I
AE
S
I
nte
rna
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
9
,
No
.
3
,
Sep
tem
b
er
2020
,
p
p
.
473
~
479
I
SS
N:
2252
-
8938
,
DOI
: 1
0
.
1
1
5
9
1
/i
j
ai.
v
9
.i
3
.
p
p
473
-
4
7
9
473
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
M
a
chine learning
-
ba
sed te
chnique
for big
data
sen
ti
m
en
ts
ex
traction
No
ra
ini
Se
m
a
n
,
Nurul
At
iqa
h Ra
z
m
i
F
a
c
u
lt
y
o
f
Co
m
p
u
ter an
d
M
a
th
e
m
a
ti
c
a
l
S
c
ien
c
e
s
,
Un
iv
e
rsiti
T
e
k
n
o
lo
g
i
M
A
R
A
(UiT
M
)
,
S
h
a
h
A
la
m
,
S
e
lan
g
o
r
,
M
a
lay
sia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
F
eb
1
5
,
2
0
20
R
ev
i
s
ed
A
pr
24
,
2
0
20
A
cc
ep
ted
M
a
y
20
,
2
0
20
A
h
u
g
e
a
m
o
u
n
t
o
f
d
a
ta
is
g
e
n
e
ra
ted
e
v
e
r
y
m
in
u
te
f
o
r
so
c
ial
n
e
t
wo
rk
in
g
a
n
d
c
o
n
ten
t
s
h
a
rin
g
v
ia
S
o
c
ial
m
e
d
ia
sites
th
a
t
c
a
n
b
e
in
a
f
o
rm
o
f
stru
c
tu
re
d
,
u
n
stru
c
t
u
re
d
o
r
se
m
i
-
stru
c
tu
re
d
d
a
ta.
On
e
o
f
th
e
larg
e
st
u
se
d
so
c
ial
m
e
d
ia
sites
is
Tw
it
ter,
w
h
e
re
e
a
c
h
a
n
d
e
v
e
r
y
d
a
y
m
il
li
o
n
s
o
f
d
a
ta
g
e
n
e
r
a
ted
in
th
e
f
o
r
m
o
f
u
n
stru
c
tu
re
d
tw
e
e
ts.
Twe
e
ts
o
r
o
p
in
io
n
s
o
f
th
e
p
e
o
p
le
c
a
n
b
e
u
se
d
to
e
x
trac
t
se
n
ti
m
e
n
ts
o
f
th
e
p
e
o
p
le.
S
e
n
ti
m
e
n
t
a
n
a
ly
sis
is
b
e
n
e
f
ici
a
l
f
o
r
o
rg
a
n
iza
ti
o
n
s
t
o
im
p
ro
v
e
th
e
ir
p
ro
d
u
c
ts
a
n
d
m
a
k
e
re
q
u
ired
c
h
a
n
g
e
s
o
n
d
e
m
a
n
d
to
in
c
re
a
se
th
e
ir
p
ro
f
it
.
In
th
is
p
a
p
e
r,
th
re
e
m
a
c
h
in
e
lea
rn
in
g
a
lg
o
rit
h
m
s
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
i
n
e
(S
V
M
),
De
c
isio
n
T
re
e
s
(DT
),
a
n
d
Na
iv
e
Ba
y
e
s
(NB)
f
o
r
c
las
si
fy
in
g
se
n
ti
m
e
n
ts
o
f
tw
it
ters
d
a
ta.
T
h
e
p
u
rp
o
se
o
f
th
is
re
se
a
rc
h
is
to
c
o
m
p
a
re
th
e
o
u
t
c
o
m
e
s
o
f
th
e
se
a
lg
o
rit
h
m
s
to
id
e
n
ti
fy
b
e
st
m
a
c
h
in
e
lea
rn
in
g
m
e
th
o
d
w
h
ich
g
iv
e
s
m
o
st
a
c
c
u
ra
te
a
n
d
e
ff
icie
n
t
re
su
lt
s
f
o
r
c
las
si
fy
in
g
t
w
it
ter
d
a
ta.
Ou
r
e
x
p
e
ri
m
e
n
tal
re
su
lt
sh
o
w
s
th
a
t
sa
m
e
p
re
p
ro
c
e
ss
in
g
m
e
th
o
d
s
o
n
a
d
if
fe
re
n
t
d
a
tas
e
t
a
ff
e
c
t
si
m
il
a
rl
y
th
e
c
las
s
if
ier
s
p
e
rf
o
r
m
a
n
c
e
.
Af
ter
a
n
a
l
y
z
in
g
th
e
re
su
lt
s
it
is
o
b
se
rv
e
d
th
a
t
S
VM
p
ro
v
id
e
s
6
4
.
9
6
%
,
7
1
.
2
6
%
a
n
d
9
1
.
2
5
%
p
re
c
isio
n
w
h
ich
is
b
e
tt
e
r
th
a
n
o
t
h
e
r
tw
o
a
lg
o
rit
h
m
s.
A
lso
,
o
v
e
ra
ll
Re
c
a
ll
a
n
d
F
-
m
e
a
su
re
ra
te
o
f
S
V
M
is
g
re
a
ter
th
a
n
NB
a
n
d
DT
f
o
r
th
re
e
d
a
ta
se
ts.
Ho
w
e
v
e
r,
it
is
i
m
p
o
rtan
t
to
f
u
rth
e
r
stu
d
y
c
u
rre
n
t
a
v
a
il
a
b
le
p
re
p
r
o
c
e
ss
in
g
t
e
c
h
n
iq
u
e
s
th
a
t
h
e
lp
u
s
t
o
im
p
ro
v
e
re
su
lt
s
o
f
v
a
rio
u
s c
las
sif
iers
.
K
ey
w
o
r
d
s
:
C
las
s
i
f
ier
F
-
m
ea
s
u
r
e
Ma
ch
i
n
e
lear
n
i
n
g
R
ec
all
Sen
ti
m
e
n
t a
n
a
l
y
s
is
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
:
No
r
ain
i Se
m
an
,
Dep
ar
t
m
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
,
Facu
lt
y
o
f
C
o
m
p
u
ter
&
Ma
t
h
e
m
atica
l Sc
ien
ce
s
,
Un
i
v
er
s
iti
T
ek
n
o
lo
g
i M
A
R
A
,
Sh
a
h
A
la
m
,
Ma
la
y
s
ia
.
E
m
ail: a
i
n
i
@
t
m
s
k
.
u
i
t
m
.
ed
u
.
m
y
1.
I
NT
RO
D
UCT
I
O
N
T
w
it
ter
,
th
e
lar
g
es
t
u
s
ed
s
o
ci
al
m
ed
ia
s
ite,
h
as
n
o
w
b
ec
o
m
e
a
v
er
y
p
o
p
u
lar
tr
en
d
o
v
er
th
e
w
o
r
ld
f
o
r
p
eo
p
le
w
h
o
w
a
n
t
to
s
h
ar
e
an
o
p
in
io
n
ab
o
u
t
t
h
eir
s
o
cial,
p
o
liti
ca
l
an
d
ec
o
n
o
m
ic
i
n
ter
es
t.
User
o
p
in
io
n
ca
n
b
e
r
elate
d
to
v
ar
io
u
s
a
s
p
ec
ts
lik
e
g
ad
g
et
s
,
p
o
liti
cs,
p
r
o
d
u
cts,
s
er
v
ices
etc.
t
h
at
ca
n
d
i
r
ec
tl
y
co
n
v
e
y
t
h
e
v
ie
w
p
o
in
t
o
f
t
h
e
u
s
er
an
d
h
e
lp
s
in
m
a
k
i
n
g
p
r
ed
ictio
n
s
o
f
a
co
n
s
u
m
er
m
ar
k
et.
S
u
ch
k
in
d
o
f
o
p
in
io
n
s
o
r
s
en
ti
m
e
n
ts
o
f
h
u
g
e
p
eo
p
le
ar
o
u
n
d
th
e
w
o
r
ld
is
ca
p
ab
le
o
f
p
er
f
o
r
m
i
n
g
an
a
l
y
s
is
a
n
d
f
u
tu
r
e
p
r
ed
ictio
n
s
.
Usu
al
l
y
,
t
w
ee
ts
co
n
ta
in
in
co
m
p
lete,
p
o
o
r
l
y
s
tr
u
ct
u
r
ed
,
n
o
is
y
,
ir
r
eg
u
lar
e
x
p
r
ess
io
n
s
,
ill
-
f
o
r
m
ed
w
o
r
d
s
a
n
d
n
o
n
-
d
ictio
n
ar
y
ter
m
s
[
1
]
.
Als
o
,
m
e
s
s
a
g
es
o
r
t
w
ee
t
s
ar
e
s
h
o
r
t
an
d
h
a
v
e
1
4
0
len
g
th
s
o
f
li
m
itatio
n
s
.
So
it r
eq
u
ir
es
p
r
ep
r
o
ce
s
s
in
g
d
o
n
e
o
n
o
u
r
co
llected
d
ataset
s
t
o
r
ed
u
ce
n
o
is
e
in
t
w
ee
t
s
b
y
r
em
o
v
i
n
g
s
to
p
-
w
o
r
d
s
,
r
e
m
o
v
i
n
g
UR
L
s
,
r
ep
lacin
g
n
e
g
atio
n
s
etc
[
2
]
.
Se
n
ti
m
en
t
d
ic
tio
n
ar
y
co
n
tai
n
s
al
l
f
o
r
m
s
o
f
a
w
o
r
d
w
it
h
ea
c
h
w
o
r
d
’
s
p
o
lar
it
y
s
tr
en
g
t
h
t
h
at
ca
n
s
a
v
e
m
o
r
e
ti
m
e.
Sen
ti
m
e
n
t
a
n
al
y
s
i
s
(
S
A
)
is
a
p
r
o
ce
s
s
o
f
d
etec
ti
n
g
th
e
co
n
tex
t
u
al
p
o
lar
it
y
o
f
te
x
t
i
n
ter
m
s
o
f
p
o
s
itiv
e,
n
eg
at
iv
e
o
r
n
e
u
tr
al
[
3
-
4
]
.
Or
g
an
izatio
n
s
ac
r
o
s
s
th
e
w
o
r
ld
wid
el
y
ad
o
p
ted
th
e
ab
ilit
y
to
ex
tr
ac
t
in
s
i
g
h
ts
f
r
o
m
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
I
n
t J
A
r
ti
f
I
n
tell
,
Vo
l.
9
,
No
.
3
,
Sep
te
m
b
er
20
20
:
4
7
3
–
4
7
9
474
th
ese
s
en
ti
m
e
n
t
s
o
f
v
ar
io
u
s
s
o
cial
m
ed
ia
s
ites
.
I
t
h
elp
s
o
r
g
a
n
izatio
n
s
to
m
a
k
e
p
r
ed
ictio
n
s
o
f
a
ce
r
tai
n
p
r
o
d
u
ct,
r
ev
ie
w
s
,
a
n
d
o
t
h
er
d
ec
is
io
n
-
m
ak
in
g
p
r
o
ce
s
s
e
s
t
h
at
w
ill
u
l
ti
m
atel
y
in
cr
ea
s
e
t
h
e
p
r
o
f
it.
So
u
lt
i
m
a
tel
y
S
A
is
b
en
ef
icia
l
f
o
r
o
r
g
an
izatio
n
s
a
n
d
in
d
iv
id
u
als
to
i
m
p
r
o
v
e
t
h
e
ir
p
r
o
f
it
as
p
er
u
s
er
o
r
m
ar
k
e
t
d
em
a
n
d
.
S
A
also
k
n
o
w
n
as
o
p
in
io
n
m
in
in
g
,
is
a
m
o
s
t
p
o
p
u
lar
tr
en
d
in
to
d
a
y
’
s
w
o
r
ld
w
h
ic
h
i
s
th
e
p
r
o
ce
s
s
o
f
id
en
ti
f
y
in
g
a
n
d
c
ateg
o
r
izin
g
o
p
in
io
n
s
o
n
t
h
e
w
eb
,
d
eter
m
in
e
s
th
e
w
r
iter
attitu
d
e
to
w
ar
d
s
a
p
ar
ticu
lar
to
p
ic
o
r
p
r
o
d
u
ct
[
5
]
.
I
t
tells
ab
o
u
t
w
h
at
au
t
h
o
r
w
a
n
ts
to
co
m
m
u
n
icate
an
d
d
e
f
i
n
e
s
h
is
s
tate
o
f
m
in
d
i
n
ter
m
s
o
f
e
m
o
tio
n
s
,
f
ee
li
n
g
s
,
an
d
s
u
b
j
ec
tiv
ities
ab
o
u
t
a
n
e
v
en
t
o
r
t
o
p
ic.
I
t
in
v
o
lv
ed
w
i
th
Nat
u
r
al
L
a
n
g
u
a
g
e
P
r
o
ce
s
s
in
g
(
N
L
P
)
p
r
o
ce
s
s
w
h
ic
h
is
t
h
e
i
n
ter
ac
tio
n
b
et
w
ee
n
t
h
e
co
m
p
u
ter
s
an
d
t
h
e
h
u
m
a
n
/
n
atu
r
al
la
n
g
u
a
g
e
[
6
-
8
]
.
NL
P
tech
n
iq
u
e
f
ac
ilit
ate
s
ea
s
y
p
r
e
-
p
r
o
ce
s
s
i
n
g
o
f
tex
t
i.e
.
N
L
P
clea
n
s
an
d
n
o
r
m
alize
s
te
x
t
f
o
r
s
en
ti
m
e
n
t
an
al
y
s
i
s
[
8
]
.
A
n
al
y
s
i
s
o
f
s
en
ti
m
en
t
s
ca
n
b
e
b
ased
o
n
s
i
n
g
le
p
h
r
ase
o
r
s
en
te
n
ce
,
w
h
er
e
th
e
s
en
ti
m
e
n
t
o
f
th
e
w
h
o
le
s
e
n
ten
ce
is
ca
lcu
lated
.
I
t c
o
n
tain
s
f
o
llo
w
i
n
g
s
tep
s
[
9
-
1
0
]
:
T
w
ee
t
s
p
o
s
ted
o
n
t
w
itter
ar
e
f
r
ee
l
y
a
v
ailab
le
t
h
r
o
u
g
h
a
s
et
o
f
A
P
I
s
o
f
t
w
itter
.
A
t
f
i
r
s
t,
w
e
co
llected
a
co
r
p
u
s
o
f
p
o
s
itiv
e,
n
eg
ati
v
e,
n
eu
tr
al
an
d
ir
r
elev
a
n
t t
w
ee
t
s
f
r
o
m
t
w
itter
A
P
I
.
T
h
en
p
r
e
-
p
r
o
ce
s
s
in
g
d
o
n
e
b
y
r
e
m
o
v
i
n
g
s
to
p
w
o
r
d
s
,
n
e
g
ati
o
n
s
,
U
R
L
,
f
u
ll
s
to
p
,
co
m
m
as
etc.
to
r
ed
u
ce
n
o
is
e
f
r
o
m
t
w
ee
ts
a
n
d
to
p
r
ep
ar
e
o
u
r
d
ata
f
o
r
s
en
ti
m
e
n
t c
la
s
s
if
ica
tio
n
.
T
h
en
,
w
e
ap
p
l
y
m
ac
h
in
e
lear
n
in
g
al
g
o
r
it
h
m
s
to
o
u
r
d
ataset
an
d
co
m
p
ar
e
th
eir
r
esu
l
ts
.
R
es
u
lts
h
elp
u
s
to
id
en
ti
f
y
w
h
i
ch
m
ac
h
i
n
e
lear
n
i
n
g
alg
o
r
it
h
m
is
b
est s
u
ited
f
o
r
class
i
f
icat
io
n
o
f
S
A
.
A
p
p
licatio
n
s
o
f
S
A
ar
e
b
r
o
ad
an
d
p
o
w
er
f
u
l
t
h
at
p
r
o
v
id
e
u
s
ea
s
ier
an
d
q
u
ick
er
s
o
c
ial
m
ed
ia
m
o
n
ito
r
i
n
g
li
k
e
i
n
:
C
o
n
s
u
m
er
m
ar
k
e
t
f
o
r
p
r
o
d
u
ct
r
ev
ie
w
s
;
Ma
r
k
etin
g
to
k
n
o
w
co
n
s
u
m
er
tr
en
d
s
an
d
at
tit
u
d
e;
So
cial
m
ed
ia
to
f
in
d
g
en
er
al
u
s
er
o
p
in
io
n
ab
o
u
t
cu
r
r
en
t
to
p
ics;
Mo
v
ie
to
k
n
o
w
w
h
et
h
er
r
elea
s
ed
m
o
v
ie
i
s
lik
ed
o
r
n
o
t,
etc
[
1
1
]
.
A
s
u
s
e
r
s
o
n
s
o
cial
m
ed
ia
s
ite
s
ar
e
r
ap
id
ly
g
r
o
w
in
g
a
n
d
p
r
o
d
u
cin
g
a
lar
g
e
a
m
o
u
n
t
o
f
d
ata
ev
er
y
d
a
y
,
s
o
th
er
e
i
s
a
n
ee
d
to
class
i
f
y
a
n
d
an
al
y
ze
t
h
ese
m
e
s
s
a
g
es
to
f
i
n
d
o
u
t
it
s
p
o
lar
it
y
ab
o
u
t
s
o
m
e
to
p
ic
o
r
ev
en
t
[
1
2
-
1
3
]
.
E
m
o
ti
o
n
s
a
n
d
o
p
in
io
n
s
ca
n
b
e
e
x
p
r
ess
ed
in
m
a
n
y
w
a
y
s
.
C
las
s
i
f
y
in
g
s
e
n
ti
m
en
ts
t
h
a
t
h
av
e
f
e
w
r
elati
v
e
cla
s
s
e
s
s
u
ch
as
“
p
o
s
i
tiv
e”
,
”n
e
g
ati
v
e
”,
o
r
”n
eu
tr
al”,
is
th
e
m
o
s
t
co
m
p
licated
ta
s
k
.
S
A
is
a
p
o
p
u
lar
to
p
ic
an
d
lo
ts
o
f
r
esear
ch
h
as
b
ee
n
g
o
in
g
o
n
f
r
o
m
a
lo
n
g
ti
m
e.
Ma
n
y
r
e
s
ea
r
ch
er
s
u
s
ed
s
u
p
er
v
i
s
ed
lear
n
i
n
g
al
g
o
r
ith
m
s
a
ls
o
w
it
h
v
ar
io
u
s
au
to
m
a
tic
clas
s
i
f
ier
s
f
o
r
c
lass
i
f
icatio
n
o
f
th
e
p
o
lar
ity
o
f
s
e
n
ti
m
e
n
ts
[
1
4
]
.
T
h
e
p
r
o
b
lem
i
s
in
ass
ig
n
i
n
g
th
e
s
tr
o
n
g
est
p
o
lar
it
y
o
f
s
e
n
ti
m
en
ts
an
d
in
f
i
n
d
i
n
g
th
e
b
est alg
o
r
ith
m
w
h
ic
h
p
r
o
v
id
es
m
o
s
t a
cc
u
r
ate
r
es
u
lt
s
.
I
n
th
i
s
p
ap
er
w
e
u
s
e
th
r
e
e
m
ac
h
in
e
lear
n
i
n
g
alg
o
r
it
h
m
s
Su
p
p
o
r
t
Vec
to
r
Ma
ch
i
n
e
(
SVM)
,
Dec
is
io
n
T
r
ee
(
DT
)
[
1
5
]
an
d
Naïv
e
B
a
y
es
C
la
s
s
i
f
ier
(
NB
)
s
en
ti
m
e
n
t
c
lass
if
ier
f
o
r
c
lass
if
y
in
g
o
u
r
d
ata
a
ls
o
h
elp
s
i
n
e
v
al
u
ati
n
g
t
h
e
p
er
f
o
r
m
an
ce
o
f
o
u
r
tr
ai
n
in
g
d
ata
s
et
.
W
e
f
o
cu
s
ed
o
n
co
m
p
ar
i
n
g
o
u
tco
m
es
o
f
th
e
s
e
alg
o
r
ith
m
s
to
id
e
n
ti
f
y
b
est
m
ac
h
in
e
lear
n
i
n
g
m
et
h
o
d
w
h
ich
g
i
v
es
m
o
s
t
ac
c
u
r
ate
a
n
d
ef
f
icien
t
r
es
u
lt
s
f
o
r
class
i
f
y
in
g
t
w
i
tter
d
ata.
2.
RE
S
E
ARCH
M
E
T
H
O
D
T
h
is
p
ap
er
p
r
esen
ts
a
m
o
d
el
p
r
esen
ted
in
F
i
g
u
r
e
1
,
w
h
ic
h
co
n
s
i
s
ts
o
f
th
r
ee
la
y
er
s
f
o
r
an
al
y
zi
n
g
s
en
ti
m
e
n
ts
.
First
Data
C
o
llect
io
n
la
y
er
,
u
s
ed
to
co
llect
t
w
ee
ts
f
r
o
m
t
w
i
tter
A
P
I
s
;
Seco
n
d
Data
p
r
ep
r
o
ce
s
s
in
g
la
y
er
w
it
h
a
s
elec
tio
n
o
f
attr
i
b
u
tes
w
h
ic
h
i
s
u
s
ed
to
r
ed
u
c
e
n
o
is
e
lev
el
f
r
o
m
t
w
ee
ts
,
a
n
d
last
S
A
o
r
Data
Mi
n
i
n
g
la
y
er
u
s
ed
to
ap
p
ly
m
a
ch
in
e
lear
n
i
n
g
al
g
o
r
ith
m
[
2
]
.
Fig
u
r
e
1
.
Gen
er
al
m
o
d
el
o
f
s
e
n
ti
m
en
t a
n
al
y
s
is
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
A
r
ti
f
I
n
tell
I
SS
N:
2252
-
8938
Ma
ch
in
e
lea
r
n
in
g
-
b
a
s
ed
te
ch
n
iq
u
e
fo
r
b
ig
d
a
ta
s
en
timen
ts
…
(
N
o
r
a
in
i S
ema
n
)
475
2
.
1
.
Da
t
a
c
o
llect
io
n
A
t
f
ir
s
t,
w
e
o
b
tain
tr
ain
i
n
g
d
ata
o
f
t
w
i
tter
s
e
n
ti
m
en
ts
f
r
o
m
2
-
d
i
f
f
er
e
n
t
t
witter
A
P
I
.
First,
d
ataset
tak
e
n
f
r
o
m
“
T
w
it
ter
S
e
n
ti
m
en
t
S
y
s
te
m
f
o
r
Se
m
E
v
al
2
0
1
6
”,
(
d
en
o
ted
b
y
“
SE
-
T
”)
co
n
tain
s
ap
p
r
o
x
1
3
5
4
1
tw
ee
t
s
w
ith
2
-
attr
ib
u
tes
n
a
m
el
y
:
clas
s
an
d
co
n
ten
t
[
1
6
]
.
Seco
n
d
d
ataset
is
tak
en
f
r
o
m
“San
d
er
s
An
al
y
tic
s
t
w
i
tter
s
en
ti
m
en
t
co
r
p
u
s
”
(
d
en
o
t
ed
b
y
T
S),
w
h
ich
co
n
tai
n
s
4
7
9
in
s
ta
n
ce
s
w
it
h
class
a
n
d
te
x
t
t
w
o
attr
ib
u
tes
[
1
7
]
as
p
r
esen
ted
i
n
T
ab
le
1
.
Ho
w
e
v
er
,
w
e
also
co
llect
o
u
r
o
w
n
t
w
it
ter
d
ata
in
Ma
la
y
lan
g
u
ag
e,
w
h
ic
h
is
s
p
o
k
e
n
in
Ma
la
y
s
ia,
Si
n
g
ap
o
r
e,
I
n
d
o
n
esia,
an
d
a
f
e
w
o
t
h
er
co
u
n
tr
ies
an
d
d
en
o
ted
as
“
O
C
”.
T
h
is
la
n
g
u
a
g
e
i
s
ac
t
u
all
y
t
h
e
f
o
u
r
th
-
m
o
s
t
p
o
p
u
lar
lan
g
u
ag
e
o
n
T
w
itter
,
ac
co
u
n
tin
g
f
o
r
8
p
er
ce
n
t o
f
all
T
w
ee
t
s
ar
e
ab
o
u
t a
ir
lin
e
s
[
1
8
]
.
T
ab
le
1
.
T
w
itter
d
ata
co
llectio
n
D
a
t
a
se
t
T
o
t
a
l
Tw
e
e
t
s
P
o
si
t
i
v
e
N
e
u
t
r
a
l
N
e
g
a
t
i
v
e
E
-
Tw
i
t
t
e
r
(
S
E
-
T)
[16]
1
3
5
4
1
5
2
3
2
6
2
4
2
2
0
6
7
Tw
i
t
t
e
r
S
a
n
d
e
r
s (T
S
)
[17]
4
7
9
1
6
3
-
3
1
6
O
w
n
C
o
l
l
e
c
t
i
o
n
(
O
C
)
6
4
5
9
3
3
2
3
1
1
9
0
1
9
4
6
2.
2
.
F
e
a
t
uriza
t
io
n
Featu
r
es
i
n
m
ac
h
in
e
lear
n
i
n
g
is
b
asicall
y
n
u
m
er
ical
attr
ib
u
te
s
f
r
o
m
w
h
ich
a
n
y
o
n
e
ca
n
p
er
f
o
r
m
s
o
m
e
m
at
h
e
m
a
tical
o
p
er
atio
n
s
u
ch
a
s
m
atr
i
x
f
ac
t
o
r
izatio
n
,
d
o
t
p
r
o
d
u
ct
etc.
B
u
t
th
er
e
ar
e
v
ar
io
u
s
s
ce
n
ar
io
w
h
e
n
d
ataset
d
o
es
n
o
t
co
n
tai
n
n
u
m
er
ical
at
tr
ib
u
te
f
o
r
ex
a
m
p
le
-
s
e
n
ti
m
en
tal
an
al
y
s
i
s
o
f
T
w
it
ter
/Faceb
o
o
k
u
s
er
,
Am
az
o
n
cu
s
to
m
er
r
ev
ie
w
,
I
MD
B
/Net
f
li
x
m
o
v
ie
r
ec
o
m
m
e
n
d
atio
n
.
I
n
all
t
h
e
ab
o
v
e
ca
s
es
d
atase
t
co
n
tai
n
n
u
m
er
ical
v
al
u
e,
s
tr
i
n
g
v
a
lu
e
,
ch
ar
ac
ter
v
al
u
e,
ca
te
g
o
r
ical
v
al
u
e,
co
n
n
ec
tio
n
(
o
n
e
u
s
er
co
n
n
ec
ted
to
an
o
th
er
u
s
er
)
.
C
o
n
v
er
s
io
n
o
f
t
h
ese
t
y
p
es
o
f
f
ea
t
u
r
e
in
to
n
u
m
er
ica
l
f
ea
t
u
r
e
is
ca
lled
f
u
t
u
r
izatio
n
.
2
.
2
.
1
.
T
ex
t
pro
ce
s
s
i
ng
s
et
up
Fo
r
th
e
p
u
r
p
o
s
e
o
f
g
etti
n
g
ac
cu
r
ate
r
es
u
lts
b
y
c
lass
if
ier
s
we
h
a
v
e
to
m
a
k
e
s
u
r
e
t
h
at
th
e
s
e
d
atasets
p
r
o
ce
s
s
ed
ef
f
icie
n
tl
y
b
y
r
e
m
o
v
in
g
u
n
r
elate
d
co
n
te
n
ts
a
n
d
th
u
s
r
elate
d
co
n
ten
ts
ar
e
ac
cu
r
atel
y
e
x
tr
ac
ted
.
As
m
o
s
t
r
e
s
ea
r
ch
er
s
co
n
s
id
er
th
at
U
R
L
d
o
es
n
’
t
h
av
e
an
y
i
n
f
o
r
m
atio
n
r
eg
ar
d
i
n
g
s
e
n
ti
m
e
n
ts
,
s
o
b
y
r
e
m
o
v
i
n
g
s
h
o
r
t
U
R
L
s
f
r
o
m
t
w
ee
t
co
n
t
en
ts
ca
n
b
e
r
ef
i
n
ed
.
P
eo
p
le
o
f
ten
u
s
e
e
m
o
tio
n
al
w
o
r
d
s
t
h
at
co
n
ta
in
r
ep
ea
ted
letter
s
to
ex
p
r
es
s
t
h
eir
s
e
n
ti
m
e
n
ts
w
h
ic
h
ar
e
v
er
y
co
m
m
o
n
tr
en
d
s
li
k
e
“
co
o
o
o
o
l”.
A
ls
o
,
n
u
m
b
er
s
ar
e
n
o
t
u
s
ed
f
o
r
an
al
y
zi
n
g
s
en
ti
m
en
t
s
s
o
tw
ee
t
co
n
te
n
t
s
ca
n
b
e
r
ef
in
ed
b
y
r
e
m
o
v
i
n
g
t
h
e
m
[
1
]
.
T
h
e
p
o
lar
it
y
o
f
th
e
w
o
r
d
w
il
l
b
e
ch
an
g
ed
w
h
e
n
t
h
e
y
ar
e
p
r
ec
ed
ed
b
y
a
n
e
g
atio
n
o
r
n
eg
atio
n
ca
n
ch
a
n
g
e/r
ev
er
s
e
t
h
e
m
ea
n
i
n
g
o
f
w
o
r
d
s
.
B
y
ch
ec
k
i
n
g
n
e
g
atio
n
s
,
R
e
m
o
v
i
n
g
o
f
U
R
L
s
,
e
m
o
tio
n
s
,
n
u
m
b
er
s
an
d
R
ep
ea
ted
W
o
r
d
;
n
o
is
e
i
n
t
w
ee
ts
ca
n
b
e
r
ed
u
ce
d
.
T
h
is
f
ilter
p
r
o
v
id
es
u
s
o
p
tio
n
s
to
d
o
co
n
f
i
g
u
r
atio
n
w
it
h
o
u
r
d
atas
et
w
h
ich
i
n
clu
d
e
s
f
o
llo
w
in
g
s
tep
s
[
1
9
-
2
0
]
:
Ste
m
m
i
n
g
:
I
t
i
s
u
s
ed
to
r
e
m
o
v
e
s
u
f
f
i
x
f
r
o
m
th
e
w
o
r
d
ac
co
r
d
in
g
to
s
o
m
e
g
r
a
m
m
atica
l
r
u
le
s
.
Her
e
w
e
ap
p
l
y
m
o
s
t p
o
p
u
lar
Sn
o
w
b
all
Ste
m
m
i
n
g
lib
r
ar
y
.
Sto
p
W
o
r
d
E
x
tr
ac
to
r
:
So
m
e
w
o
r
d
s
t
h
at
d
o
n
’
t
h
a
v
e
p
o
lar
it
y
s
o
t
h
e
y
d
o
n
’
t
n
ee
d
to
b
e
f
u
r
th
er
an
al
y
ze
d
lik
e:
ab
le,
ar
e,
b
o
th
,
w
h
ic
h
,
h
as,
b
ec
o
m
e
,
a
f
ter
etc.
So
a
f
ter
eli
m
in
a
tio
n
o
f
th
e
s
e
w
o
r
d
s
,
o
u
r
r
esu
lt
w
i
ll
n
o
t b
e
af
f
ec
ted
.
W
e
u
s
ed
R
a
in
b
o
w
lis
t
f
o
r
o
u
r
ex
p
er
i
m
e
n
t.
T
o
k
en
izatio
n
:
I
t
i
s
u
s
ed
to
s
p
li
t
a
d
o
cu
m
en
t
in
to
a
w
o
r
d
o
r
ter
m
s
a
n
d
m
ak
e
a
w
o
r
d
v
ec
to
r
.
Her
e
w
e
u
s
ed
NGr
a
m
T
o
k
en
izer
.
Featu
r
e
Sele
ctio
n
:
T
h
is
p
r
o
ce
s
s
d
ec
r
ea
s
es
th
e
n
u
m
b
er
o
f
attr
ib
u
tes
i
n
to
a
b
etter
s
u
b
s
et
w
h
ic
h
ca
n
in
cr
ea
s
e
ac
cu
r
ac
y
also
it b
r
in
g
s
a
r
ed
u
ctio
n
in
tr
ai
n
i
n
g
ti
m
e.
I
t is d
o
n
e
b
y
u
s
i
n
g
Fil
ter
s
an
d
W
r
ap
p
e
r
s
.
2
.
3
.
Senti
m
ent
c
la
s
s
if
ier
T
o
class
if
y
s
e
n
ti
m
e
n
ts
m
ac
h
in
e
lear
n
i
n
g
(
ML
)
al
g
o
r
ith
m
s
ar
e
u
s
ed
i.e
.
a
b
r
an
ch
o
f
A
r
ti
f
icia
l
I
n
telli
g
en
ce
(
A
I
)
co
n
ce
r
n
ed
w
it
h
t
h
e
s
t
u
d
y
o
f
clas
s
i
f
icati
o
n
an
d
p
atter
n
a
n
al
y
s
is
,
allo
w
s
th
e
co
m
p
u
ter
to
lear
n
b
eh
a
v
io
r
s
o
f
e
m
p
ir
ical
d
ata
tak
e
n
f
r
o
m
s
en
s
o
r
s
o
r
d
atab
ase
[
2
1
]
.
ML
alg
o
r
it
h
m
allo
w
s
u
s
to
au
to
m
at
icall
y
r
ec
o
g
n
ize
co
m
p
lex
p
atter
n
s
a
n
d
m
a
k
e
in
tel
l
ig
en
t
d
ec
is
io
n
s
b
ased
o
n
d
ata.
I
n
th
is
p
ap
er
,
w
e
u
s
ed
v
ar
io
u
s
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
s
u
c
h
as
Naiv
e
B
a
y
es
(
NB
)
,
Su
p
p
o
r
t
Vec
to
r
M
ac
h
in
e
(
SVM)
[
2
2
]
,
an
d
Dec
is
io
n
T
r
ee
(
DT
)
[
1
5
]
.
2
.
3
.
1
.
Na
ïv
e
ba
y
es c
la
s
s
if
ier
I
t
r
ef
er
s
to
co
u
n
ti
n
g
th
e
f
r
e
q
u
en
c
y
o
f
w
o
r
d
s
t
h
at
ar
e
r
el
ated
to
th
e
s
e
n
ti
m
e
n
ts
in
th
e
m
es
s
ag
e
.
As
B
a
y
es
th
eo
r
e
m
b
ased
o
n
p
r
o
b
ab
ilis
tic
class
i
f
ier
s
o
it
all
o
w
s
u
s
to
ca
p
tu
r
e
u
n
ce
r
ta
in
t
y
ab
o
u
t
th
e
m
o
d
el
to
d
eter
m
in
e
t
h
e
p
r
o
b
ab
ilit
y
o
f
t
h
e
o
u
tco
m
e.
E
x
p
licit
p
r
o
b
ab
ilit
ies
ca
n
b
e
ca
lc
u
lated
b
y
it
f
o
r
th
e
test
ed
d
ataset
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
I
n
t J
A
r
ti
f
I
n
tell
,
Vo
l.
9
,
No
.
3
,
Sep
te
m
b
er
20
20
:
4
7
3
–
4
7
9
476
an
d
it
h
elp
s
to
r
ed
u
ce
n
o
is
e
r
o
b
u
s
tl
y
.
I
t
is
n
u
m
er
ic
al
b
ased
ap
p
r
o
ac
h
w
i
th
ea
s
y
,
f
a
s
t
an
d
h
i
g
h
ac
cu
r
ac
y
f
ea
tu
r
e
s
.
2
.
3
.
2
.
Su
pp
o
rt
v
ec
t
o
r
m
a
chi
ne
(
SVM
)
I
t
y
ield
s
m
o
r
e
ac
cu
r
ate
r
es
u
lt
s
w
h
e
n
it
i
s
u
s
ed
f
o
r
cla
s
s
i
f
y
in
g
te
x
t.
T
h
e
b
asic
id
ea
b
eh
i
n
d
i
t
is
to
f
in
d
th
e
h
y
p
er
p
lan
e
(
o
r
v
ec
to
r
w
)
,
w
h
ic
h
i
s
r
esp
o
n
s
ib
le
f
o
r
s
ep
ar
atin
g
o
n
e
clas
s
d
o
cu
m
e
n
t
v
ec
t
o
r
f
r
o
m
t
h
e
v
ec
to
r
in
o
th
er
class
[
7
]
.
I
t
is
s
u
cc
e
s
s
f
u
l
l
y
e
m
p
lo
y
ed
in
te
x
t
clas
s
if
ica
t
io
n
an
d
v
ar
io
u
s
o
th
er
s
eq
u
en
ce
p
r
o
ce
s
s
in
g
ap
p
licatio
n
s
as it i
s
a
t
y
p
e
o
f
li
n
ea
r
class
i
f
ier
.
2
.
3
.
3
.
Dec
is
io
n
t
re
e
(
DT
)
I
t
is
a
f
lo
w
c
h
ar
t
u
s
ed
to
o
u
t
p
u
t
lab
els
f
o
r
ce
r
tai
n
f
ea
t
u
r
es,
ac
t
as
i
n
p
u
t
v
al
u
es.
I
t
ca
t
eg
o
r
ies
a
d
o
cu
m
en
t
as
b
y
,
s
tar
ti
n
g
f
r
o
m
t
h
e
t
r
ee
r
o
o
t
(
lab
ele
d
as
f
ea
t
u
r
es),
f
o
llo
w
ed
d
o
w
n
w
ar
d
b
y
b
r
an
ch
e
s
(
lab
eled
as f
ea
tu
r
es
w
ei
g
h
t)
an
d
last
r
ea
ch
ed
a
leaf
n
o
d
e
(
lab
eled
b
y
ca
te
g
o
r
ies).
2
.
4
.
E
x
peri
m
ent
a
l
s
et
up
W
e
u
s
e
W
aik
ato
E
n
v
ir
o
n
m
en
t
f
o
r
K
n
o
w
led
g
e
An
a
l
y
s
i
s
(
W
E
KA
)
to
i
m
p
le
m
en
t
d
ata
m
i
n
in
g
al
g
o
r
it
h
m
s
f
o
r
p
r
ep
r
o
ce
s
s
in
g
,
c
lass
if
icatio
n
,
c
lu
s
ter
i
n
g
,
an
d
a
n
al
y
s
i
s
o
f
r
esu
lt
s
[
2
3
-
2
4
]
.
T
h
is
en
v
ir
o
n
m
en
t
in
cl
u
d
es
j
av
a
lib
r
ar
ies
t
h
at
i
m
p
le
m
e
n
t
a
lg
o
r
ith
m
s
a
n
d
p
r
o
v
id
e
th
e
b
e
s
t
en
v
ir
o
n
m
e
n
t
to
r
esear
ch
er
s
f
o
r
clas
s
i
f
y
in
g
d
a
tasets
.
W
e
ap
p
l
y
“
Stri
n
g
T
o
W
o
r
d
V
ec
to
r
”
f
ilter
an
d
d
o
n
e
lo
ts
o
f
p
r
ep
r
o
ce
s
s
in
g
w
it
h
o
u
r
d
atasets
[
2
5
-
2
6
]
.
Usi
n
g
n
-
g
r
a
m
to
k
en
izer
o
p
tio
n
a
n
d
attr
ib
u
te
s
e
lectio
n
m
et
h
o
d
d
if
f
er
e
n
t
n
u
m
b
er
o
f
attr
ib
u
tes
ar
e
cr
ea
ted
.
W
ith
attr
ib
u
tes
s
elec
tio
n
m
et
h
o
d
5
0
a
ttrib
u
tes
ar
e
tak
en
f
o
r
test
in
g
o
u
t
o
f
1
6
1
3
w
o
r
d
s
f
r
o
m
f
ir
s
t
d
ataset
SE
-
T
[
1
6
]
an
d
1
0
5
attr
ib
u
tes
o
u
t
o
f
2
0
6
5
w
o
r
d
s
ar
e
tak
en
f
r
o
m
s
ec
o
n
d
d
ataset
T
S
[
1
7
]
.
T
h
is
m
e
th
o
d
in
cr
ea
s
e
s
ac
cu
r
ac
y
r
ate
o
f
o
u
r
tr
ain
in
g
d
ata
s
et
also
it
b
r
in
g
s
a
r
ed
u
ctio
n
in
ex
ec
u
tio
n
t
i
m
e.
Fo
llo
w
i
n
g
T
ab
le
2
s
h
o
w
s
r
ed
u
ctio
n
in
s
ize
o
f
f
ile
af
ter
p
r
ep
r
o
ce
s
s
in
g
:
T
ab
le
2
.
Data
co
llectio
n
cr
iter
i
a
D
a
t
a
se
t
E
-
Tw
i
t
t
e
r
(
S
E
-
T)
[
1
5
]
Tw
i
t
t
e
r
S
a
n
d
e
r
s (T
S
)
[
1
6
]
O
w
n
C
o
l
l
e
c
t
i
o
n
(
O
C
)
S
i
z
e
o
f
f
i
l
e
b
e
f
o
r
e
p
r
e
p
r
o
c
e
ssi
n
g
1
.
7
M
B
9
3
.
7
K
B
1
.
5
M
B
S
i
z
e
o
f
f
i
l
e
a
f
t
e
r
p
r
e
p
r
o
c
e
ssi
n
g
(
f
e
a
t
u
r
e
se
l
e
c
t
i
o
n
)
1
8
1
K
B
9
.
9
K
B
1
5
0
K
B
T
o
ev
alu
ate
p
er
f
o
r
m
a
n
ce
w
e
ap
p
ly
1
0
-
f
o
ld
cr
o
s
s
v
a
lid
atio
n
tech
n
iq
u
e
w
h
ich
s
p
lit
s
t
h
e
o
r
ig
i
n
al
s
et
in
to
tr
ain
in
g
s
a
m
p
le
to
tr
ai
n
t
h
e
m
o
d
el
an
d
a
tes
t
s
et
to
e
v
al
u
ate
r
esu
lt
s
.
Fo
r
co
m
p
u
t
in
g
s
e
n
ti
m
e
n
ts
q
u
ic
k
l
y
o
f
t
w
ee
t
s
w
it
h
o
u
t
co
m
p
r
o
m
is
i
n
g
ac
cu
r
ac
y
,
an
ap
p
r
o
ac
h
k
n
o
wn
as
“
I
n
f
o
r
m
at
io
n
R
e
tr
iev
al
Me
tr
ics”
ca
n
b
e
u
s
ed
to
ev
alu
ate
e
x
p
er
i
m
e
n
tal
r
e
s
u
lts
in
ter
m
s
o
f
p
r
ec
is
io
n
,
r
ec
all,
f
-
m
ea
s
u
r
e,
a
n
d
ac
cu
r
a
c
y
w
it
h
t
h
e
u
s
e
o
f
f
o
llo
w
in
g
f
o
r
m
u
las [
9
,
2
7
]
:
P
r
ec
is
io
n
=
T
P
/ (
T
P
+
FP
)
(
1
)
R
ec
al
=
T
P
/
(
T
P
+F
P
)
(
2
)
F
-
m
ea
s
u
r
e
=
2
*
P
r
ec
is
io
n
*
R
e
ca
ll/ (
P
r
ec
is
io
n
+
R
ec
all)
(
3
)
A
cc
u
r
ac
y
=
T
P
+
T
N/
(
T
P
+
T
N
+
FP
+
FN)
(
4
)
Her
e
(
T
P
=
T
r
u
e
P
o
s
itiv
e;
T
N
=
T
r
u
e
Neg
ati
v
e;
FP
=False P
o
s
iti
v
e;
FN=
Fal
s
e
Ne
g
ativ
e)
3.
RE
SU
L
T
S
A
ND
AN
AL
Y
SI
S
W
e
o
b
s
er
v
ed
th
at
o
u
r
clas
s
if
ic
atio
n
r
es
u
lts
i
m
p
r
o
v
ed
in
ter
m
s
o
f
ti
m
e
an
d
ac
c
u
r
ac
y
u
s
i
n
g
p
r
o
ce
s
s
ed
an
d
s
m
all
f
ea
t
u
r
es
d
ata
th
a
n
s
i
m
p
le
d
ataset
s
.
Fo
r
e
x
a
m
p
le
i
n
f
ir
s
t
SE
-
T
d
ataset,
ti
m
e
tak
e
n
to
b
u
ild
a
m
o
d
el
f
o
r
NB
alg
o
r
ith
m
tak
e
s
1
0
.
5
6
s
ec
o
n
d
s
,
ac
c
u
r
ac
y
5
3
.
7
3
%
an
d
af
ter
p
r
o
ce
s
s
i
n
g
ti
m
e
ta
k
e
n
to
tes
t
m
o
d
el
o
n
tr
ain
i
n
g
d
ata
i
s
r
ed
u
ce
d
at
0
.
3
5
s
ec
o
n
d
s
o
n
l
y
,
ac
c
u
r
ac
y
i
m
p
r
o
v
ed
b
y
5
7
.
4
6
%.
T
a
b
le
3
d
em
o
n
s
tr
ates
th
e
ac
cu
r
ac
y
o
f
cla
s
s
i
f
ier
s
o
n
th
r
e
e
d
atasets
af
ter
ap
p
l
y
i
n
g
v
ar
io
u
s
p
r
ep
r
o
ce
s
s
m
et
h
o
d
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
A
r
ti
f
I
n
tell
I
SS
N:
2252
-
8938
Ma
ch
in
e
lea
r
n
in
g
-
b
a
s
ed
te
ch
n
iq
u
e
fo
r
b
ig
d
a
ta
s
en
timen
ts
…
(
N
o
r
a
in
i S
ema
n
)
477
T
ab
le
3
.
A
cc
u
r
ac
y
cr
iter
ia
f
o
r
d
atasets
Ev
a
l
u
a
t
i
o
n
C
r
i
t
e
r
i
a
D
a
t
a
se
t
S
V
M
DT
NB
C
o
r
r
e
c
t
l
y
c
l
a
ssi
f
i
e
d
i
n
s
t
a
n
c
e
s
SE
-
T
8
3
3
5
8
1
1
8
7
7
7
6
TS
4
1
9
4
1
2
3
5
5
OC
5
3
3
4
5
0
3
5
5
3
2
4
I
n
c
o
r
r
e
c
t
l
y
c
l
a
ssi
f
i
e
d
i
n
st
a
n
c
e
s
SE
-
T
5
2
0
6
5
4
2
3
5
7
6
5
TS
60
67
1
2
4
OC
1
1
2
5
1
4
2
4
1
1
3
5
A
c
c
u
r
a
c
y
(
%)
SE
-
T
6
1
.
5
5
5
9
.
9
5
5
7
.
4
2
TS
8
7
.
4
7
8
6
.
0
1
7
4
.
1
1
OC
8
2
.
5
8
7
7
.
9
5
8
2
.
4
3
Er
r
o
r
SE
-
T
0
.
3
8
0
.
4
0
0
.
4
2
TS
0
.
1
3
0
.
1
4
0
.
2
6
OC
0
.
1
7
0
.
2
2
0
.
1
8
Fo
llo
w
i
n
g
p
er
f
o
r
m
an
ce
m
ea
s
u
r
es
ar
e
r
ep
o
r
ted
in
T
a
b
le
4
b
y
o
u
r
ex
p
er
i
m
en
tal
r
esu
l
t
u
s
in
g
th
r
e
e
d
ataset,
af
ter
co
n
d
u
cti
n
g
1
0
-
f
o
ld
cr
o
s
s
v
alid
atio
n
tec
h
n
iq
u
e.
T
ab
le
4
.
P
er
f
o
r
m
a
n
ce
m
ea
s
u
r
e
s
o
f
clas
s
if
ier
s
C
l
a
ssi
f
i
e
r
D
a
t
a
se
t
T
P
R
a
t
e
F
P
R
a
t
e
P
r
e
c
i
si
o
n
R
e
c
a
l
l
F
-
M
e
a
su
r
e
P
o
l
a
r
i
t
y
S
V
M
SE
-
T
0
.
3
5
2
0
.
0
3
6
0
.
8
5
9
0
.
3
5
2
0
.
5
0
0
p
o
si
t
i
v
e
TS
0
.
9
0
6
0
.
5
6
2
0
.
5
8
0
0
.
9
0
6
0
.
7
0
7
n
e
u
t
r
a
l
OC
0
.
6
4
4
0
.
0
0
6
0
.
9
8
1
0
.
6
4
4
0
.
7
7
8
p
o
si
t
i
v
e
DT
SE
-
T
0
.
3
4
0
.
0
4
8
0
.
8
2
1
0
.
3
4
9
0
.
4
8
9
p
o
si
t
i
v
e
TS
0
.
2
8
4
0
.
0
5
4
0
.
4
8
6
0
.
2
8
4
0
.
3
5
9
n
e
g
a
t
i
v
e
OC
0
.
9
9
4
0
.
3
9
9
0
.
8
2
8
0
.
9
9
4
0
.
9
0
4
n
e
g
a
t
i
v
e
NB
SE
-
T
0
.
7
8
0
0
.
4
7
9
0
.
5
8
2
0
.
7
8
0
0
.
6
6
7
n
e
u
t
r
a
l
TS
0
.
2
5
2
0
.
0
0
6
0
.
9
5
3
0
.
2
5
2
0
.
3
9
8
p
o
si
t
i
v
e
OC
0
.
4
3
7
0
.
1
8
6
0
.
5
9
7
0
.
4
3
7
0
.
5
0
5
p
o
si
t
i
v
e
T
h
e
n
u
m
b
er
o
f
co
r
r
ec
tl
y
cla
s
s
if
ied
i
n
s
ta
n
ce
s
an
d
ac
cu
r
ac
y
r
ate
is
g
r
ea
ter
f
o
r
th
r
ee
d
at
asets
w
it
h
SVM
alg
o
r
it
h
m
.
I
n
o
u
r
ex
p
er
i
m
en
t
o
b
tain
ed
ac
cu
r
ac
y
u
s
i
n
g
SVM
alg
o
r
ith
m
is
6
1
.
5
5
%,
8
7
.
4
7
%
an
d
8
2
.
5
8
%
r
esp
ec
tiv
el
y
(
w
it
h
5
0
f
ea
t
u
r
e
SE
-
T
,
1
0
5
f
ea
tu
r
e
T
S
d
ataset
s
an
d
1
0
0
f
ea
tu
r
e
o
f
O
C
d
atase
ts
)
w
h
ich
i
s
g
r
ea
ter
th
an
o
th
er
t
w
o
alg
o
r
it
h
m
s
.
O
u
r
ex
p
er
i
m
en
tal
r
es
u
lt
s
h
o
w
s
th
at
s
a
m
e
p
r
ep
r
o
ce
s
s
in
g
m
et
h
o
d
s
o
n
a
d
i
f
f
er
e
n
t
d
ataset
af
f
ec
t
s
i
m
ilar
l
y
t
h
e
c
la
s
s
i
f
ier
s
p
er
f
o
r
m
an
ce
.
Af
ter
an
al
y
zi
n
g
r
esu
lts
o
f
T
ab
le
4
it i
s
o
b
s
er
v
ed
th
at
SVM
p
r
o
v
id
es
6
4
.
9
6
%,
7
1
.
2
6
%
a
n
d
9
1
.
2
5
%
o
v
er
all
p
r
ec
is
io
n
w
h
ic
h
is
b
etter
t
h
an
o
th
er
t
w
o
alg
o
r
it
h
m
s
.
A
l
s
o
,
o
v
er
all
R
ec
all
a
n
d
F
-
m
e
asu
r
e
r
ate
o
f
SVM
is
g
r
ea
ter
t
h
an
NB
an
d
DT
f
o
r
t
h
r
ee
d
atasets
.
F
u
r
t
h
er
m
o
r
e,
ti
m
e
tak
e
n
to
b
u
i
ld
a
m
o
d
el
i
s
g
r
ea
tl
y
r
e
d
u
ce
d
b
y
ap
p
l
y
i
n
g
f
ea
t
u
r
e
s
e
lectio
n
m
e
th
o
d
.
T
i
m
e
tak
e
n
to
b
u
ild
m
o
d
el
i
n
f
ir
s
t
SE
-
T
d
atasets
is
0
.
4
5
,
2
9
.
4
3
,
4
.
4
7
s
ec
o
n
d
s
r
esp
ec
tiv
el
y
w
i
th
NB
,
SVM,
an
d
DT
alg
o
r
ith
m
;
in
s
ec
o
n
d
T
S d
ataset,
it is
0
.
0
1
,
0
.
0
6
,
0
.
0
1
s
ec
o
n
d
s
w
it
h
NB
,
SVM
an
d
DT
alg
o
r
ith
m
s
r
esp
ec
tiv
el
y
.
4.
CO
NCLU
SI
O
N
I
n
th
i
s
p
ap
er
,
w
e
d
is
c
u
s
s
s
e
n
ti
m
en
t
an
al
y
s
is
w
h
ic
h
ca
n
t
ell
u
s
t
h
e
t
h
o
u
g
h
t
o
f
w
r
iter
s
ab
o
u
t
th
e
p
ar
ticu
lar
en
tit
y
.
T
h
ese
d
ay
s
,
it
b
ec
o
m
e
s
a
r
o
u
tin
e
tas
k
to
f
in
d
p
eo
p
le
s
en
ti
m
e
n
t
s
ab
o
u
t
a
r
ea
l
w
o
r
ld
en
tit
y
f
r
o
m
s
o
cial
m
ed
ia
s
ite
s
li
k
e
T
w
it
ter
,
f
ac
e
b
o
o
k
o
r
b
lo
g
s
etc.
T
o
ef
f
icie
n
tl
y
an
a
l
y
ze
t
h
is
lar
g
e
a
m
o
u
n
t
o
f
d
atasets
i
t
i
s
es
s
en
tial
to
ac
cu
r
atel
y
c
lass
if
y
it.
I
n
t
h
is
p
ap
er
,
w
e
h
a
v
e
p
r
ese
n
ted
a
m
et
h
o
d
o
lo
g
y
o
f
te
x
t
m
in
i
n
g
u
s
i
n
g
W
ek
a
to
o
l
f
o
r
class
if
y
i
n
g
s
e
n
ti
m
e
n
ts
o
f
t
w
itter
.
W
e
u
s
e
t
h
r
ee
m
ac
h
i
n
e
lear
n
i
n
g
al
g
o
r
ith
m
s
SVM,
DT
,
an
d
NB
f
o
r
class
if
y
i
n
g
s
e
n
ti
m
en
ts
o
f
t
w
it
ter
s
d
ata.
W
e
co
n
d
u
ct
an
e
x
p
er
i
m
e
n
t
o
n
t
h
r
ee
t
w
itter
’
s
d
ataset
s
to
v
er
if
y
t
h
e
e
f
f
ec
ti
v
e
n
ess
o
f
p
r
e
-
p
r
o
ce
s
s
in
g
.
O
u
r
e
x
p
er
i
m
e
n
tal
r
es
u
lt
s
i
n
d
icate
t
h
at
b
y
r
e
m
o
v
i
n
g
u
n
w
a
n
ted
wo
r
d
s
an
d
s
e
lecti
n
g
f
ea
t
u
r
es
i
n
th
e
p
r
eli
m
i
n
ar
y
p
h
ase
o
f
p
r
ep
r
o
ce
s
s
in
g
,
t
i
m
e
to
b
u
ild
m
o
d
el
is
r
ed
u
ce
d
an
d
also
it
p
r
o
v
id
es
m
o
r
e
ac
c
u
r
at
e
r
esu
lt
s
i
n
ap
p
lied
alg
o
r
ith
m
s
.
T
h
e
r
esu
lt
m
a
y
b
e
a
f
f
ec
te
d
b
y
t
h
e
c
h
o
ice
o
f
f
ea
t
u
r
es
f
o
r
tr
ain
i
n
g
a
n
d
ch
o
ic
e
o
f
alg
o
r
ith
m
f
o
r
s
en
ti
m
en
t
c
lass
i
f
icatio
n
.
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
SVM,
DT
,
an
d
NB
alg
o
r
ith
m
s
i
m
p
r
o
v
e
o
n
d
a
tasets
af
ter
r
e
m
o
v
in
g
u
n
w
a
n
te
d
w
o
r
d
s
.
T
h
er
ef
o
r
e,
r
e
m
o
v
i
n
g
u
n
w
a
n
ted
w
o
r
d
s
i
s
u
s
e
f
u
l
to
i
m
p
r
o
v
e
t
h
e
p
er
f
o
r
m
an
ce
o
f
s
en
t
i
m
e
n
t
cla
s
s
i
f
icat
io
n
.
W
e
d
is
cu
s
s
t
h
e
co
m
p
ar
ati
v
e
an
al
y
s
i
s
o
f
t
h
r
ee
alg
o
r
ith
m
s
an
d
ca
lc
u
late
o
v
e
r
all
p
er
f
o
r
m
an
ce
m
ea
s
u
r
e
s
i
n
ter
m
s
o
f
p
r
ec
is
io
n
,
r
ec
all,
a
n
d
f
-
m
ea
s
u
r
e.
O
u
r
ex
p
er
i
m
e
n
tal
r
es
u
lt
s
i
n
d
icate
t
h
at
S
VM
p
r
o
v
id
es
m
o
r
e
ac
c
u
r
ate
r
esu
lt
s
t
h
an
o
t
h
er
al
g
o
r
ith
m
s
.
Ho
w
e
v
er
,
it
i
s
i
m
p
o
r
tan
t
to
f
u
r
t
h
er
s
t
u
d
y
c
u
r
r
en
t
a
v
ailab
le
p
r
ep
r
o
ce
s
s
in
g
tec
h
n
iq
u
es
t
h
at
h
elp
u
s
to
i
m
p
r
o
v
e
r
esu
lts
o
f
v
ar
io
u
s
clas
s
i
f
ier
s
.
A
m
eth
o
d
s
h
o
u
ld
b
e
f
o
u
n
d
to
a
u
to
m
a
ti
c
in
co
r
p
o
r
ate
f
ea
t
u
r
e
s
elec
tio
n
at
ti
m
e
o
f
m
o
d
el
b
u
ild
in
g
ac
co
r
d
in
g
to
an
y
la
n
g
u
ag
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
I
n
t J
A
r
ti
f
I
n
tell
,
Vo
l.
9
,
No
.
3
,
Sep
te
m
b
er
20
20
:
4
7
3
–
4
7
9
478
ACK
NO
WL
E
D
G
E
M
E
NT
S
T
h
e
au
th
o
r
s
w
o
u
ld
lik
e
to
th
a
n
k
U
n
i
v
er
s
iti
T
ek
n
o
lo
g
i
M
AR
A
(
UiT
M)
f
o
r
th
e
r
esear
ch
f
u
n
d
i
n
g
an
d
s
u
p
p
o
r
t v
ia
g
r
a
n
t n
u
m
b
er
6
0
0
-
I
R
MI
/P
E
R
D
A
N
A
5
/3
B
E
ST
A
R
I
(
1
0
7
/2
0
1
8
)
.
RE
F
E
R
E
NC
E
S
[1
]
J.
Zh
a
o
,
G
.
X
ia
o
li
n
,
“
Co
m
p
a
riso
n
Re
se
a
rc
h
o
n
T
e
x
t
P
re
-
p
ro
c
e
s
sin
g
M
e
th
o
d
s
o
n
T
w
it
ter
S
e
n
ti
m
e
n
t
A
n
a
l
y
sis,”
IEE
E
Acc
e
ss
,
DO
I
1
0
.
1
1
0
9
/A
CC
ES
S
.
2
0
1
7
,
2
6
7
2
6
7
7
.
[2
]
A
.
Kro
u
sk
a
e
t
a
l.
,
“
T
h
e
e
ff
e
c
t
o
f
p
re
p
ro
c
e
ss
in
g
tec
h
n
iq
u
e
s
o
n
T
w
it
ter
S
A
,
”
in
7
t
h
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
In
fo
rm
a
t
io
n
,
I
n
telli
g
e
n
c
e
,
S
y
ste
ms
&
Ap
p
li
c
a
ti
o
n
s (
IIS
A)
,
Re
se
a
rc
h
G
a
te Co
n
f
e
r
e
n
c
e
:
2
0
1
6
.
[3
]
B
.
D.
S
a
v
it
a
,
e
t
a
l,
“
S
e
n
ti
m
e
n
t
An
a
ly
sis
o
n
Tw
it
ter
Da
ta
Us
in
g
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
i
n
e
,
”
in
I
n
ter
n
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
Co
m
p
u
ter
S
c
ien
c
e
T
re
n
d
s
a
n
d
T
e
c
h
n
o
l
o
g
y
(
IJ
CS
T
)
,
v
o
l
.
4
(3
),
2
0
1
6
.
[4
]
He
m
a
lath
a
,
e
t
a
l
,
“
S
e
n
ti
m
e
n
t
An
a
ly
sis
T
o
o
l
u
sin
g
M
a
c
h
in
e
L
e
a
rn
in
g
A
lg
o
rit
h
m
s
,
”
in
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Eme
rg
in
g
T
re
n
d
s &
T
e
c
h
n
o
lo
g
y
i
n
Co
m
p
u
ter
S
c
ie
n
c
e
(
IJ
ET
T
CS
)
,
v
o
l.
2
(2
),
2
0
1
3
.
[5
]
G
.
V
in
o
d
h
i
n
i,
e
t
a
l
,
“
S
e
n
ti
m
e
n
t
A
n
a
l
y
si
s
a
n
d
Op
i
n
io
n
M
in
i
n
g
:
A
S
u
rv
e
y
,
”
in
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Ad
v
a
n
c
e
d
Res
e
a
rc
h
in
C
o
mp
u
ter
S
c
ien
c
e
a
n
d
S
o
ft
w
a
re
En
g
in
e
e
rin
g
,
v
o
l
2
(6
)
,
2
0
1
2
.
[6
]
A
.
P
a
k
e
t
a
l,
“
Tw
it
ter
a
s
a
C
o
rp
u
s
f
o
r
S
e
n
ti
m
e
n
t
A
n
a
l
y
sis
a
n
d
Op
i
n
io
n
M
in
i
n
g
,
”
Un
iv
e
rsite
d
e
P
a
ris
-
S
u
d
,
L
a
b
o
ra
to
ire L
IM
S
I
-
CNRS,
F
RA
NCE,
p
p
.
1
3
2
0
-
1
3
2
6
.
[7
]
M
.
Ra
n
i
e
t
a
l,
“
A
Re
v
i
e
w
o
f
D
a
t
a
A
n
a
l
y
sis
o
f
Tw
it
ter,”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Ad
v
a
n
c
e
d
Res
e
a
rc
h
in
C
o
mp
u
ter
S
c
ien
c
e
a
n
d
S
o
ft
w
a
re
En
g
in
e
e
rin
g
,
v
o
l.
6
(
5
),
2
0
1
6
.
[8
]
L
.
Ba
rb
o
sa
,
e
t
a
l
,
“
Ro
b
u
st
S
e
n
ti
m
e
n
t
De
tec
ti
o
n
o
n
Tw
it
ter
f
ro
m
Bias
e
d
a
n
d
No
isy
Da
ta,”
Co
li
n
g
2
0
1
0
:
P
o
ste
r,
p
p
.
36
-
4
4
,
Be
ij
in
g
,
2
0
1
0
[9
]
B.
P
a
n
g
,
e
t
a
l
,
“
T
h
u
m
b
s
u
p
?
S
e
n
ti
m
e
n
t
Clas
si
f
ic
a
ti
o
n
u
sin
g
M
a
c
h
in
e
L
e
a
rn
in
g
T
e
c
h
n
iq
u
e
s,”
Pro
c
e
e
d
in
g
s
o
f
EM
NL
P
2
0
0
2
,
p
p
.
7
9
-
86
,
2
0
0
2
.
[1
0
]
H.
A
n
b
e
r,
e
t
a
l
,
“
A
L
it
e
ra
tu
re
Re
v
ie
w
o
n
T
w
it
ter
Da
ta
A
n
a
l
y
sis
”
,
In
ter
n
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
Co
mp
u
ter
a
n
d
E
lec
trica
l
En
g
i
n
e
e
rin
g
,
v
o
l.
5
(3
)
,
p
p
:
5
3
-
6
0
,
2
0
1
6
.
[1
1
]
S
.
A
.
M
u
la
y
,
e
t
a
l
,
“
S
e
n
ti
m
e
n
t
A
n
a
l
y
si
s
a
n
d
Op
in
io
n
M
in
i
n
g
Wi
th
S
o
c
ial
Ne
tw
o
rk
in
g
f
o
r
P
re
d
ictin
g
Bo
x
O
ff
ice
Co
ll
e
c
ti
o
n
o
f
M
o
v
ie,”
in
In
ter
n
a
t
io
n
a
l
J
o
u
r
n
a
l
o
f
Eme
rg
in
g
Res
e
a
rc
h
in
M
a
n
a
g
e
me
n
t
&
T
e
c
h
n
o
lo
g
y
,
v
o
l.
5
(1
),
p
p
.
227
-
2
3
5
.
[1
2
]
Y.
Ye
n
g
i,
e
t
a
l
.
,
“
Distrib
u
ted
Re
c
o
m
m
e
n
d
e
r
S
y
ste
m
s
w
it
h
S
e
n
ti
m
e
n
t
A
n
a
l
y
sis”
,
Eu
ro
p
e
a
n
J
o
u
rn
a
l
o
f
S
c
ien
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
v
o
l.
4
(7
),
p
p
.
5
1
-
57.
[1
3
]
S
.
W
a
k
a
d
e
,
e
t
a
l
,
“
T
e
x
t
M
in
in
g
f
o
r
S
e
n
ti
m
e
n
t
A
n
a
l
y
sis
o
f
Tw
it
ter
Da
ta”
,
T
h
e
Un
iv
e
rsity
o
f
Ak
ro
n
,
De
p
a
rtm
e
n
t
o
f
Co
m
p
u
ter S
c
ien
c
e
.
[1
4
]
R.
Niv
e
d
h
a
a
n
d
N.
S
a
iram
,
“
A
M
a
c
h
in
e
L
e
a
rn
in
g
b
a
se
d
Clas
si
f
ica
t
io
n
o
r
S
o
c
ial
M
e
d
ia
M
e
ss
a
g
e
s
”
,
In
d
i
a
n
J
o
u
rn
a
l
o
f
S
c
ien
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
v
o
l
8
(
1
6
),
p
p
.
1
0
2
-
1
1
0
.
[1
5
]
H.
S
h
a
m
su
d
in
,
e
t
a
l
,
“
H
y
b
rid
is
a
ti
o
n
o
f
RF
(X
g
b
)
T
o
Im
p
ro
v
e
T
h
e
T
re
e
-
b
a
s
e
d
A
lg
o
rit
h
m
s
in
L
e
a
rn
in
g
S
t
y
l
e
P
re
d
ictio
n
,
”
i
n
IAE
S
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Art
if
icia
l
In
tell
ig
e
n
c
e
(
IJ
-
AI
),
v
o
l.
8
(4
)
,
p
p
.
4
2
2
-
4
2
8
,
2
0
1
9
.
[1
6
]
W
.
S
id
o
re
n
k
o
,
“
S
e
m
Ev
a
l
-
2
0
1
6
T
a
sk
4
:
S
e
n
ti
m
e
n
t
A
n
a
l
y
sis
o
n
T
w
it
ter,
T
ra
in
in
g
+
De
v
d
a
tas
e
t,
”
h
tt
p
s://
g
it
h
u
b
.
c
o
m/W
la
d
imirS
id
o
re
n
k
o
/S
e
mEv
a
l
-
2
0
1
6
.
[1
7
]
S
.
S
a
n
d
e
rs,
“
S
a
n
d
e
rs A
n
a
l
y
ti
c
s t
w
it
ter se
n
ti
m
e
n
t
c
o
rp
u
s,”
h
tt
p
s:/
/
g
it
h
u
b
.
c
o
m/g
u
y
z/twi
tt
e
r
-
se
n
ti
me
n
t
d
a
t
a
se
t
.
[1
8
]
M
.
M
.
A
lt
a
w
a
ier
a
n
d
S
.
T
iu
n
,
“
Co
m
p
a
riso
n
o
f
M
a
c
h
in
e
L
e
a
rn
in
g
A
p
p
ro
a
c
h
e
s
o
n
A
ra
b
ic
Tw
i
tt
e
r
S
e
n
ti
m
e
n
t
A
n
a
l
y
si
s,”
v
o
l.
6
(6
)
,
p
p
.
1
0
6
7
-
1
0
7
3
,
2
0
1
6
.
[1
9
]
S
.
Zain
u
d
in
,
D.
S
.
Ja
sim
,
a
n
d
A
.
A
.
Ba
k
a
r,
“
Co
m
p
a
ra
ti
v
e
A
n
a
ly
sis
o
f
Da
ta
M
in
in
g
T
e
c
h
n
iq
u
e
s
f
o
r
M
a
lay
sia
n
Ra
in
f
a
ll
P
re
d
icti
o
n
,
”
In
t.
J
.
A
d
v
.
S
c
i.
En
g
.
I
n
f.
T
e
c
h
n
o
l
.
,
v
o
l
.
6
,
n
o
.
6
,
p
p
.
1
1
4
8
-
1
1
5
3
,
2
0
1
6
.
[2
0
]
M
.
S
.
Ne
e
th
u
,
a
n
d
R
.
Ra
jas
re
e
,
"
S
e
n
ti
m
e
n
t
a
n
a
l
y
sis
in
t
w
it
ter
u
sin
g
m
a
c
h
in
e
lea
rn
in
g
tec
h
n
i
q
u
e
s
.
"
Co
mp
u
ti
n
g
,
Co
mm
u
n
ica
ti
o
n
s a
n
d
Ne
two
rk
in
g
T
e
c
h
n
o
lo
g
ies
(
ICCCNT
),
2
0
1
3
F
o
u
rth
In
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
.
IEE
E,
2
0
1
3
.
[2
1
]
M
.
A
.
A
l
-
Ha
g
e
r
y
,
"
Ex
trac
ti
n
g
Hi
d
d
e
n
P
a
tt
e
r
n
s
F
ro
m
Da
tes
’
P
ro
d
u
c
t
Da
ta
Us
in
g
A
M
a
c
h
in
e
Lea
rn
in
g
T
e
c
h
n
iq
u
e
,
"
in
IAE
S
I
n
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
o
f
Arti
fi
c
ia
l
I
n
telli
g
e
n
c
e
(
IJ
-
AI
),
v
o
l
.
8
(3
)
,
p
p
.
2
0
5
-
2
1
4
,
2
0
1
9
.
[2
2
]
S
.
Ib
ra
h
im
,
e
t
a
l
,
"
Rice
G
ra
in
Clas
sif
ic
a
ti
o
n
Us
in
g
M
u
l
ti
-
Cla
ss
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
in
e
(S
V
M
),
"
in
IAE
S
In
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
o
f
Art
if
icia
l
In
telli
g
e
n
c
e
(
IJ
-
AI
),
v
o
l.
8
(
3
),
p
p
.
2
1
5
-
2
2
0
,
2
0
1
9
.
[2
3
]
M
.
Ha
ll
,
E.
F
ra
n
k
,
G
.
Ho
l
m
e
s,
B.
P
f
a
h
rin
g
e
r,
P
.
Re
u
tem
a
n
n
,
a
n
d
I.
H.
W
it
ten
,
“
T
h
e
W
EKA
Da
ta
M
i
n
in
g
S
o
ft
w
a
re
:
An
Up
d
a
te;
S
IGKD
D E
x
p
l
o
ra
ti
o
n
s
,
”
v
o
l.
1
1
(1
),
2
0
0
9
.
[2
4
]
R.
A
ro
r
a
a
n
d
S
.
S
u
m
a
n
,
“
Co
m
p
a
ra
ti
v
e
A
n
a
l
y
sis
o
f
Clas
si
f
ica
ti
o
n
A
l
g
o
rit
h
m
s
o
n
Dif
f
e
re
n
t
Da
tas
e
ts
u
sin
g
W
EKA
,
”
In
ter
n
a
t
io
n
a
l
J
o
u
rn
a
ls C
o
mp
u
ter
Ap
p
li
c
a
ti
o
n
,
v
o
l.
5
4
(
1
3
),
p
p
.
2
1
-
2
5
,
2
0
1
2
.
[2
5
]
N.
M
a
ll
io
s,
E.
P
a
p
a
g
e
o
rg
io
u
,
M
.
S
a
m
a
rin
a
s,
a
n
d
K.
S
k
riap
a
s,
“
Co
m
p
a
riso
n
o
f
m
a
c
h
in
e
lea
rn
in
g
t
e
c
h
n
iq
u
e
s
u
sin
g
th
e
W
EK
A
e
n
v
iro
n
m
e
n
t
f
o
r
p
ro
sta
te
c
a
n
c
e
r
th
e
ra
p
y
p
lan
,
”
in
Pro
c
e
e
d
in
g
s
o
f
th
e
2
0
1
1
2
0
th
IEE
E
In
ter
n
a
t
io
n
a
l
W
o
rk
sh
o
p
s
o
n
E
n
a
b
li
n
g
T
e
c
h
n
o
l
o
g
ies
:
In
fra
stru
c
tu
re
f
o
r
Co
l
la
b
o
ra
ti
v
e
En
ter
p
rise
s,
W
ET
ICE
2
0
1
1
,
p
p
.
1
5
1
-
1
5
5
,
2
0
1
1
.
[2
6
]
T
.
Ga
rg
a
n
d
S
.
S
.
Kh
u
ra
n
a
,
“
Co
m
p
a
riso
n
o
f
c
las
sif
ica
ti
o
n
tec
h
n
i
q
u
e
s
f
o
r
in
tru
si
o
n
d
e
tec
ti
o
n
d
a
tas
e
t
u
sin
g
W
EKA
,
”
In
t.
C
o
n
f
.
Rec
e
n
t
A
d
v
.
In
n
o
v
.
En
g
.
ICRA
IE
2
0
1
4
,
2
0
1
4
.
[2
7
]
B.
M
.
P
a
ti
l,
D
.
T
o
sh
n
iw
a
l,
a
n
d
R.
C.
Jo
sh
i,
“
P
re
d
ictin
g
b
u
rn
p
a
t
ien
t
su
rv
iv
a
b
il
it
y
u
sin
g
d
e
c
is
io
n
tree
in
W
EK
A
e
n
v
iro
n
m
e
n
t,
”
2
0
0
9
I
EE
E
I
n
t.
Ad
v
.
Co
mp
u
t.
C
o
n
f
.
IACC
2
0
0
9
,
M
a
rc
h
,
p
p
.
1
3
5
3
-
1
3
5
6
,
2
0
0
9
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
A
r
ti
f
I
n
tell
I
SS
N:
2252
-
8938
Ma
ch
in
e
lea
r
n
in
g
-
b
a
s
ed
te
ch
n
iq
u
e
fo
r
b
ig
d
a
ta
s
en
timen
ts
…
(
N
o
r
a
in
i S
ema
n
)
479
B
I
O
G
RAP
H
I
E
S
O
F
AUTH
O
RS
No
ra
in
i
S
e
m
a
n
re
c
e
iv
e
d
th
e
B
a
c
h
e
lo
r
d
e
g
re
e
in
Co
m
p
u
ter
S
c
ien
c
e
f
ro
m
Un
iv
e
rsiti
P
u
tra
M
a
la
y
sia
(UP
M
),
M
a
lay
sia
in
1
9
9
9
;
t
h
e
M
S
c
d
e
g
re
e
f
ro
m
Qu
e
e
n
sl
a
n
d
Un
iv
e
rsity
o
f
Tec
h
n
o
lo
g
y
(QU
T
),
A
u
stra
li
a
(2
0
0
2
),
a
n
d
t
h
e
P
h
.
D
.
d
e
g
re
e
f
ro
m
th
e
Un
iv
e
rsiti
T
e
k
n
o
lo
g
i
M
A
RA
(Ui
T
M
),
M
a
la
y
sia
(2
0
1
2
)
.
S
h
e
is
p
re
se
n
tl
y
H
e
a
d
o
f
Ac
a
d
e
m
ic
P
ro
g
ra
m
Ac
c
re
d
it
a
ti
o
n
u
n
d
e
r
Cu
rr
icu
lu
m
Aff
a
irs
Un
it
a
n
d
se
n
io
r
lec
tu
re
r
a
t
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
,
F
a
c
u
lt
y
o
f
Co
m
p
u
ter
a
n
d
M
a
th
e
m
a
ti
c
a
l
S
c
ien
c
e
s.
He
r
r
e
se
a
rc
h
in
tere
sts
in
c
lu
d
e
A
I
a
p
p
li
c
a
ti
o
n
t
o
d
ig
it
a
l
sig
n
a
l
p
ro
c
e
ss
in
g
p
ro
b
lem
s,
sp
e
e
c
h
su
m
m
a
riz
a
ti
o
n
,
m
a
c
h
in
e
tran
sl
a
ti
o
n
a
n
d
m
a
c
h
in
e
lea
rn
in
g
tec
h
n
iq
u
e
s in
sp
e
e
c
h
re
c
o
g
n
it
io
n
tec
h
n
o
lo
g
y
a
n
d
tex
t
m
in
in
g
w
it
h
v
a
st
tec
h
n
iq
u
e
s/a
p
p
ro
a
c
h
e
s
o
f
Na
tu
ra
l
L
a
n
g
u
a
g
e
P
r
o
c
e
ss
in
g
(NL
P
).
Nu
ru
l
A
ti
q
a
h
Bin
ti
Ra
z
m
i
o
b
tain
e
d
h
e
r
Ba
c
h
e
lo
r
o
f
En
g
in
e
e
rin
g
(M
e
c
h
a
tro
n
ics
)
in
2
0
1
2
f
ro
m
M
o
n
a
sh
U
n
iv
e
rsity
M
a
la
y
sia
.
S
h
e
h
a
s
b
e
e
n
w
o
rk
in
g
in
M
a
n
u
f
a
c
tu
rin
g
Qu
a
li
ty
in
ra
re
-
e
a
rth
p
ro
d
u
c
ts
in
d
u
stry
f
o
r
5
y
e
a
rs
b
e
f
o
re
p
u
rsu
i
n
g
h
e
r
m
a
ste
r’s
d
e
g
re
e
.
S
h
e
h
a
s
c
o
m
p
lete
d
h
e
r
M
a
ste
r
o
f
Da
ta
S
c
i
e
n
c
e
stu
d
y
a
t
Ui
T
M
S
h
a
h
A
la
m
in
Ja
n
u
a
r
y
2
0
2
0
a
n
d
is
n
o
w
w
o
rk
in
g
in
so
f
t
wa
re
in
d
u
stry
to
d
e
v
e
lo
p
A
rti
f
icia
l
In
telli
g
e
n
c
e
p
ro
d
u
c
ts f
o
r
b
a
n
k
in
g
se
c
t
o
r.
Evaluation Warning : The document was created with Spire.PDF for Python.