m
e
th
o
d
s,”
S
t
a
t
isti
c
s
S
u
rv
eys
,
v
o
l.
2
,
p
p
.
9
4
-
1
1
2
,
2
0
0
8
.
[8
]
A
.
Am
a
d
o
,
P
a
u
l
o
Co
rtez
,
P
a
u
lo
R
.
,
S
e
rg
io
M
.
,
“
Re
se
a
rc
h
tren
d
s
o
n
Big
Da
t
a
in
M
a
rk
e
ti
n
g
:
A
t
e
x
t
m
in
in
g
a
n
d
to
p
ic
m
o
d
e
li
n
g
b
a
se
d
li
tera
tu
re
a
n
a
ly
si
s,”
Eu
ro
p
e
a
n
Res
e
a
rc
h
o
n
M
a
n
a
g
e
me
n
t
a
n
d
Bu
si
n
e
ss
Eco
n
o
mic
s
,
v
o
l.
2
4
,
n
o
.
1
,
p
p
.
1
-
7
,
2
0
1
8
.
[9
]
H.
Ha
sh
im
i,
e
t
a
l.
,
“
S
e
lec
ti
o
n
c
rit
e
ria
f
o
r
tex
t
m
in
in
g
a
p
p
ro
a
c
h
e
s,”
Co
mp
u
t
e
rs
in
Hu
ma
n
Beh
a
v
io
r
,
v
o
l.
5
1
,
p
p
.
7
2
9
-
7
3
3
,
2
0
1
5
.
[1
0
]
M
.
Ja
k
u
b
ik
,
“
Be
c
o
m
in
g
to
k
n
o
w
.
S
h
if
ti
n
g
th
e
k
n
o
w
led
g
e
c
re
a
ti
o
n
p
a
ra
d
ig
m
,
”
J
o
u
rn
a
l
o
f
Kn
o
wled
g
e
M
a
n
a
g
e
me
n
t
,
v
o
l.
1
5
,
n
o
.
3
,
p
p
.
3
7
4
-
4
0
2
,
2
0
1
1
.
[1
1
]
S
.
S
a
n
g
a
m
a
n
d
S
.
S
h
in
d
e
,
“
S
e
n
t
im
e
n
t
Clas
si
f
ica
ti
o
n
Of
S
o
c
ial
M
e
d
ia
Re
v
iew
s
Us
in
g
A
n
En
se
m
b
le
Clas
si
f
ier
,
”
In
d
o
n
e
s
ia
n
J
o
u
rn
a
l
o
f
El
e
c
tr
ica
l
En
g
i
n
e
e
rin
g
a
n
d
Co
mp
u
t
er
S
c
i
e
n
c
e
(
IJ
EE
CS
)
,
v
o
l.
1
6
,
n
o
.
1
,
p
p
.
3
5
5
-
3
6
3
,
Oc
t
.
2
0
1
9
.
[1
2
]
C.
G
.
R
e
d
d
ick
,
e
t
a
l.
,
“
A
S
o
c
ial
M
e
d
ia
T
e
x
t
A
n
a
l
y
ti
c
s
F
ra
m
e
w
o
r
k
f
o
r
Do
u
b
le
-
L
o
o
p
L
e
a
rn
in
g
f
o
r
Cit
ize
n
-
Ce
n
tri
c
P
u
b
l
ic S
e
rv
ice
s:
A
Ca
se
S
tu
d
y
o
f
a
L
o
c
a
l
G
o
v
e
rn
m
e
n
t
F
a
c
e
b
o
o
k
U
se
,
”
Go
v
e
rn
me
n
t
In
f
o
rm
a
ti
o
n
Q
u
a
rte
rly
,
v
o
l.
3
4
,
n
o
.
1
,
p
p
.
1
1
0
-
1
2
5
,
2
0
1
7
.
[1
3
]
O.
M
ü
ll
e
r,
e
t
a
l.
,
“
Us
in
g
T
e
x
t
An
a
ly
t
ics
T
o
De
riv
e
Cu
sto
m
e
r
S
e
rv
ice
M
a
n
a
g
e
m
e
n
t
Be
n
e
f
it
s
F
ro
m
Un
stru
c
tu
re
d
Da
ta
,
”
M
IS
Q
u
a
rte
rly
Exe
c
u
ti
v
e
,
v
o
l.
1
5
,
n
o
.
4
,
p
p
.
2
4
3
-
2
5
8
,
2
0
1
6
.
[1
4
]
J.
S
.
S
o
w
m
i
y
a
a
n
d
S
.
Ch
a
n
d
ra
k
a
la
,
“
Jo
in
t
S
e
n
ti
m
e
n
t/
T
o
p
ic
Ex
trac
ti
o
n
F
ro
m
T
e
x
t
,
”
2
0
1
4
IEE
E
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
Ad
v
a
n
c
e
d
C
o
mm
u
n
ica
ti
o
n
s C
o
n
tr
o
l
a
n
d
Co
m
p
u
ti
n
g
T
e
c
h
n
o
lo
g
ies
,
p
p
.
6
1
1
-
6
1
5
,
2
0
1
4
.
[1
5
]
A
.
S
.
Ha
li
b
a
s,
e
t
a
l.
,
“
A
p
p
li
c
a
ti
o
n
o
f
tex
t
c
las
sif
i
c
a
ti
o
n
a
n
d
c
l
u
ste
rin
g
o
f
Tw
it
ter
d
a
ta
f
o
r
b
u
sin
e
ss
a
n
a
l
y
ti
c
s,”
in
Pro
c
e
e
d
i
n
g
s
o
f
2
0
1
8
M
a
ja
n
In
t
e
rn
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
(
M
IC
2
0
1
8
)
,
p
p
.
1
-
7,
2
0
1
8
.
[1
6
]
R.
A
h
u
ja
a
n
d
W
.
A
n
a
n
d
,
“
S
e
n
ti
m
e
n
t
c
l
a
ss
i
f
ica
ti
o
n
o
f
m
o
v
ie
re
v
ie
w
s
u
sin
g
d
u
a
l
train
in
g
a
n
d
d
u
a
l
p
re
d
it
i
o
n
,
”
in
2
0
1
7
4
th
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
Im
a
g
e
I
n
fo
rm
a
ti
o
n
Pr
o
c
e
ss
in
g
(
ICIIP
2
0
1
7
)
,
p
p
.
5
9
4
-
5
9
7
,
2
0
1
7
.
[1
7
]
P
.
Zo
la,
e
t
a
l.
,
“
S
o
c
ial
M
e
d
ia
Cro
ss
-
S
o
u
rc
e
a
n
d
Cro
ss
-
Do
m
a
in
S
e
n
ti
m
e
n
t
Clas
si
f
ica
ti
o
n
,
”
in
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
In
fo
rm
a
t
io
n
T
e
c
h
n
o
l
o
g
y
a
n
d
D
e
c
isio
n
M
a
k
in
g
,
v
o
l
.
1
8
,
n
o
.
5
,
p
p
.
1
4
6
9
-
1
4
9
9
,
2
0
1
9
.
[1
8
]
R.
W
a
n
g
,
e
t
a
l.
,
“
Co
n
v
o
lu
ti
o
n
a
l
Re
c
u
rre
n
t
Ne
u
ra
l
Ne
t
w
o
rk
s
f
o
r
T
e
x
t
Clas
si
f
ic
a
ti
o
n
,
”
in
Pro
c
e
e
d
i
n
g
s
o
f
th
e
2
0
1
9
In
ter
n
a
t
io
n
a
l
J
o
in
t
Co
n
fer
e
n
c
e
o
n
Ne
u
ra
l
Ne
two
rk
s
,
p
p
.
1
-
6
,
2
0
1
9
.
[1
9
]
A
.
W
.
Ha
r
y
a
n
to
,
e
t
a
l.
,
“
In
f
lu
e
n
c
e
o
f
W
o
rd
No
rm
a
li
z
a
ti
o
n
a
n
d
C
h
i
-
S
q
u
a
re
d
F
e
a
tu
re
S
e
lec
ti
o
n
o
n
S
u
p
p
o
rt
V
e
c
t
o
r
M
a
c
h
in
e
(S
V
M
)
T
e
x
t
Clas
sif
ic
a
ti
o
n
,
”
i
n
Pro
c
e
e
d
i
n
g
s
of
2
0
1
8
I
n
ter
n
a
ti
o
n
a
l
S
e
min
a
r o
n
Ap
p
li
c
a
ti
o
n
fo
r
T
e
c
h
n
o
lo
g
y
o
f
In
fo
rm
a
t
io
n
a
n
d
Co
mm
u
n
ica
ti
o
n
,
p
p
.
2
2
9
-
2
3
3
,
2
0
1
8
.
[2
0
]
M
.
F
ik
ri
a
n
d
R.
S
a
rn
o
,
“
A
Co
m
p
a
ra
ti
v
e
S
tu
d
y
o
f
S
e
n
ti
m
e
n
t
A
n
a
l
y
sis
u
sin
g
S
V
M
a
n
d
S
e
n
ti
W
o
r
d
Ne
t,
”
In
d
o
n
e
s
ia
n
J
o
u
rn
a
l
o
f
El
e
c
tr
ica
l
En
g
in
e
e
r
in
g
a
n
d
Co
m
p
u
t
er
S
c
i
e
n
c
e
(
IJ
EE
CS
)
,
v
o
l.
1
3
,
n
o
.
3
,
p
p
.
9
0
2
-
9
0
9
,
2
0
1
9
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
10
,
No
.
5
,
Octo
b
e
r
2
0
2
0
:
5
5
2
6
-
5534
5534
[2
1
]
B.
T
a
n
g
,
e
t
a
l.
,
“
T
o
w
a
rd
Op
ti
m
a
l
F
e
a
tu
re
S
e
lec
ti
o
n
i
n
Na
iv
e
Ba
y
e
s
f
o
r
T
e
x
t
Ca
teg
o
riza
ti
o
n
,
”
IEE
E
T
ra
n
s
a
c
ti
o
n
s o
n
Kn
o
wl
e
d
g
e
a
n
d
D
a
ta
En
g
i
n
e
e
rin
g
,
v
o
l.
2
8
,
n
o
.
9
,
p
p
.
2
5
0
8
-
2
5
2
1
,
S
e
p
.
2
0
1
6
.
[2
2
]
Z.
T
a
n
g
,
W
.
X
iao
,
Bin
L
u
,
a
n
d
Y.
Zu
o
,
“
A
P
a
ra
ll
e
l
A
lg
o
rit
h
m
f
o
r
Ba
y
e
sia
n
Tex
t
Clas
sif
i
c
a
ti
o
n
Ba
se
d
on
No
ise
El
im
in
a
ti
o
n
a
n
d
Dim
e
n
sio
n
Re
d
u
c
ti
o
n
i
n
S
p
a
rk
Co
m
p
u
ti
n
g
En
v
iro
n
m
e
n
t
,
”
in
L
e
c
tu
re
No
tes
in
Co
mp
u
te
r
S
c
ien
c
e
(
in
c
lu
d
in
g
s
u
b
se
rie
s
L
e
c
tu
re
No
tes
in
A
rtif
icia
l
In
telli
g
e
n
c
e
a
n
d
L
e
c
t
u
re
No
tes
in
B
io
in
f
o
rm
a
ti
c
s)
,
v
o
l.
1
1
5
1
3
L
NCS,
p
p
.
2
2
2
-
2
3
9
,
2
0
1
9
.
[2
3
]
F
.
Zab
li
th
a
n
d
I.
H.
Os
m
a
n
,
“
R
e
v
ie
w
M
o
d
u
s:
T
e
x
t
c
la
ss
i
f
ica
ti
o
n
a
n
d
se
n
ti
m
e
n
t
p
re
d
ictio
n
o
f
u
n
str
u
c
tu
re
d
re
v
iew
s
u
sin
g
a
h
y
b
rid
c
o
m
b
in
a
ti
o
n
o
f
m
a
c
h
in
e
lea
rn
in
g
a
n
d
e
v
a
lu
a
ti
o
n
m
o
d
e
ls,”
Ap
p
l
ied
M
a
th
e
ma
ti
c
a
l
M
o
d
e
l
li
n
g
,
v
o
l.
7
1
,
p
p
.
5
6
9
-
5
8
3
,
Ju
l
.
2
0
1
9
.
[2
4
]
D.
P
e
l
leg
a
n
d
A
.
W
.
M
o
o
re
,
“
X
-
m
e
a
n
s:
Ex
ten
d
in
g
K
-
m
e
a
n
s
w
it
h
Eff
icie
n
t
Esti
m
a
ti
o
n
o
f
th
e
Nu
m
b
e
r
o
f
Clu
ste
rs,”
Pro
c
e
e
d
in
g
s
o
f
t
h
e
S
e
v
e
n
tee
n
t
h
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
M
a
c
h
i
n
e
L
e
a
rn
i
n
g
,
p
p
.
7
2
7
-
7
3
4
,
20
00
.
[2
5
]
Y.
Ca
i,
e
t
a
l.
,
“
Disc
ip
li
n
e
H
o
t
sp
o
ts
M
in
i
n
g
Ba
se
d
o
n
Hie
ra
rc
h
ica
l
Dirich
let
T
o
p
ic
Clu
ste
ri
n
g
a
n
d
C
o
-
w
o
r
d
Ne
tw
o
rk
,
”
J
o
u
rn
a
l
o
f
S
o
ft
wa
re
,
v
o
l.
1
1
,
n
o
.
1
1
,
p
p
.
1
0
8
9
-
1
1
0
1
,
2
0
1
6
.
[2
6
]
M
.
Ha
ll
,
“
Clas
s
Ho
tS
p
o
t
,
”
we
k
a
.
so
u
rc
e
fo
rg
e
,
2
0
1
5
.
[
O
n
li
n
e
]
,
A
v
a
il
a
b
le:
h
tt
p
:
//
w
e
k
a
.
so
u
rc
e
f
o
rg
e
.
n
e
t/
d
o
c
.
p
a
c
k
a
g
e
s/h
o
tS
p
o
t
/w
e
k
a
/a
ss
o
c
iatio
n
s/Ho
tS
p
o
t
.
h
tm
l
.
[2
7
]
J.
H.
F
ried
m
a
n
a
n
d
N.
I.
F
ish
e
r,
“
Bu
m
p
h
u
n
ti
n
g
in
h
ig
h
-
d
im
e
n
sio
n
a
l
d
a
ta,”
S
ta
t
isti
c
s
a
n
d
Co
m
p
u
t
i
n
g
,
v
o
l.
9
,
p
p
.
1
2
3
-
1
4
3
,
1
9
9
9
.
[2
8
]
M
.
A
.
Bu
rh
a
n
u
d
d
in
,
R
.
Ism
a
il
,
N
.
Iz
z
a
i
m
a
h
,
A
li
A
b
d
u
l
-
Ja
b
b
a
r
M
.
,
a
n
d
N
.
Zain
o
l
.
,
“
A
n
a
l
y
sis
o
f
M
o
b
il
e
S
e
rv
ice
P
r
o
v
id
e
rs
P
e
rf
o
rm
a
n
c
e
Us
in
g
N
a
iv
e
Ba
y
e
s
Da
ta
M
in
in
g
T
e
c
h
n
iq
u
e
,
”
In
t
e
rn
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
El
e
c
tr
ica
l
a
n
d
Co
mp
u
t
er
E
n
g
in
e
e
rin
g
(
IJ
ECE
)
,
v
o
l.
8
,
n
o
.
6
,
p
p
.
5
1
5
3
-
5
1
6
1
,
2
0
1
8
.
[2
9
]
B.
P
a
n
g
a
n
d
L
.
L
e
e
,
“
A
se
n
ti
m
e
n
tal
e
d
u
c
a
ti
o
n
,
”
in
Pro
c
e
e
d
in
g
s
o
f
th
e
4
2
n
d
A
n
n
u
a
l
M
e
e
ti
n
g
o
n
Asso
c
ia
ti
o
n
f
o
r
Co
mp
u
t
a
ti
o
n
a
l
L
in
g
u
isti
c
s
,
p
p
.
2
7
1
-
e
s,
2
0
0
4
.
Evaluation Warning : The document was created with Spire.PDF for Python.