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.
9
,
No
.
5
,
Octo
b
e
r
2
0
1
9
,
p
p
.
4
4
5
2
~
4
4
5
9
I
SS
N:
2088
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v
9
i
5
.
p
p
4
4
5
2
-
4
459
4452
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
co
r
e
.
co
m/
jo
u
r
n
a
ls
/in
d
ex
.
p
h
p
/
I
JE
C
E
H
y
brid app
ro
a
ch
:
na
iv
e bay
es a
nd
senti
m
en
t
VA
DE
R f
o
r
a
na
ly
z
ing
senti
m
e
nt
o
f
m
o
bile
unbo
x
ing
video
co
m
m
e
nts
Cha
it
hra
V
.
D
.
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter S
c
ien
c
e
,
CHRIST
(De
e
m
e
d
to
b
e
Un
iv
e
rsity
),
In
d
ia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Dec
3
1
,
2
0
1
8
R
ev
i
s
ed
A
p
r
24
,
2
0
1
9
A
cc
ep
ted
Ma
y
3
,
2
0
1
9
Re
v
o
lu
ti
o
n
in
s
o
c
ial
m
e
d
ia
h
a
s
a
t
trac
ted
th
e
u
se
rs
to
w
a
rd
s
v
id
e
o
sh
a
rin
g
sites
li
k
e
Yo
u
T
u
b
e
.
It
is
th
e
m
o
st
p
o
p
u
lar
so
c
ial
m
e
d
ia
site
w
h
e
re
p
e
o
p
le
v
iew
,
sh
a
re
a
n
d
in
tera
c
t
b
y
c
o
m
m
e
n
ti
n
g
o
n
th
e
v
id
e
o
s.
T
h
e
re
a
re
v
a
rio
u
s
ty
p
e
s
o
f
v
id
e
o
s
th
a
t
a
re
sh
a
re
d
b
y
th
e
u
se
rs
li
k
e
so
n
g
s,
m
o
v
ie
trail
e
rs,
n
e
w
s,
e
n
terta
in
m
e
n
t
e
tc.
No
w
a
d
a
y
s th
e
m
o
st t
re
n
d
in
g
v
id
e
o
s is
th
e
u
n
b
o
x
in
g
v
id
e
o
s
a
n
d
i
n
p
a
rti
c
u
lar
u
n
b
o
x
in
g
o
f
m
o
b
il
e
p
h
o
n
e
s
w
h
ich
g
e
ts
m
o
re
v
iew
s,
li
k
e
s/d
isli
k
e
s
a
n
d
c
o
m
m
e
n
ts.
A
n
a
l
y
z
in
g
th
e
c
o
m
m
e
n
ts
o
f
t
h
e
m
o
b
il
e
u
n
b
o
x
in
g
v
id
e
o
s
p
ro
v
id
e
s
t
h
e
o
p
in
i
o
n
o
f
th
e
v
iew
e
rs
to
w
a
rd
s
t
h
e
m
o
b
il
e
p
h
o
n
e
.
S
t
u
d
y
in
g
th
e
se
n
ti
m
e
n
t
e
x
p
re
ss
e
d
in
th
e
se
c
o
m
m
e
n
ts
s
h
o
w
if
th
e
m
o
b
il
e
p
h
o
n
e
is
g
e
tt
in
g
p
o
sit
iv
e
o
r
n
e
g
a
ti
v
e
f
e
e
d
b
a
c
k
.
A
H
y
b
rid
a
p
p
r
o
a
c
h
c
o
m
b
in
in
g
th
e
lex
ico
n
a
p
p
ro
a
c
h
S
e
n
ti
m
e
n
t
V
A
DER
a
n
d
m
a
c
h
in
e
lea
rn
in
g
a
lg
o
rit
h
m
Na
iv
e
Ba
y
e
s
is
a
p
p
li
e
d
o
n
t
h
e
c
o
m
m
e
n
ts
to
p
re
d
ict
th
e
se
n
ti
m
e
n
t.
S
e
n
ti
m
e
n
t
V
A
DER
h
a
s
a
g
o
o
d
im
p
a
c
t
o
n
t
h
e
Na
iv
e
Ba
y
e
s
c
las
sif
i
e
r
in
p
re
d
ictin
g
t
h
e
se
n
ti
m
e
n
t
o
f
th
e
c
o
m
m
e
n
t.
T
h
e
c
las
si
f
ier
a
c
h
iev
e
s
a
n
a
c
c
u
ra
c
y
o
f
7
9
.
7
8
%
a
n
d
F
1
sc
o
re
o
f
8
3
.
7
2
%
.
K
ey
w
o
r
d
s
:
Mo
b
ile
u
n
b
o
x
i
n
g
Naiv
e
b
a
y
e
s
Sen
ti
m
e
n
t a
n
a
l
y
s
is
Sen
ti
m
e
n
t V
A
DE
R
Yo
u
T
u
b
e
Co
p
y
rig
h
t
©
2
0
1
9
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
:
Siv
a
k
u
m
ar
R
,
Dep
ar
t
m
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
,
C
HR
I
ST
(
Dee
m
ed
to
b
e
Un
iv
er
s
it
y
)
,
B
en
g
al
u
r
u
,
I
n
d
ia
.
E
m
ail:
s
iv
a
k
u
m
ar
.
r
@
ch
r
i
s
tu
n
i
v
er
s
it
y
.
in
1.
I
NT
RO
D
UCT
I
O
N
So
cial
m
ed
ia
s
ites
ac
t
as
a
m
ed
iu
m
f
o
r
th
e
u
s
er
s
to
p
o
s
t
r
e
v
ie
w
s
o
f
t
h
e
p
r
o
d
u
ct,
s
er
v
ice
s
,
is
s
u
es
o
r
ev
en
t
s
.
T
h
e
u
s
er
g
e
n
er
ated
co
m
m
en
t
s
ar
e
o
f
g
r
ea
t
h
elp
f
o
r
p
eo
p
le
lo
o
k
in
g
f
o
r
t
h
e
v
ie
w
s
o
f
th
e
u
s
er
s
w
h
o
h
av
e
u
s
ed
th
e
p
r
o
d
u
ct
o
r
s
er
v
ices.
W
ith
th
e
g
r
o
w
i
n
g
p
o
p
u
lar
it
y
o
f
v
ar
io
u
s
s
o
cial
m
ed
ia
s
ites
Y
o
u
T
u
b
e
is
g
ain
i
n
g
a
lo
t
o
f
atte
n
tio
n
f
r
o
m
t
h
e
u
s
er
s
.
Yo
u
T
u
b
e
n
o
t
o
n
l
y
p
r
o
v
id
e
s
v
id
eo
s
b
u
t
also
a
p
lat
f
o
r
m
f
o
r
th
e
v
ie
w
er
s
to
ex
p
r
ess
t
h
eir
o
p
in
io
n
to
w
ar
d
s
th
e
v
id
eo
in
t
h
e
f
o
r
m
o
f
li
k
es,
d
is
lik
e
s
an
d
co
m
m
e
n
ts
.
Yo
u
T
u
b
e
h
as
v
ar
iet
y
o
f
u
s
er
g
en
er
ated
an
d
co
r
p
o
r
ate
m
ed
i
a
v
id
eo
s
w
h
ic
h
in
c
lu
d
es
m
o
v
i
e
tr
ailer
s
,
m
u
s
ic,
T
V
s
h
o
w
s
,
ed
u
ca
tio
n
a
l,
v
lo
g
g
i
n
g
a
n
d
u
n
b
o
x
i
n
g
v
id
eo
s
.
Un
b
o
x
i
n
g
v
id
eo
s
ar
e
th
e
tr
en
d
i
n
g
an
d
m
o
s
t
v
ie
w
ed
a
m
o
n
g
t
h
e
v
id
eo
s
a
v
ai
lab
le
o
n
Yo
u
T
u
b
e.
Un
b
o
x
in
g
v
id
eo
s
ar
e
r
eg
ar
d
in
g
p
r
o
d
u
cts
th
at
ar
e
n
e
w
i
n
th
e
m
ar
k
et,
l
ik
e
elec
tr
o
n
ic
g
ad
g
et
s
,
m
o
b
ile
p
h
o
n
e
s
,
clo
th
e
s
an
d
ac
ce
s
s
o
r
ies.
T
h
e
co
n
ce
p
t
o
f
u
n
b
o
x
in
g
v
id
eo
is
a
p
er
s
o
n
u
n
b
o
x
es
t
h
e
n
e
w
p
r
o
d
u
ct,
r
ev
ie
w
s
a
n
d
ex
p
r
ess
e
s
h
is
/h
er
o
p
in
io
n
to
w
ar
d
s
t
h
e
p
r
o
d
u
ct
as
a
n
e
n
d
u
s
e
r
.
W
ith
th
e
f
r
eq
u
e
n
t
r
elea
s
es
o
f
m
o
b
ile
p
h
o
n
es
b
y
v
ar
io
u
s
m
a
n
u
f
ac
t
u
r
er
e
m
b
ed
d
ed
w
it
h
tr
e
n
d
in
g
tech
n
o
lo
g
ie
s
,
it
b
ec
o
m
e
s
d
if
f
ic
u
lt
to
p
u
r
ch
a
s
e
a
m
o
b
ile
w
it
h
tech
n
o
lo
g
ie
s
w
o
r
th
t
h
e
m
o
n
e
y
.
R
e
v
ie
w
e
x
p
r
ess
ed
in
m
o
b
ile
u
n
b
o
x
in
g
v
id
eo
alo
n
g
w
it
h
t
h
e
co
m
m
en
ts
f
r
o
m
t
h
e
v
ie
w
er
s
to
w
ar
d
s
t
h
e
v
id
eo
g
u
id
e
o
n
e
t
o
k
n
o
w
m
o
r
e
ab
o
u
t
th
e
m
o
b
ile.
Sen
ti
m
e
n
t
a
n
al
y
s
is
o
f
m
o
b
ile
u
n
b
o
x
in
g
v
id
eo
co
m
m
e
n
t
h
el
p
s
an
al
y
zin
g
t
h
e
u
s
er
’
s
r
ea
cti
o
n
to
w
ar
d
s
th
e
m
o
b
ile
p
h
o
n
e.
T
h
e
m
eta
d
ata
o
f
th
e
v
id
eo
(
li
k
es,
d
is
li
k
es,
v
ie
w
s
an
d
co
m
m
e
n
t
s
)
e
x
p
r
ess
t
h
e
v
ie
w
er
’
s
r
ea
ctio
n
to
w
ar
d
s
t
h
e
p
h
o
n
e.
Me
tad
ata
s
u
ch
as
li
k
es
a
n
d
d
is
li
k
es
d
o
n
o
t
p
r
o
v
id
e
d
etaile
d
s
en
t
i
m
e
n
t
o
f
th
e
v
ie
w
er
s
,
t
h
er
ef
o
r
e
co
m
m
e
n
t
s
ar
e
co
n
s
id
er
ed
to
an
al
y
ze
t
h
e
v
ie
w
er
’
s
o
p
in
io
n
.
T
h
e
co
m
m
en
t
b
y
th
e
v
ie
w
er
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
Hyb
r
id
a
p
p
r
o
a
c
h
:
n
a
ive
b
a
ye
s
a
n
d
s
en
timen
t V
A
DE
R
f
o
r
a
n
a
lyzi
n
g
s
en
timen
t
…
(
C
h
a
ith
r
a
V
D
)
4453
ex
p
r
ess
es
a
p
o
s
iti
v
e
o
r
a
n
eg
a
tiv
e
o
p
in
io
n
to
w
ar
d
s
t
h
e
m
o
b
ile
p
h
o
n
e.
Fo
r
s
tu
d
y
i
n
g
t
h
is
u
n
b
o
x
i
n
g
o
f
L
G
G7
m
o
b
ile
p
h
o
n
e
is
r
a
n
d
o
m
l
y
co
n
s
id
er
ed
.
T
h
e
m
etad
ata
(
li
k
es,
d
is
lik
e
s
,
co
m
m
e
n
ts
etc.
)
ass
o
ciate
d
w
i
th
th
e
L
G
G7
p
h
o
n
e
is
ex
tr
ac
ted
u
s
i
n
g
an
o
n
li
n
e
Yo
u
T
u
b
e
s
cr
ap
p
er
.
H
y
b
r
id
ap
p
r
o
ac
h
w
h
ich
co
m
b
in
es
b
o
th
m
ac
h
i
n
e
lear
n
in
g
a
n
d
lex
ico
n
ap
p
r
o
ac
h
is
f
o
llo
w
ed
to
p
er
f
o
r
m
th
e
s
en
ti
m
e
n
t
clas
s
i
f
icatio
n
o
f
t
h
e
co
m
m
e
n
ts
.
T
h
e
r
esu
lt
s
o
f
t
h
e
clas
s
i
f
icatio
n
will
h
elp
t
h
e
v
ie
w
er
s
to
k
n
o
w
ab
o
u
t
th
e
m
o
b
ile
p
h
o
n
e
a
n
d
d
ec
id
e
is
i
t
w
o
r
t
h
b
u
y
i
n
g
o
r
n
o
t.
2
.
T
H
E
O
R
E
T
I
CA
L
B
ACK
G
R
O
UND
2
.
1
.
Senti
m
ent
a
na
ly
s
is
Sen
ti
m
e
n
t
an
a
l
y
s
is
[
1
]
is
a
s
t
u
d
y
f
o
r
k
n
o
w
in
g
t
h
e
s
en
ti
m
en
t
ex
p
r
ess
ed
in
t
h
e
u
s
er
’
s
o
p
in
io
n
to
w
ar
d
s
a
p
r
o
d
u
ct
o
r
an
is
s
u
e.
Sen
ti
m
e
n
t
ex
p
r
es
s
ed
b
y
a
n
in
d
i
v
id
u
a
l
ca
n
b
e
p
o
s
itiv
e,
n
e
g
ati
v
e
o
r
n
eu
tr
al.
I
n
th
is
p
ap
er
s
en
ti
m
e
n
t
is
co
n
s
id
er
ed
to
b
e
eith
er
p
o
s
itiv
e
o
r
n
e
g
ati
v
e.
P
o
s
iti
v
e
s
e
n
ti
m
en
t
s
h
o
w
s
th
a
t
th
e
u
s
er
h
a
s
lik
ed
t
h
e
p
r
o
d
u
ct
o
r
ag
r
ee
s
w
i
th
t
h
e
i
s
s
u
e
w
h
er
ea
s
n
e
g
ati
v
e
s
en
t
i
m
en
t
s
h
o
w
s
t
h
e
d
i
s
lik
e
to
w
a
r
d
s
th
e
p
r
o
d
u
ct
o
r
d
is
ag
r
ee
m
e
n
t
w
it
h
th
e
is
s
u
e.
Stu
d
y
in
g
t
h
e
s
e
n
ti
m
e
n
t
e
x
p
r
ess
ed
to
w
ar
d
s
th
e
p
r
o
d
u
ct
a
n
s
w
er
s
q
u
esti
o
n
s
h
o
w
th
e
p
r
o
d
u
ct
is
d
o
in
g
i
n
t
h
e
m
ar
k
et
w
h
e
th
er
it i
s
g
ett
in
g
p
o
s
iti
v
e
o
r
n
eg
ati
v
e
r
esp
o
n
s
e.
2
.
2
.
L
ev
els
o
f
s
ent
i
m
ent
a
na
ly
s
is
Sen
ti
m
e
n
t a
n
a
l
y
s
is
ca
n
b
e
p
er
f
o
r
m
ed
o
n
3
d
if
f
er
en
t le
v
el
s
as
s
h
o
w
n
in
F
ig
u
r
e
1
.
Fig
u
r
e
1
.
L
ev
el
s
o
f
Se
n
ti
m
e
n
t
An
al
y
s
i
s
2
.
2
.
1
.
Do
cu
m
ent
lev
el
Do
cu
m
e
n
t
le
v
el
s
e
n
ti
m
en
t
an
al
y
s
i
s
clas
s
if
ies
en
tire
d
o
cu
m
en
t
eit
h
er
as
p
o
s
itiv
e
o
r
n
eg
a
tiv
e.
T
h
is
lev
el
o
f
s
e
n
ti
m
e
n
t a
n
al
y
s
is
i
s
u
s
ed
w
h
e
n
t
h
e
en
t
ir
e
d
o
cu
m
e
n
t is r
elate
d
to
a
s
in
g
le
en
tit
y
o
r
to
p
ic.
2
.
2
.
2
.
Sente
nce
lev
el
Sen
te
n
ce
lev
e
l
s
en
ti
m
en
t
a
n
al
y
s
i
s
clas
s
i
f
ies
ea
c
h
s
en
ten
c
e
in
d
iv
id
u
all
y
b
ased
o
n
t
h
e
s
en
ti
m
e
n
t
ex
p
r
ess
ed
.
T
h
is
lev
el
o
f
s
en
ti
m
e
n
t
clas
s
i
f
icatio
n
ac
t
s
as
s
u
b
j
ec
tiv
e
clas
s
if
icatio
n
.
T
h
e
class
if
icatio
n
i
n
s
en
te
n
ce
le
v
el
ca
n
b
e
e
x
p
r
ess
ed
as
p
o
s
itiv
e,
n
e
g
ati
v
e
o
r
n
e
u
tr
al.
I
t
d
if
f
er
en
tia
tes
b
et
w
ee
n
th
e
s
u
b
j
ec
tiv
e
an
d
th
e
o
b
j
ec
tiv
e
in
f
o
r
m
atio
n
p
r
es
en
t in
t
h
e
s
e
n
te
n
ce
s
.
2
.
2
.
3
.
E
ntit
y
o
r
a
s
pect
lev
el
Asp
ec
t
lev
el
co
n
ce
n
tr
ates
o
n
t
h
e
o
p
in
io
n
,
w
it
h
th
e
id
ea
th
at
an
o
p
in
io
n
co
n
s
i
s
ts
o
f
a
s
en
t
i
m
en
t
an
d
a
tar
g
et.
I
t
p
er
f
o
r
m
s
a
f
i
n
e
-
g
r
ai
n
ed
s
en
ti
m
e
n
t
an
al
y
s
i
s
an
d
d
if
f
er
en
tiate
s
w
h
at
a
u
s
er
w
a
n
ts
an
d
d
o
es
n
o
t
w
an
t.
T
h
is
lev
el
o
f
s
e
n
ti
m
e
n
t
a
n
al
y
s
is
i
s
p
r
ef
er
r
ed
w
h
e
n
m
o
r
e
t
h
an
t
w
o
p
r
o
d
u
cts
r
ev
ie
w
s
ar
e
to
an
al
y
ze
d
o
r
th
e
f
ea
t
u
r
es
o
f
t
h
e
p
r
o
d
u
ct
h
as
to
b
e
an
al
y
ze
d
to
k
n
o
w
w
h
a
t
f
ea
tu
r
e
is
g
etti
n
g
m
o
r
e
p
o
s
iti
v
e
r
esp
o
n
s
e
o
r
n
eg
a
tiv
e
r
esp
o
n
s
e.
2
.
3
.
T
ec
hn
i
qu
e
s
o
f
s
ent
i
m
ent
cla
s
s
if
ica
t
io
n
Sen
ti
m
e
n
t
o
f
a
d
o
cu
m
e
n
t
o
r
a
s
en
te
n
ce
ca
n
b
e
cla
s
s
i
f
ied
b
y
t
h
e
f
o
l
lo
w
in
g
m
et
h
o
d
s
as
s
h
o
w
n
in
Fig
u
r
e
2
.
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.
9
,
No
.
5
,
Octo
b
er
2
0
1
9
:
4
4
5
2
-
4459
4454
Fig
u
r
e
2
.
C
lass
if
ica
tio
n
tec
h
n
i
q
u
es
2
.
3
.
1
.
M
a
chine
lea
rning
a
pp
ro
a
ch
Sen
ti
m
e
n
t
class
if
icatio
n
u
s
i
n
g
m
ac
h
i
n
e
lear
n
i
n
g
ap
p
r
o
ac
h
is
d
iv
id
ed
in
to
s
u
p
er
v
is
ed
lea
r
n
in
g
an
d
u
n
s
u
p
er
v
is
ed
lear
n
i
n
g
.
S
u
p
er
v
is
ed
lear
n
in
g
w
a
y
o
f
s
e
n
ti
m
en
t
clas
s
if
icatio
n
i
s
u
s
ed
f
o
r
d
ataset
w
h
er
ea
s
u
n
s
u
p
er
v
is
ed
lear
n
i
n
g
m
e
th
o
d
s
ar
e
u
s
ed
f
o
r
d
ataset
w
i
th
o
u
t
th
e
lab
el
s
.
T
h
er
e
ar
e
m
a
n
y
s
u
p
er
v
is
ed
clas
s
i
f
ier
s
an
d
th
e
m
o
s
t c
o
m
m
o
n
l
y
u
s
ed
class
i
f
ier
s
ar
e
v
ar
io
u
s
p
r
o
b
ab
ilis
tic
clas
s
i
f
ier
s
an
d
Nai
v
e
B
a
y
e
s
class
i
f
ier
s
.
2
.
3
.
2
.
L
ex
ico
n
ba
s
ed
a
p
pro
a
ch
Sen
ti
m
e
n
t
cla
s
s
i
f
icat
io
n
u
s
i
n
g
lex
ico
n
b
ased
h
as
t
w
o
s
u
b
-
a
p
p
r
o
ac
h
es
d
ictio
n
ar
y
b
ased
a
n
d
co
r
p
u
s
b
ased
.
Dictio
n
ar
y
b
ased
ap
p
r
o
ac
h
d
ep
en
d
s
o
n
t
h
e
d
ict
io
n
a
r
y
a
n
d
s
ea
r
ch
e
s
f
o
r
s
ee
d
w
o
r
d
s
in
t
h
e
d
ictio
n
ar
y
.
C
o
r
p
u
s
b
ased
ap
p
r
o
ac
h
h
as
a
lis
t
o
f
w
o
r
d
s
a
n
d
f
i
n
d
s
t
h
e
o
p
i
n
io
n
w
o
r
d
s
i
n
t
h
e
lar
g
e
co
r
p
u
s
i
n
o
r
d
er
to
ex
tr
ac
t
th
e
s
e
m
a
n
tic
o
r
ien
tatio
n
.
2
.
3
.
3
.
H
y
brid
a
pp
ro
a
ch
H
y
b
r
id
ap
p
r
o
ac
h
co
m
b
in
es
b
o
th
th
e
m
ac
h
i
n
e
lear
n
i
n
g
a
n
d
lex
ico
n
b
ased
ap
p
r
o
ac
h
in
o
r
d
er
to
p
er
f
o
r
m
t
h
e
s
e
n
ti
m
e
n
t c
las
s
i
f
i
ca
tio
n
.
3.
RE
L
AT
E
D
WO
RK
T
h
e
w
o
r
k
[
2
]
b
y
Han
if
B
h
u
i
y
an
et
al.
a
n
al
y
ze
d
t
h
e
m
eta
d
ata
(
lik
es/d
is
li
k
e
s
/v
ie
w
s
/co
m
m
en
ts
)
i
n
o
r
d
er
to
r
etr
iev
e
th
e
m
o
s
t
r
elev
an
t
an
d
p
o
p
u
lar
v
id
eo
b
ased
o
n
th
e
s
ea
r
ch
b
y
u
s
i
n
g
L
e
x
ico
n
s
en
t
i
m
e
n
t
class
i
f
ier
Sen
ti
Stre
n
g
t
h
.
B
y
ap
p
ly
i
n
g
th
i
s
ap
p
r
o
ac
h
m
a
x
i
m
u
m
o
f
7
5
.
4
%
ac
cu
r
ac
y
is
o
b
tain
ed
in
o
r
d
er
to
r
etr
iev
e
th
e
r
elev
a
n
t
an
d
ef
f
ec
tiv
e
v
id
eo
s
.
Au
th
o
r
s
Fi
k
to
r
et
al.
in
[
3
]
p
r
o
p
o
s
ed
Su
p
p
o
r
t
Vec
to
r
Ma
ch
in
e
f
o
r
class
i
f
icatio
n
an
d
a
L
e
x
ico
n
m
et
h
o
d
to
f
i
n
d
th
e
p
er
ce
n
ta
g
e
o
f
th
e
s
e
n
ti
m
en
t
clas
s
to
u
n
d
er
s
tan
d
t
h
e
ch
ar
ac
ter
an
d
th
e
p
er
f
o
r
m
a
n
ce
o
f
A
h
o
k
as
a
g
o
v
er
n
o
r
r
esu
lted
i
n
an
a
cc
u
r
ac
y
o
f
8
4
%.
T
h
e
w
o
r
k
b
y
Y.
Han
et
al.
i
n
[
4
]
p
r
o
p
o
s
ed
a
m
o
r
p
h
o
lo
g
ical
s
en
ten
ce
p
atter
n
m
o
d
el
w
h
ic
h
u
s
es
asp
ec
t
b
ased
s
e
n
ti
m
e
n
t
le
x
i
co
n
p
ar
t
o
f
s
p
e
ec
h
to
in
cr
ea
s
e
th
e
ac
c
u
r
ac
y
o
f
t
h
e
p
r
o
b
ab
ilit
y
m
o
d
el.
T
h
e
e
x
is
tin
g
m
o
d
el
ac
cu
r
ac
y
w
a
s
i
n
cr
ea
s
ed
u
p
to
9
1
.
2
%
an
d
also
h
e
lp
s
i
n
f
i
n
d
i
n
g
th
e
asp
ec
ts
t
h
at
ar
e
ad
j
ac
en
t
w
it
h
r
esp
ec
t
to
p
ar
t
o
f
s
p
ee
c
h
.
T
h
e
p
ap
er
[
5
]
b
y
S.
R
an
g
as
w
a
m
y
et
al.
p
r
o
p
o
s
ed
a
tech
n
iq
u
e
to
ex
tr
ac
t
a
n
d
an
al
y
ze
m
e
tad
ata
an
d
m
a
k
e
d
ec
is
io
n
s
b
y
u
s
in
g
t
h
e
v
id
eo
UR
L
.
Me
tad
ata
r
elate
d
t
o
th
e
v
id
eo
is
e
x
tr
ac
ted
w
it
h
t
h
e
s
ta
g
es o
f
ca
te
g
o
r
izatio
n
,
p
ar
s
in
g
an
d
lo
o
k
u
p
o
f
th
e
m
etad
ata.
T
h
e
w
o
r
k
[
6
]
b
y
Vip
u
l
et
a
l.
u
s
e
th
e
r
u
le
b
ased
lex
ico
n
m
o
d
el
s
e
n
ti
m
e
n
t
V
ADE
R
to
a
n
al
y
ze
th
e
t
w
i
tter
d
ata.
T
ex
t B
lo
b
is
u
s
ed
to
in
cr
ea
s
e
th
e
ac
c
u
r
ac
y
an
d
a
n
ef
f
icie
n
c
y
o
f
8
5
-
9
0
% is
r
ea
c
h
ed
.
T
h
e
p
ap
er
[
7
]
b
y
A
s
h
o
k
e
t
al
.
p
r
o
p
o
s
ed
a
f
r
a
m
e
w
o
r
k
u
s
in
g
t
h
e
m
ac
h
i
n
e
lear
n
i
n
g
ap
p
r
o
ac
h
es
a
n
d
s
en
ti
m
e
n
t
a
n
al
y
s
i
s
to
o
p
ti
m
ize
th
e
s
ea
r
c
h
q
u
er
ies
f
o
r
s
u
g
g
est
in
g
t
h
e
r
esta
u
r
an
ts
.
SVM,
Nai
v
e
B
a
y
e
s
,
R
a
n
d
o
m
Fo
r
est
an
d
Ma
x
i
m
u
m
E
n
tr
o
p
y
ar
e
u
s
ed
to
eli
m
i
n
ate
th
e
u
n
w
a
n
ted
d
ata
a
n
d
co
n
s
id
er
t
h
o
s
e
d
ata
w
h
ic
h
y
ield
s
b
etter
r
esu
lt
s
to
t
h
e
u
s
er
s
ea
r
c
h
q
u
er
ie
s
.
S
m
ita
s
h
r
ee
C
h
o
u
d
h
u
r
y
e
t
al.
i
n
[
8
]
p
r
o
p
o
s
e
d
an
u
n
s
u
p
er
v
is
ed
lex
ico
n
ap
p
r
o
ac
h
to
f
in
d
th
e
p
o
lar
ity
o
f
th
e
u
s
er
co
m
m
e
n
t.
Se
n
ti
W
o
r
d
n
et
is
u
s
ed
to
f
in
d
th
e
s
en
ti
m
en
t
p
o
lar
ity
a
n
d
lis
t
i
s
p
r
ep
ar
ed
b
y
n
e
g
ati
n
g
t
h
e
w
o
r
d
s
i
n
t
h
e
co
m
m
e
n
t.
C
o
m
b
i
n
atio
n
o
f
Se
n
ti
W
o
r
d
n
et
an
d
t
h
e
ad
d
ed
lis
t
p
er
f
o
r
m
s
r
es
u
lts
i
n
b
etter
ca
te
g
o
r
izatio
n
o
f
t
h
e
co
m
m
e
n
ts
a
n
d
th
e
o
u
tco
m
e
i
s
th
a
t n
e
g
ati
v
e
s
en
ti
m
e
n
t i
s
p
o
o
r
e
r
th
an
t
h
e
p
o
s
iti
v
e.
Au
t
h
o
r
s
Am
ar
Kr
is
h
n
a
et
a
l.
in
[
9
]
p
r
o
p
o
s
ed
m
ac
h
i
n
e
lear
n
i
n
g
tech
n
iq
u
es
to
a
n
al
y
ze
t
h
e
s
en
ti
m
e
n
t o
f
th
e
co
m
m
en
t in
o
r
d
er
to
id
en
tify
t
h
e
tr
e
n
d
s
,
s
ea
s
o
n
alit
y
a
n
d
f
o
r
ec
asts
.
Naiv
e
B
ay
e
s
clas
s
i
f
icatio
n
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
Hyb
r
id
a
p
p
r
o
a
c
h
:
n
a
ive
b
a
ye
s
a
n
d
s
en
timen
t V
A
DE
R
f
o
r
a
n
a
lyzi
n
g
s
en
timen
t
…
(
C
h
a
ith
r
a
V
D
)
4455
is
u
s
ed
to
ca
lcu
la
te
th
e
p
o
lar
ity
o
f
co
m
m
e
n
ts
,
d
ec
o
m
p
o
s
e
(
)
f
u
n
ctio
n
o
f
R
is
u
s
ed
to
f
i
n
d
ti
m
e
s
er
ies
a
n
d
to
g
iv
e
t
h
e
o
v
er
all
tr
en
d
.
W
ek
a
i
s
u
s
ed
f
o
r
f
o
r
ec
asti
n
g
o
f
2
6
w
ee
k
s
.
Asad
Ulla
h
R
a
fi
q
K
h
a
n
et
al.
i
n
[
1
0
]
u
s
ed
Naiv
e
B
a
y
e
s
al
g
o
r
ith
m
a
m
ac
h
in
e
lear
n
i
n
g
ap
p
r
o
ac
h
to
p
er
f
o
r
m
m
u
lti
-
lab
el
cla
s
s
i
f
icatio
n
.
ME
KA
to
o
l is u
s
ed
f
o
r
m
u
lt
i
-
lab
el
clas
s
if
icatio
n
.
Naiv
e
as
s
u
m
p
tio
n
r
eg
ar
d
in
g
n
ei
g
h
b
o
r
h
o
o
d
k
e
y
w
o
r
d
s
p
er
f
o
r
m
ed
w
ell
a
s
co
m
p
ar
e
to
o
th
er
e
x
p
er
i
m
e
n
t
al
s
etti
n
g
.
T
h
e
p
ap
er
w
o
r
k
[
1
1
]
b
y
R
i
s
h
a
n
k
i
J
ain
u
s
ed
S
AS
E
M
T
ex
t
Mi
n
er
s
o
f
t
w
ar
e
to
p
er
f
o
r
m
s
en
t
i
m
e
n
t
an
al
y
s
i
s
o
n
m
o
v
ie
tr
ailer
co
m
m
en
ts
to
p
r
ed
ict
th
e
co
llectio
n
o
f
t
h
e
b
o
x
-
o
f
f
ice
.
Sen
ti
m
e
n
t
s
co
r
es
ar
e
ca
lc
u
lat
ed
an
d
g
r
o
s
s
tr
e
n
d
i
s
p
lo
tted
w
h
ic
h
s
h
o
w
s
th
at
i
n
itial
l
y
p
eo
p
le
w
er
e
lo
o
k
in
g
f
o
r
w
ar
d
to
th
e
r
elea
s
e
,
a
f
ter
th
e
m
o
v
ie
la
u
n
c
h
ed
v
ie
w
er
s
d
id
n
o
t e
n
j
o
y
.
Au
t
h
o
r
s
Sier
s
d
o
r
f
er
et
al.
i
n
[
1
2
]
u
s
ed
Naiv
e
B
a
y
e
s
class
i
f
ier
f
o
r
p
r
ed
ictin
g
th
e
co
m
m
e
n
t
ac
ce
p
tan
ce
.
St
u
d
ies
o
f
lar
g
e
d
ataset
u
s
i
n
g
Se
n
tiW
o
r
d
Net
an
d
Yo
u
T
u
b
e
m
e
tad
at
a
r
ev
ea
led
s
tr
o
n
g
d
ep
en
d
en
cies
b
et
w
ee
n
d
i
f
f
er
en
t
s
e
n
ti
m
e
n
ts
e
x
p
r
ess
ed
in
co
m
m
en
t
s
,
co
m
m
e
n
t
r
atin
g
s
a
n
d
to
p
ic
o
r
ien
tatio
n
d
is
cu
s
s
ed
i
n
t
h
e
v
id
eo
co
n
te
n
t
.
T
h
e
p
ap
er
[
1
3
]
b
y
Kau
s
h
ik
p
r
o
p
o
s
ed
a
b
etter
tex
t
-
b
ased
s
en
ti
m
en
t
m
o
d
el
f
o
r
lar
g
e
d
ataset
w
i
th
t
h
e
ai
m
to
r
ed
u
ce
th
e
co
m
p
le
x
it
y
o
f
s
e
n
ti
m
e
n
t
m
o
d
el.
ME
m
o
d
elin
g
tech
n
iq
u
e
u
s
ed
f
o
r
d
eter
m
in
i
n
g
th
e
p
o
lar
it
y
o
f
th
e
co
m
m
e
n
ts
an
d
is
o
b
s
er
v
ed
t
h
at
ME
s
y
s
te
m
o
u
tp
er
f
o
r
m
s
th
e
Naiv
e
B
a
y
e
s
an
d
th
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
ME
te
ch
n
iq
u
e
is
s
lig
h
tl
y
i
n
f
er
io
r
to
th
e
SVM
tec
h
n
iq
u
e.
W
o
r
k
b
y
A
lia
k
s
ei
Se
v
er
y
n
et
al.
in
[
1
4
]
d
ef
in
ed
a
s
y
s
te
m
a
tic
ap
p
r
o
ac
h
f
o
r
o
p
in
io
n
m
i
n
in
g
a
n
d
a
r
o
b
u
s
t
s
h
allo
w
s
y
n
tactic
s
tr
u
ct
u
r
e
f
o
r
i
m
p
r
o
v
i
n
g
m
o
d
el
ad
ap
tab
ilit
y
.
A
s
u
p
er
v
is
ed
m
u
lti
-
cla
s
s
c
la
s
s
i
f
ier
i
s
u
s
ed
to
ca
r
r
y
o
u
t
t
h
e
ef
f
ec
ti
v
e
o
p
in
io
n
m
i
n
in
g
a
n
d
k
er
n
els
ar
e
d
e
f
i
n
ed
to
i
m
p
r
o
v
e
th
e
f
ea
t
u
r
e
v
ec
to
r
s
.
T
h
e
p
ap
er
[
1
5
]
b
y
T
u
r
n
e
y
p
r
o
v
id
es
t
h
e
u
n
s
u
p
er
v
is
ed
lear
n
i
n
g
al
g
o
r
it
h
m
s
to
class
i
f
y
t
h
e
r
ev
ie
w
s
a
s
r
ec
o
m
m
e
n
d
ed
o
r
n
o
t
r
ec
o
m
m
en
d
ed
.
Se
m
a
n
tic
o
r
ien
tatio
n
o
f
th
e
p
h
r
ase
s
i
s
esti
m
ated
u
s
i
n
g
t
h
e
P
MI
-
I
R
alg
o
r
ith
m
an
d
i
t
atta
in
s
a
n
a
v
er
ag
e
ac
cu
r
ac
y
o
f
7
4
%.T
h
e
w
o
r
k
[
1
6
]
b
y
P
an
g
c
lass
i
f
ies
a
d
o
cu
m
e
n
t
n
o
t
b
y
to
p
ic
b
u
t
in
s
tead
b
y
o
v
er
all
s
e
n
t
i
m
e
n
t.
T
o
p
er
f
o
r
m
th
is
m
ac
h
i
n
e
lear
n
i
n
g
al
g
o
r
it
h
m
s
l
ik
e
Naï
v
e
B
a
y
e
s
,
Ma
x
i
m
u
m
E
n
tr
o
p
y
an
d
Su
p
p
o
r
t
Ma
ch
i
n
e
v
ec
to
r
s
ar
e
u
s
ed
.
I
n
ter
m
s
o
f
r
elati
v
e
p
er
f
o
r
m
a
n
ce
,
Naï
v
e
B
a
y
es p
er
f
o
r
m
s
w
o
r
s
t a
n
d
SVMs p
er
f
o
r
m
s
th
e
b
est.
4.
M
E
T
H
O
DO
L
O
G
Y
Fig
u
r
e
3
is
th
e
p
r
o
p
o
s
ed
f
r
am
e
w
o
r
k
to
p
er
f
o
r
m
s
e
n
ti
m
e
n
t
an
al
y
s
is
o
f
th
e
m
o
b
ile
u
n
b
o
x
in
g
v
id
eo
co
m
m
e
n
t
s
.
R
u
le
b
ased
lex
ico
n
ap
p
r
o
ac
h
Sen
ti
m
e
n
t
V
A
DE
R
is
u
s
ed
to
lab
el
th
e
e
x
tr
ac
te
d
d
ata.
L
ab
elled
d
ata
is
p
r
e
-
p
r
o
ce
s
s
ed
to
to
k
en
ize,
r
em
o
v
e
s
to
p
w
o
r
d
s
,
p
er
f
o
r
m
s
te
m
m
i
n
g
an
d
co
u
n
t
v
ec
to
r
izer
.
Naiv
e
B
a
y
es
alg
o
r
ith
m
is
ap
p
lied
o
n
p
r
e
-
p
r
o
ce
s
s
ed
d
ata
to
class
if
y
t
h
e
co
m
m
en
t
s
as e
it
h
er
p
o
s
itiv
e
o
r
n
eg
ati
v
e.
Fig
u
r
e
3
.
P
r
o
p
o
s
ed
f
r
am
e
w
o
r
k
4
.
1
.
Da
t
a
ex
t
ra
ct
io
n
T
h
e
d
ata
r
eq
u
ir
ed
f
o
r
th
e
p
er
f
o
r
m
i
n
g
s
en
ti
m
en
t
a
n
al
y
s
is
o
f
th
e
m
o
b
ile
u
n
b
o
x
i
n
g
v
id
eo
is
ex
tr
ac
ted
f
r
o
m
Yo
u
T
u
b
e.
T
h
e
m
etad
ata
(
lik
e
s
,
d
is
li
k
es
an
d
co
m
m
en
t
s
)
ass
o
ciate
d
w
it
h
t
h
e
v
id
eo
is
ex
tr
ac
ted
b
y
u
s
i
n
g
o
n
lin
e
Yo
u
T
u
b
e
s
cr
ap
p
er
(
h
ttp
://
y
tco
m
m
e
n
t
s
.
k
lo
s
ter
m
a
n
n
.
c
a/)
.
Yo
u
T
u
b
e
s
cr
a
p
p
er
ex
tr
ac
ts
all
th
e
co
m
m
en
t
s
an
d
m
etad
ata
as
s
o
ciate
d
w
it
h
it.
T
h
e
ex
tr
ac
ted
d
ata
ar
e
a
v
ail
ab
le
in
th
e
f
o
r
m
o
f
C
SV
an
d
J
SON
f
o
r
m
at,
C
SV
f
o
r
m
at
i
s
co
n
s
id
er
ed
f
o
r
an
al
y
zin
g
t
h
e
d
ata
a
n
d
clas
s
i
f
y
in
g
t
h
e
co
m
m
en
ts
.
I
n
th
is
p
ap
er
u
n
b
o
x
i
n
g
o
f
L
G
G7
m
o
b
ile
i
s
r
an
d
o
m
l
y
co
n
s
i
d
er
ed
to
s
tu
d
y
t
h
e
s
en
ti
m
e
n
t
ass
o
ciate
d
w
it
h
t
h
e
v
id
eo
co
m
m
en
ts
.
T
h
e
o
n
lin
e
Yo
u
T
u
b
e
s
cr
ap
p
er
is
u
s
ed
to
ex
tr
ac
t
6
2
4
8
co
m
m
e
n
t
s
an
d
th
e
m
etad
ata
ass
o
ciate
d
w
it
h
th
e
v
id
eo
.
T
h
e
e
x
tr
ac
ted
d
ataset
co
n
s
is
t
s
o
f
1
4
m
etad
ata
li
k
e
u
s
er
id
,
u
s
er
n
a
m
e,
d
ate,
ti
m
e
s
ta
m
p
,
li
k
es,
co
m
m
en
t
e
tc.
,
w
h
ich
p
r
o
v
id
es
th
e
d
etai
ls
o
f
t
h
e
u
s
er
n
a
m
e,
ti
m
e
a
n
d
d
ate
w
h
e
n
th
e
co
m
m
e
n
t
w
as
m
ad
e
an
d
also
p
r
o
v
id
es
d
etails
o
f
th
e
n
u
m
b
e
r
o
f
lik
es
an
d
d
is
li
k
es
f
o
r
th
e
co
m
m
e
n
t
s
.
W
ith
all
th
ese
m
etad
ata
a
v
ailab
le
o
n
l
y
th
e
co
m
m
en
t
is
co
n
ce
n
tr
at
ed
to
s
tu
d
y
t
h
e
s
e
n
ti
m
en
t
ex
p
r
ess
ed
b
y
t
h
e
u
s
er
to
w
ar
d
s
t
h
e
m
o
b
ile
p
h
o
n
e
b
ei
n
g
u
n
b
o
x
ed
.
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.
9
,
No
.
5
,
Octo
b
er
2
0
1
9
:
4
4
5
2
-
4459
4456
4
.
2
.
Da
t
a
la
belin
g
us
ing
s
ent
i
m
e
nt
VADER
L
G
G7
u
n
b
o
x
i
n
g
v
id
eo
d
atase
t
d
o
es
n
o
t
co
n
tain
o
u
tp
u
t
attr
i
b
u
te
th
at
lab
el
s
th
e
co
m
m
e
n
t
as
p
o
s
itiv
e
o
r
n
eg
ati
v
e
in
o
r
d
er
to
tr
ain
a
s
u
p
er
v
is
ed
clas
s
i
f
ier
.
A
r
u
le
b
ased
lex
ico
n
ap
p
r
o
ac
h
Sen
ti
m
e
n
t
V
A
DE
R
(
Vale
n
ce
A
w
ar
e
Dictio
n
ar
y
f
o
r
s
E
n
ti
m
e
n
t
R
ea
s
o
n
i
n
g
)
[
1
7
-
1
8
]
is
ap
p
lied
to
g
et
th
e
o
u
tp
u
t
attr
ib
u
te
th
at
lab
els
th
e
co
m
m
e
n
ts
a
s
p
o
s
itiv
e,
n
e
u
tr
al
o
r
n
eg
at
iv
e.
Se
n
ti
m
e
n
t
VADE
R
n
o
t
o
n
l
y
a
s
s
i
g
n
s
p
o
lar
it
y
to
th
e
co
m
m
e
n
t
b
u
t
also
a
s
s
i
g
n
s
th
e
i
n
ten
s
it
y
v
alu
e.
U
n
lik
e
p
o
lar
it
y
b
ased
m
et
h
o
d
w
h
ic
h
cla
s
s
i
f
ie
s
t
h
e
s
en
ten
ce
s
a
s
p
o
s
iti
v
e,
n
eg
at
iv
e
o
r
n
e
u
tr
al
w
i
th
t
h
e
v
alu
es
1
,
-
1
a
n
d
0
r
esp
ec
ti
v
el
y
,
th
e
v
ale
n
ce
o
r
i
n
te
n
s
i
t
y
b
ase
d
m
et
h
o
d
co
n
s
id
er
s
th
e
in
ten
s
it
y
v
al
u
e
as
s
o
ciate
d
w
it
h
t
h
e
w
o
r
d
s
to
r
an
g
e
f
r
o
m
-
1
to
1
.
Fo
r
ex
a
m
p
le,
t
h
e
w
o
r
d
“
g
o
o
d
”
an
d
“
e
x
ce
lle
n
t”
w
i
ll
h
a
v
e
s
a
m
e
p
o
la
r
it
y
o
f
1
i
n
p
o
lar
it
y
b
a
s
ed
ap
p
r
o
ac
h
,
w
h
er
ea
s
i
n
v
ale
n
ce
b
ased
“
ex
ce
llen
t”
is
co
n
s
id
er
e
d
to
b
e
m
o
r
e
p
o
s
itiv
e
th
a
n
“g
o
o
d
”
s
o
th
e
in
te
n
s
it
y
v
alu
e
f
o
r
ex
ce
lle
n
t
w
il
l b
e
m
o
r
e
th
an
t
h
e
in
ten
s
it
y
v
al
u
e
o
f
g
o
o
d
.
I
n
th
i
s
p
ap
er
Sen
ti
m
e
n
t
V
AD
E
R
w
h
ich
i
s
a
r
u
le
b
ased
ap
p
r
o
ac
h
is
ap
p
lied
u
s
in
g
t
h
e
p
y
th
o
n
p
ac
k
ag
e
Sen
ti
m
e
n
tI
n
ten
s
it
y
An
a
l
y
ze
r
.
T
h
is
ass
i
g
n
s
th
e
i
n
ten
s
it
y
o
f
in
d
iv
id
u
a
l
co
m
m
en
t
to
w
h
a
t
ex
ten
t
i
s
p
o
s
itiv
e,
n
eg
at
iv
e
a
n
d
n
e
u
tr
al.
A
lo
n
g
w
it
h
t
h
e
3
v
al
u
es
it
al
s
o
p
r
o
v
id
es
th
e
co
m
p
o
u
n
d
v
al
u
e
w
h
ic
h
is
t
h
e
n
o
r
m
alize
d
v
alu
e
as
s
o
ciate
d
w
it
h
t
h
e
co
m
m
en
ts
th
a
t
r
an
g
e
f
r
o
m
-
1
to
1
.
C
u
t
-
o
f
f
v
alu
e
ar
e
ch
o
s
e
n
f
o
r
th
e
co
m
p
o
u
n
d
v
al
u
e
w
h
er
e
th
e
v
al
u
e
s
m
o
r
e
th
a
n
0
.
2
ar
e
m
ad
e
1
an
d
v
alu
es
w
h
ic
h
ar
e
less
th
a
n
-
0
.
2
ar
e
m
ad
e
-
1
in
d
icati
n
g
p
o
s
itiv
e
a
n
d
n
e
g
ati
v
e
co
m
m
e
n
t
r
esp
ec
ti
v
el
y
.
T
h
e
v
alu
e
s
b
et
w
ee
n
0
.
2
an
d
-
0
.
2
ar
e
m
ad
e
0
m
a
k
i
n
g
n
e
u
tr
al
co
m
m
e
n
t
s
.
I
n
cr
ea
s
e
i
n
t
h
e
r
a
n
g
e
(
-
0
.
2
an
d
0
.
2
)
w
i
ll
i
n
cr
ea
s
e
th
e
n
e
u
tr
al
co
m
m
en
t
s
a
n
d
d
e
cr
ea
s
e
th
e
p
o
s
iti
v
e
an
d
n
e
g
ati
v
e
co
m
m
e
n
ts
.
I
t
b
e
co
m
e
s
d
i
f
f
icu
lt
to
tr
ain
th
e
cla
s
s
i
f
ier
w
it
h
le
s
s
n
e
g
ativ
e
an
d
p
o
s
itiv
e
co
m
m
e
n
t
s
.
T
h
er
ef
o
r
e,
th
e
r
an
g
e
is
co
n
s
id
er
ed
to
b
e
0
.
2
an
d
-
0
.
2
.
4
.
3
.
Da
t
a
pre
-
pro
ce
s
s
ing
C
o
m
m
e
n
ts
ar
e
p
r
e
-
p
r
o
ce
s
s
ed
to
ap
p
ly
th
e
m
ac
h
in
e
lear
n
i
n
g
alg
o
r
ith
m
Nai
v
e
B
a
y
es.
T
h
e
co
m
m
e
n
t
s
w
it
h
t
h
e
n
eu
tr
al
v
alu
e
0
ar
e
r
e
m
o
v
ed
an
d
co
m
m
en
t
s
w
ith
v
alu
es
1
an
d
-
1
ar
e
co
n
s
id
er
ed
to
p
r
e
-
p
r
o
ce
s
s
an
d
ap
p
ly
t
h
e
Nai
v
e
B
a
y
es
al
g
o
r
ith
m
.
Data
is
p
r
e
-
p
r
o
ce
s
s
ed
to
r
em
o
v
e
s
to
p
w
o
r
d
s
,
s
te
m
t
h
e
w
o
r
d
s
to
th
e
r
o
o
t
w
o
r
d
u
s
i
n
g
th
e
P
o
r
ter
Ste
m
m
e
r
alg
o
r
ith
m
.
C
o
u
n
t
v
ec
to
r
izer
is
ap
p
lied
to
co
n
v
er
t th
e
w
o
r
d
s
as v
ec
to
r
s
.
4
.
3
.
1
.
T
o
k
eniz
ing
T
o
k
en
izin
g
i
s
co
n
s
id
er
ed
as
a
n
i
m
p
o
r
tan
t
p
r
e
-
p
r
o
ce
s
s
in
g
s
t
ep
in
cla
s
s
i
f
y
i
n
g
a
te
x
t
d
ata.
T
r
ain
in
g
a
n
alg
o
r
ith
m
to
clas
s
i
f
y
te
x
t
d
at
a
b
y
u
s
i
n
g
en
tire
d
o
cu
m
e
n
t
o
r
a
s
en
te
n
ce
is
v
er
y
h
ar
d
.
So
,
it
is
n
ec
e
s
s
ar
y
to
to
k
en
ize
t
h
e
s
e
n
te
n
ce
in
to
w
o
r
d
s
an
d
tr
ain
th
e
cla
s
s
i
f
ier
w
it
h
th
e
p
o
s
iti
v
e
an
d
n
e
g
ati
v
e
w
o
r
d
s
.
4
.
3
.
2
.
Sto
pwo
rds
T
h
ese
ar
e
th
e
w
o
r
d
s
t
h
at
d
o
n
o
t
ad
d
to
o
m
u
ch
v
al
u
e
in
cla
s
s
if
y
i
n
g
it
as
eit
h
er
p
o
s
itiv
e
o
r
n
eg
ati
v
e
w
o
r
d
.
T
h
e
y
ar
e
n
o
t
n
ec
e
s
s
ar
y
in
cla
s
s
i
f
y
i
n
g
t
h
e
s
en
te
n
ce
o
r
d
o
cu
m
en
t
s
o
s
to
p
w
o
r
d
s
o
f
E
n
g
lis
h
(
a,
as,
is
,
t
h
e
etc.
,
)
ar
e
r
em
o
v
ed
f
r
o
m
t
h
e
s
e
n
ten
ce
s
.
4
.
3
.
3
.
Ste
m
m
i
ng
Ste
m
m
i
n
g
is
a
p
r
o
ce
s
s
o
f
f
i
n
d
in
g
t
h
e
r
o
o
t
w
o
r
d
o
f
th
e
wo
r
d
s
.
T
h
is
is
d
o
n
e
b
y
u
s
i
n
g
th
e
P
o
r
ter
Ste
m
m
er
al
g
o
r
ith
m
.
T
h
is
r
ed
u
ce
s
t
h
e
ti
m
e
ta
k
en
b
y
th
e
a
l
g
o
r
ith
m
i
n
tr
ain
in
g
all
t
h
e
te
n
s
e
o
f
t
h
e
w
o
r
d
i
n
to
eith
er
p
o
s
itiv
e
o
r
n
e
g
ati
v
e.
4
.
3
.
4
.
Co
un
t
v
ec
t
o
rize
r
C
o
u
n
t
v
ec
to
r
izer
h
elp
s
in
co
n
v
er
tin
g
th
e
w
o
r
d
s
in
to
v
ec
to
r
s
w
h
ich
a
m
ac
h
i
n
e
lear
n
i
n
g
al
g
o
r
ith
m
ca
n
ea
s
il
y
u
n
d
er
s
ta
n
d
.
A
p
p
l
y
in
g
c
o
u
n
t
v
ec
to
r
izer
o
n
th
e
d
ata
cr
ea
tes
a
m
atr
ix
o
f
v
ec
to
r
s
f
o
r
t
h
e
w
o
r
d
s
p
r
esen
t
i
n
th
e
d
ataset.
4
.
4
.
Senti
m
ent
cla
s
s
if
ica
t
io
n us
in
g
na
iv
e
ba
y
es
Su
p
er
v
i
s
ed
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
Nai
v
e
B
a
y
es
[
1
9
-
2
1
]
is
ap
p
lied
o
n
th
e
p
r
e
-
p
r
o
ce
s
s
ed
d
ata
t
o
p
r
ed
ict
th
e
s
e
n
ti
m
e
n
t
o
f
t
h
e
co
m
m
en
t
s
.
T
h
e
cla
s
s
i
f
ier
i
s
tr
ai
n
ed
w
it
h
d
ata
an
d
te
s
ted
o
n
t
h
e
u
n
s
ee
n
d
ata
o
r
th
e
test
i
n
g
d
ata.
T
h
e
d
ataset
is
d
i
v
id
ed
in
t
h
e
r
atio
o
f
7
:3
w
it
h
7
0
%
d
ata
f
o
r
tr
ai
n
i
n
g
an
d
3
0
%
f
o
r
test
i
n
g
.
Fro
m
th
e
d
ata
all
th
e
n
eu
tr
al
co
m
m
en
ts
ar
e
r
e
m
o
v
ed
m
ak
in
g
it
a
b
in
ar
y
clas
s
i
f
icatio
n
.
Fo
r
p
r
ed
ictin
g
th
e
co
m
m
e
n
t
as
eith
er
p
o
s
itiv
e
o
r
n
e
g
ativ
e
th
e
p
r
o
b
ab
ilit
y
o
f
p
r
ed
ictin
g
t
h
e
class
h
as
to
b
e
5
0
%
ea
ch
.
I
f
th
e
ac
cu
r
ac
y
o
f
p
r
ed
ictin
g
s
i
n
g
le
clas
s
i
s
m
o
r
e
th
a
n
5
0
%,
SMOT
E
alg
o
r
ith
m
is
ap
p
lied
i
n
o
r
d
er
to
b
alan
ce
th
e
p
r
o
b
ab
ilit
y
o
f
p
r
ed
ictin
g
a
s
en
ti
m
en
t
to
5
0
%.
T
h
e
b
alan
ce
d
d
ataset
is
t
h
en
ap
p
lied
w
it
h
Nai
v
e
B
a
y
e
s
alg
o
r
ith
m
to
p
r
ed
ict
th
e
s
e
n
ti
m
e
n
t
o
f
th
e
co
m
m
e
n
ts
.
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
Naiv
e
B
a
y
e
s
clas
s
i
f
ier
is
c
alcu
lated
u
s
i
n
g
th
e
co
n
f
u
s
io
n
m
a
tr
ix
.
T
h
e
co
n
f
u
s
i
o
n
m
atr
i
x
g
i
v
e
s
th
e
co
u
n
t o
f
th
e
c
o
r
r
ec
tly
a
n
d
in
co
r
r
ec
tl
y
p
r
ed
icted
p
o
s
itiv
e
an
d
n
eg
at
iv
e
co
m
m
e
n
t
s
b
y
t
h
e
clas
s
if
ier
.
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
Hyb
r
id
a
p
p
r
o
a
c
h
:
n
a
ive
b
a
ye
s
a
n
d
s
en
timen
t V
A
DE
R
f
o
r
a
n
a
lyzi
n
g
s
en
timen
t
…
(
C
h
a
ith
r
a
V
D
)
4457
5.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
Sen
ti
m
e
n
t
V
A
DE
R
ap
p
lied
t
o
lab
el
th
e
co
m
m
e
n
ts
w
it
h
t
h
e
r
a
n
g
e
v
al
u
e
o
f
0
.
2
a
n
d
-
0
.
2
r
esu
lt
s
ef
f
ec
tiv
e
l
y
lab
elin
g
t
h
e
co
m
m
en
ts
.
T
ab
le
1
s
h
o
w
s
t
h
e
r
e
s
u
lt
o
f
ap
p
l
y
i
n
g
t
h
e
Se
n
ti
m
e
n
t
V
ADE
R
o
n
th
e
d
ataset.
T
h
e
lab
eled
d
ata
is
co
u
n
ted
f
o
r
t
h
e
n
u
m
b
er
o
f
p
o
s
iti
v
e,
n
eg
at
iv
e
an
d
n
e
u
tr
al
v
al
u
e
s
w
h
ich
is
s
h
o
w
n
i
n
th
e
T
ab
le
1
.
Fig
u
r
e
4
s
h
o
w
s
t
h
e
g
r
ap
h
icall
y
r
ep
r
esen
tatio
n
o
f
Sen
ti
m
en
t
V
A
DE
R
r
esu
lt
w
h
e
n
ap
p
lied
o
n
th
e
d
ataset.
C
o
m
m
e
n
ts
t
h
at
lab
eled
n
eu
tr
al
ar
e
r
e
m
o
v
ed
m
a
k
in
g
a
b
in
ar
y
cla
s
s
i
f
ica
tio
n
f
o
r
p
r
ed
ictin
g
th
e
s
en
ti
m
e
n
t
a
s
eit
h
er
p
o
s
iti
v
e
o
r
n
e
g
ati
v
e.
T
h
e
d
ataset
w
it
h
p
o
s
iti
v
e
an
d
n
e
g
ati
v
e
co
m
m
e
n
t
i
s
u
s
ed
to
tr
ain
th
e
Naiv
e
B
a
y
e
s
clas
s
if
ier
.
Fro
m
T
ab
le
1
i
t
is
clea
r
th
at
t
h
e
p
o
s
itiv
e
co
m
m
e
n
ts
ar
e
m
o
r
e
th
an
th
e
n
eg
a
tiv
e
co
m
m
e
n
t
s
in
th
e
tr
ai
n
i
n
g
d
at
aset.
T
h
e
ac
cu
r
ac
y
o
f
p
r
ed
ictin
g
p
o
s
iti
v
e
co
m
m
en
ts
is
6
2
.
9
9
%
w
h
ich
i
s
m
o
r
e
th
an
5
0
%
ca
u
s
i
n
g
an
i
m
b
ala
n
ce
.
Fo
r
a
b
in
ar
y
c
lass
if
ica
tio
n
th
e
ac
cu
r
ac
y
o
f
p
r
ed
ictin
g
a
cl
ass
h
a
s
to
b
e
5
0
%.
Sin
ce
ac
c
u
r
ac
y
o
f
p
r
ed
ictin
g
p
o
s
itiv
e
co
m
m
e
n
ts
is
6
2
.
9
9
%
cr
ea
tin
g
i
m
b
ala
n
ce
SMOT
E
alg
o
r
ith
m
i
s
u
s
ed
to
r
esto
r
e
th
e
b
alan
ce
.
S
MO
T
E
alg
o
r
ith
m
ap
p
lied
o
n
t
h
e
d
ata
s
et
cr
ea
tes
a
b
ala
n
ce
f
o
r
p
r
ed
ictin
g
t
h
e
s
e
n
ti
m
en
t
class
as e
it
h
er
p
o
s
iti
v
e
o
r
n
eg
a
tiv
e.
T
ab
le
1
.
R
esu
lts
o
f
s
en
ti
m
en
t
VADE
R
S
e
n
t
i
me
n
t
V
a
l
u
e
s
C
o
mm
e
n
t
C
o
u
n
t
P
e
r
c
e
n
t
a
g
e
(
%)
P
o
si
t
i
v
e
(
V
a
l
u
e
1
)
1
6
3
1
3
7
.
2
9
N
e
g
a
t
i
v
e
(
V
a
l
u
e
-
1)
9
5
5
2
1
.
.
8
3
N
e
u
t
r
a
l
(
V
a
l
u
e
0
)
1
7
8
7
4
0
.
8
6
Fig
u
r
e
4
.
Gr
ap
h
ical
r
ep
r
esen
ta
tio
n
o
f
s
e
n
ti
m
e
n
t
V
A
DE
R
r
es
u
lt
Naiv
e
B
a
y
e
s
clas
s
i
f
ier
is
tr
ai
n
ed
w
it
h
t
h
e
co
m
m
e
n
ts
h
a
v
in
g
th
e
lab
els.
T
h
e
clas
s
i
f
ier
is
t
h
en
ap
p
lied
o
n
th
e
test
in
g
d
ataset
a
n
d
ac
cu
r
ac
y
o
f
t
h
e
clas
s
i
f
ier
is
f
o
u
n
d
to
b
e
7
9
.
7
8
%
w
it
h
th
e
F
1
s
co
r
e
o
f
8
3
.
7
2
%.
C
o
n
f
u
s
io
n
m
atr
i
x
also
ca
lled
er
r
o
r
m
atr
ix
v
i
s
u
a
ll
y
s
h
o
w
s
t
h
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
clas
s
i
f
i
er
f
o
r
th
e
test
d
ata
w
h
o
s
e
ac
t
u
al
v
al
u
es
ar
e
k
n
o
w
n
.
I
t
p
r
o
v
id
es
th
e
n
u
m
b
er
o
f
co
r
r
ec
t
an
d
in
co
r
r
ec
t
class
p
r
ed
ictio
n
.
T
ab
le
2
s
h
o
w
s
th
e
co
n
f
u
s
io
n
m
atr
i
x
a
s
s
o
ciate
d
w
it
h
t
h
e
Nai
v
e
B
a
y
es
class
if
ier
ap
p
lies
o
n
t
h
e
te
s
t
d
ata.
Ou
t
o
f
4
1
0
n
eg
at
iv
e
co
m
m
en
t
s
i
n
th
e
te
s
t
d
ataset
3
0
5
w
er
e
co
r
r
ec
tl
y
p
r
ed
icted
as
n
eg
at
iv
e
cla
s
s
a
n
d
1
0
5
w
er
e
p
r
ed
icted
p
o
s
itiv
e
b
u
t
w
a
s
n
eg
at
iv
e.
W
it
h
6
8
8
p
o
s
itiv
e
co
m
m
en
ts
in
th
e
test
d
ata
s
et
5
7
1
w
er
e
p
r
ed
icted
as
p
o
s
iti
v
e
a
n
d
1
1
7
p
r
ed
icted
n
eg
ati
v
e
b
u
t
w
a
s
p
o
s
itiv
e.
T
ab
le
2
.
C
o
n
f
u
s
io
n
m
atr
i
x
P
r
e
d
i
c
t
e
d
N
e
g
a
t
i
v
e
P
r
e
d
i
c
t
e
d
P
o
si
t
i
v
e
A
c
t
u
a
l
N
e
g
a
t
i
v
e
3
0
5
(
T
r
u
e
N
e
g
a
t
i
v
e
)
1
0
5
(
F
a
l
se
P
o
si
t
i
v
e
)
A
c
t
u
a
l
P
o
si
t
i
v
e
1
1
7
(
F
a
l
se
N
e
g
a
t
i
v
e
)
5
7
1
(
T
r
u
e
P
o
si
t
i
v
e
)
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.
9
,
No
.
5
,
Octo
b
er
2
0
1
9
:
4
4
5
2
-
4459
4458
6.
CO
NCLU
SI
O
N
H
y
b
r
id
ap
p
r
o
ac
h
f
o
r
p
er
f
o
r
m
i
n
g
s
en
ti
m
en
t
a
n
al
y
s
i
s
o
f
L
G
G7
m
o
b
ile
u
n
b
o
x
i
n
g
v
id
eo
w
a
s
i
m
p
le
m
en
ted
.
Fro
m
t
h
e
e
x
tr
ac
ted
d
ataset
o
f
L
G
G7
m
o
b
ile
u
n
b
o
x
in
g
v
id
eo
o
n
l
y
t
h
e
co
m
m
e
n
t
s
w
er
e
co
n
ce
n
tr
ated
to
p
er
f
o
r
m
s
en
ti
m
en
t
an
al
y
s
i
s
.
T
h
e
o
b
j
ec
tiv
e
o
f
co
n
s
id
er
in
g
co
m
m
e
n
ts
f
o
r
s
tu
d
y
in
g
s
e
n
ti
m
e
n
t
w
a
s
th
a
t
it
e
x
p
r
ess
es
ac
t
u
al
o
p
in
io
n
t
h
a
n
li
k
es/d
i
s
li
k
es.
R
u
le
b
ased
lex
ico
n
ap
p
r
o
ac
h
Sen
ti
m
en
t
V
A
DE
R
w
as
ap
p
lied
to
lab
el
th
e
co
m
m
e
n
t
s
.
I
n
o
r
d
er
to
p
er
f
o
r
m
b
in
ar
y
cl
ass
i
f
icatio
n
n
e
u
tr
al
co
m
m
e
n
t
s
w
er
e
r
e
m
o
v
ed
an
d
th
e
Nai
v
e
B
a
y
es c
la
s
s
i
f
ier
w
as
tr
ain
ed
w
i
th
7
0
% o
f
t
h
e
d
ata.
T
h
e
class
if
ier
w
a
s
th
e
n
te
s
ted
o
n
th
e
3
0
% u
n
s
ee
n
d
ata
an
d
an
ac
cu
r
ac
y
o
f
7
9
.
7
8
% a
n
d
F1
Sco
r
e
o
f
8
3
.
7
2
%
w
a
s
ac
h
iev
ed
.
T
h
e
co
n
f
u
s
io
n
m
at
r
ix
o
b
tain
ed
b
y
t
h
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
clas
s
i
f
ier
s
h
o
w
s
t
h
at
th
e
clas
s
i
f
ier
h
a
s
p
er
f
o
r
m
ed
s
i
g
n
i
f
ica
n
tl
y
w
e
l
l
i
n
p
r
ed
ictin
g
t
h
e
s
en
ti
m
e
n
t
o
f
t
h
e
co
m
m
en
ts
.
I
t
is
also
o
b
s
er
v
ed
th
at
th
e
le
x
ico
n
ap
p
r
o
ac
h
Sen
ti
m
e
n
t
V
A
DE
R
u
s
ed
f
o
r
th
e
s
o
cial
m
ed
ia
te
x
t h
a
s
a
g
o
o
d
im
p
ac
t o
n
th
e
Na
iv
e
B
a
y
e
s
cla
s
s
i
f
ier
in
p
r
ed
ictin
g
t
h
e
s
e
n
ti
m
en
t.
RE
F
E
R
E
NC
E
S
[1
]
L
iu
B,
“
S
e
n
ti
m
e
n
t
a
n
a
l
y
sis
a
n
d
o
p
in
i
o
n
m
in
in
g
,
”
in
S
y
n
th
e
sis L
e
c
tu
re
s o
n
Hu
ma
n
L
a
n
g
u
a
g
e
T
e
c
h
n
o
lo
g
ies
:
M
o
rg
a
n
&
Cla
y
p
o
o
l
P
u
b
l
ish
e
rs
,
M
a
y
2
0
1
2
.
[2
]
Ha
n
if
Bh
u
iy
a
n
,
Jin
a
t
A
ra
,
Ra
jo
n
Ba
rd
h
a
n
,
e
t
a
l.
,
“
Re
tri
e
v
in
g
Yo
u
T
u
b
e
V
id
e
o
b
y
S
e
n
ti
m
e
n
t
A
n
a
l
y
si
s
o
n
Us
e
r
Co
m
m
e
n
t,
”
in
IEE
E
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
S
ig
n
a
l
a
n
d
Ima
g
e
Pro
c
e
ss
in
g
Ap
p
li
c
a
ti
o
n
s
,
Ku
c
h
in
g
,
M
a
lay
sia
,
S
e
p
2
0
1
7
.
[3
]
F
ik
to
r
Im
a
n
u
e
l
T
a
n
e
sa
b
,
Ir
w
a
n
S
e
m
b
iri
n
g
,
Hin
d
riy
a
n
to
Dw
i
P
u
rn
o
m
o
,
“
S
e
n
ti
m
e
n
t
A
n
a
l
y
sis
M
o
d
e
l
Ba
se
d
On
Yo
u
t
u
b
e
Co
m
m
e
n
t
Us
in
g
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
i
n
e
,
”
in
In
ter
n
a
t
io
n
a
l
J
o
u
r
n
a
l
o
f
Co
mp
u
ter
S
c
ien
c
e
a
n
d
S
o
ft
w
a
re
En
g
i
n
e
e
rin
g
(
IJ
CS
S
E)
,
v
o
l.
6
,
n
o
.
8
,
p
p
.
1
8
0
-
1
8
5
,
2
0
1
7
.
[4
]
Yo
u
n
g
su
b
Ha
n
,
Kw
a
n
g
m
i
Ko
Ki
m
,
“
S
e
n
ti
m
e
n
t
A
n
a
l
y
sis
o
n
S
o
c
ial
M
e
d
ia
Us
in
g
M
o
rp
h
o
lo
g
ica
l
S
e
n
ten
c
e
P
a
tt
e
rn
M
o
d
e
l,
”
in
IEE
E
1
5
th
In
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
S
o
ft
w
a
re
En
g
i
n
e
e
rin
g
Res
e
a
rc
h
,
M
a
n
a
g
e
me
n
t
a
n
d
A
p
p
li
c
a
ti
o
n
s
(
S
ER
A)
,
L
o
n
d
o
n
,
UK
,
Ju
n
2
0
1
7
.
[5
]
S
h
a
n
ta
Ra
n
g
a
sw
a
m
y
,
e
t
a
l.
,
“
M
e
tad
a
ta
Ex
trac
ti
o
n
a
n
d
Clas
sif
ica
ti
o
n
o
f
Yo
u
T
u
b
e
V
id
e
o
s
Us
in
g
S
e
n
ti
m
e
n
t
A
n
a
l
y
si
s,”
in
IEE
E
In
ter
n
a
ti
o
n
a
l
Ca
rn
a
h
a
n
Co
n
fer
e
n
c
e
o
n
S
e
c
u
rity
T
e
c
h
n
o
lo
g
y
(
ICCS
T
)
,
Orla
n
d
o
,
F
L
,
USA
,
Oc
t
2
0
1
6
.
[6
]
V
ip
u
l
Ku
m
a
r
Ch
a
u
h
a
n
,
A
sh
ish
Ba
n
sa
l,
Dr.
A
m
it
a
G
o
e
l,
“
Tw
it
ter
S
e
n
ti
m
e
n
t
A
n
a
l
y
sis
Us
i
n
g
V
a
d
e
r,
”
i
n
In
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
o
f
A
d
v
a
n
c
e
Res
e
a
rc
h
,
Id
e
a
s
a
n
d
I
n
n
o
v
a
ti
o
n
s
i
n
T
e
c
h
n
o
l
o
g
y
(
IJ
AR
IIT
)
,
v
o
l.
4
,
n
o
.
1
,
p
p
.
4
8
5
-
4
8
9
,
2
0
1
8
.
[7
]
M
e
g
h
a
n
a
A
sh
o
k
,
e
t
a
l.
,
“
A
P
e
rso
n
a
li
z
e
d
Re
c
o
m
m
e
n
d
e
r
S
y
ste
m
u
s
in
g
M
a
c
h
in
e
L
e
a
rn
in
g
b
a
se
d
S
e
n
ti
m
e
n
t
A
n
a
l
y
sis
o
v
e
r
S
o
c
ial
Da
ta,”
in
IEE
E
S
tu
d
e
n
ts’
Co
n
fer
e
n
c
e
o
n
E
lec
trica
l,
El
e
c
tro
n
ics
a
n
d
C
o
mp
u
ter
S
c
ien
c
e
,
Bh
o
p
a
l,
I
n
d
ia
,
M
a
r
2
0
1
6
.
[8
]
S
m
it
a
sh
re
e
C
h
o
u
d
h
u
ry
,
Jo
h
n
G
.
Bre
slin
,
“
Us
e
r
S
e
n
ti
m
e
n
t
De
tec
ti
o
n
:
A
Yo
u
T
u
b
e
Us
e
Ca
se
,
”
In
:
2
1
st
Na
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
Arti
fi
c
i
a
l
I
n
telli
g
e
n
c
e
a
n
d
Co
g
n
it
ive
S
c
ien
c
e
,
G
a
lwa
y
,
Ire
lan
d
,
A
u
g
-
S
e
p
2
0
1
0
.
[9
]
Am
a
r
Krish
n
a
,
Jo
se
p
h
Zam
b
re
n
o
,
S
a
n
d
e
e
p
Kris
h
n
a
n
,
“
P
o
lari
ty
T
r
e
n
d
A
n
a
l
y
sis
o
f
P
u
b
li
c
S
e
n
ti
m
e
n
t
o
n
Y
o
u
T
u
b
e
,
”
in
COM
AD
'
1
3
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
1
9
t
h
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
M
a
n
a
g
e
me
n
t
o
f
Da
t
a
,
A
h
m
e
d
a
b
a
d
,
In
d
ia
,
De
c
2
0
1
3
,
p
p
.
1
2
5
-
1
2
8
.
[1
0
]
A
s
a
d
Ullah
Ra
fi
q
Kh
a
n
,
M
a
d
i
h
a
K
h
a
n
,
M
o
h
a
m
m
a
d
Ba
d
ru
d
d
in
K
h
a
n
,
“
Na
iv
e
M
u
lt
i
-
lab
e
l
c
las
sifi
c
a
ti
o
n
o
f
Yo
u
T
u
b
e
c
o
m
m
e
n
ts u
sin
g
c
o
m
p
a
ra
ti
v
e
o
p
in
io
n
m
in
in
g
,
”
Pro
c
e
d
i
a
Co
m
p
u
te
r S
c
ien
c
e
,
El
se
v
ier
.
v
o
l.
8
2
,
p
p
.
57
-
6
4
,
2
0
1
6
.
[1
1
]
Rish
a
n
k
i
Ja
in
,
“
S
e
n
ti
m
e
n
t
A
n
a
l
y
sis
o
n
Y
o
u
T
u
b
e
M
o
v
ie
T
ra
il
e
r
c
o
m
m
e
n
ts
to
d
e
term
in
e
th
e
im
p
a
c
t
o
n
Bo
x
-
Of
fi
c
e
Earn
in
g
,
”
Ok
la
h
o
m
a
S
t
a
te
Un
i
v
e
rs
ity
,
2
0
1
8
.
A
v
a
il
a
b
le
:
h
tt
p
s:/
/www
.
s
a
s.c
o
m
/co
n
ten
t/
d
a
m
/S
A
S
/
su
p
p
o
rt/
e
n
/sa
s
-
g
lo
b
a
l
-
f
o
ru
m
-
p
ro
c
e
e
d
in
g
s/2
0
1
8
/
2
7
1
9
-
2
0
1
8
.
p
d
f
.
[1
2
]
S
iers
d
o
rf
e
r,
S
e
rg
iu
Ch
e
l
a
ru
,
Jo
s
e
S
a
n
P
e
d
ro
,
“
H
o
w
Us
e
f
u
l
a
re
Yo
u
r
Co
m
m
e
n
ts?
A
n
a
l
y
z
in
g
a
n
d
P
re
d
ictin
g
Yo
u
T
u
b
e
C
o
m
m
e
n
ts
a
n
d
Co
m
m
e
n
t
Ra
ti
n
g
s,”
in
Pro
c
e
e
d
in
g
s
o
f
t
h
e
1
7
t
h
I
n
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
W
o
rl
d
W
i
d
e
W
e
b
,
Ra
leig
h
,
No
rth
Ca
ro
li
n
a
,
U
S
A
,
A
p
r
2
0
1
0
,
p
p
.
8
9
1
-
9
0
0
,
2
0
1
0
.
[1
3
]
L
a
k
sh
m
ish
Ka
u
sh
ik
,
A
b
h
ij
e
e
t
S
a
n
g
wa
n
,
Jo
h
n
H.L
.
Ha
n
se
n
,
“
Au
to
m
a
ti
c
S
e
n
ti
m
e
n
t
Ex
trac
ti
o
n
f
ro
m
Yo
u
T
u
b
e
V
id
e
o
s,”
in
IEE
E
W
o
rk
sh
o
p
o
n
Au
to
ma
t
ic
S
p
e
e
c
h
Rec
o
g
n
it
io
n
a
n
d
Un
d
e
rs
ta
n
d
in
g
(
AS
RU)
,
Olo
m
o
u
c
,
Cz
e
c
h
Re
p
u
b
li
c
,
De
c
2
0
1
3
,
p
p
.
2
3
9
–
2
4
4
.
[1
4
]
A
li
a
k
se
iS
e
v
e
r
y
n
,
e
t
a
l.
,
“
Op
in
io
n
M
i
n
i
n
g
o
n
Yo
u
T
u
b
e
,
”
i
n
Pr
o
c
e
e
d
in
g
s
o
f
th
e
5
2
n
d
An
n
u
a
l
M
e
e
ti
n
g
o
f
th
e
Asso
c
ia
ti
o
n
f
o
r Co
m
p
u
t
a
ti
o
n
a
l
L
i
n
g
u
isti
c
s
,
Ba
lt
im
o
re
,
M
a
ry
lan
d
,
USA
,
Ju
n
2
0
1
4
,
p
p.
1
2
5
2
–
1
2
6
1
.
[1
5
]
P
e
ter
D.
T
u
rn
e
y
,
“
T
h
u
m
b
s
Up
o
r
T
h
u
m
b
s
Do
w
n
?
S
e
m
a
n
ti
c
Orie
n
tatio
n
A
p
p
li
e
d
to
Un
s
u
p
e
rv
ise
d
Clas
sifi
c
a
ti
o
n
o
f
Re
v
ie
w
s,”
in
Pro
c
e
e
d
in
g
s
o
f
th
e
4
0
t
h
An
n
u
a
l
M
e
e
ti
n
g
o
f
t
h
e
Asso
c
ia
ti
o
n
f
o
r
Co
mp
u
ta
ti
o
n
a
l
L
i
n
g
u
isti
c
s
(
ACL
),
P
h
il
a
d
e
l
p
h
ia
,
Ju
l
2
0
0
2
,
p
p.
4
1
7
-
4
2
4
.
[1
6
]
Bo
P
a
n
g
a
n
d
L
il
li
a
n
L
e
e
,
S
h
iv
a
k
u
m
a
r
V
a
it
h
y
a
n
a
th
a
n
,
“
T
h
u
m
b
s
u
p
?
S
e
n
ti
m
e
n
t
Clas
sifi
c
a
ti
o
n
u
sin
g
M
a
c
h
i
n
e
L
e
a
rn
in
g
T
e
c
h
n
iq
u
e
s,”
in
Pro
c
e
e
d
in
g
s
o
f
th
e
Co
n
fer
e
n
c
e
o
n
Emp
i
ric
a
l
M
e
th
o
d
s
in
Na
t
u
ra
l
L
a
n
g
u
a
g
e
Pro
c
e
ss
in
g
(
EM
NL
P)
,
P
h
il
a
d
e
lp
h
ia,
J
u
l
2
0
0
2
,
p
p
.
79
–
86.
[1
7
]
Hu
tt
o
C
,
G
il
b
e
rt
E,
“
V
A
DER:
A
P
a
rsim
o
n
io
u
s
Ru
le
-
b
a
se
d
M
o
d
e
l
f
o
r
S
e
n
ti
m
e
n
t
A
n
a
l
y
sis
o
f
S
o
c
ial
M
e
d
ia
T
e
x
t,
”
in
8
th
in
ter
n
a
ti
o
n
a
l
AA
AI
c
o
n
fer
e
n
c
e
o
n
we
b
lo
g
s a
n
d
s
o
c
ia
l
me
d
ia
(
ICW
S
M
)
,
2
0
1
4
.
[1
8
]
Ch
a
u
h
a
n
Vip
u
l
Ku
m
a
r
,
e
t
a
l.
,
“
Tw
it
ter
S
e
n
ti
m
e
n
t
A
n
a
l
y
sis
U
sin
g
V
a
d
e
r,
”
i
n
I.
J
I
d
e
a
s
a
n
d
In
n
o
v
a
ti
o
n
s
i
n
T
e
c
h
n
o
l
o
g
y
,
p
p
.
4
8
5
-
4
8
9
,
2
0
1
8
.
[1
9
]
L
o
p
a
m
u
d
ra
De
y
,
e
t
a
l.
,
“
S
e
n
t
im
e
n
t
A
n
a
l
y
sis
o
f
Re
v
ie
w
Da
t
a
se
ts
Us
in
g
Na
ïv
e
Ba
y
e
s‘
a
n
d
K
-
NN
Clas
sif
ier,”
in
I.
J
.
In
fo
rm
a
t
io
n
E
n
g
i
n
e
e
rin
g
a
n
d
El
e
c
tro
n
ic B
u
si
n
e
ss
,
p
p
5
4
-
62,
Ju
l
2
0
1
6
.
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
Hyb
r
id
a
p
p
r
o
a
c
h
:
n
a
ive
b
a
ye
s
a
n
d
s
en
timen
t V
A
DE
R
f
o
r
a
n
a
lyzi
n
g
s
en
timen
t
…
(
C
h
a
ith
r
a
V
D
)
4459
[2
0
]
No
rm
a
n
J
a
s
m
in
e
,
e
t
a
l.
,
“
A
N
a
iv
e
-
Ba
y
e
s
S
trate
g
y
f
o
r
S
e
n
ti
m
e
n
t
A
n
a
l
y
si
s
on
De
m
o
n
e
ti
z
a
ti
o
n
a
n
d
In
d
ian
B
u
d
g
e
t,
”
in
I.
J
P
u
re
a
n
d
Ap
p
li
e
d
M
a
t
h
e
ma
ti
c
s
,
p
p
.
23
-
3
1
,
2
0
1
7
.
[2
1
]
Bh
a
n
a
p
a
n
d
Ka
w
th
e
k
a
r
,
“
S
e
n
ti
m
e
n
t
A
n
a
l
y
sis
O
f
M
o
b
il
e
Da
tas
e
ts
Us
in
g
Na
ïv
e
Ba
y
e
s
A
l
g
o
rit
h
m
,
”
i
n
I.
J
Ad
v
a
n
c
e
d
Res
e
a
rc
h
in
C
o
mp
u
ter
S
c
ien
c
e
,
p
p
7
8
5
-
7
8
7
,
M
a
r
-
A
p
r
2
0
1
8
.
B
I
O
G
RAP
H
Y
O
F
AUTHO
R
Ch
a
it
h
ra
V
D
is
P
G
S
c
h
o
lar
a
t
C
HRIST
(De
e
m
e
d
to
b
e
Un
iv
e
rsit
y
),
Be
n
g
a
lu
ru
.
S
h
e
re
c
e
iv
e
d
a
B.
S
c
in
M
a
t
h
e
m
a
ti
c
s,
El
e
c
tro
n
ics
a
n
d
Co
m
p
u
ter
S
c
ien
c
e
f
ro
m
J
y
o
ti
Niv
a
s
Co
ll
e
g
e
(
A
u
to
n
o
m
o
u
s) i
n
2
0
1
6
,
Be
n
g
a
lu
r
u
.
He
r
re
se
a
rc
h
in
tere
sts a
re
Big
d
a
ta an
d
Da
ta A
n
a
l
y
ti
c
s.
Evaluation Warning : The document was created with Spire.PDF for Python.