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
.
4
,
A
u
g
u
s
t
201
9
,
p
p
.
2614
~
261
9
I
SS
N:
2088
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v
9
i
4
.
p
p
2
6
1
4
-
2619
2
614
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
Bus
iness
r
eco
mm
enda
tion
b
a
sed
o
n
c
o
lla
bo
ra
tive
f
il
tering
a
nd
f
ea
ture
e
ng
ineeri
ng
–
ap
ro
po
sed
a
ppro
a
ch
P
ra
k
a
s
h P
.
Ro
k
a
de
,
Aru
na
K
u
m
a
ri
D
.
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
&
En
g
in
e
e
rin
g
,
KL
EF
De
e
m
e
d
Un
iv
e
rsity
V
a
d
d
e
sw
a
ra
m
,
In
d
ia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ti
cle
his
to
r
y:
R
ec
eiv
ed
J
an
9
,
2
0
1
8
R
ev
i
s
ed
No
v
2
8
,
2
0
1
8
A
cc
ep
ted
Ma
r
4
,
2
0
1
9
Bu
sin
e
ss
d
e
c
isio
n
s
f
o
r
a
n
y
se
r
v
ice
o
r
p
ro
d
u
c
t
d
e
p
e
n
d
o
n
se
n
ti
m
e
n
ts
b
y
p
e
o
p
le.
W
e
g
e
t
th
e
se
se
n
ti
m
e
n
t
s
o
r
ra
ti
n
g
o
n
so
c
ial
w
e
b
sites
li
k
e
tw
it
ter,
k
a
g
g
le.
T
h
e
m
o
o
d
o
f
p
e
o
p
le
t
o
w
a
rd
s
a
n
y
e
v
e
n
t,
se
rv
ic
e
a
n
d
p
ro
d
u
c
t
a
re
e
x
p
re
ss
e
d
in
th
e
se
s
e
n
ti
m
e
n
ts
o
r
ra
ti
n
g
.
T
h
e
tex
t
o
f
se
n
ti
m
e
n
t
c
o
n
tai
n
s
d
if
fe
re
n
t
li
n
g
u
isti
c
f
e
a
tu
re
s
o
f
s
e
n
ten
c
e
.
A
se
n
ti
m
e
n
t
se
n
ten
c
e
a
ls
o
c
o
n
tai
n
s
o
th
e
r
f
e
a
tu
re
s
w
h
ich
a
re
p
la
y
in
g
a
v
it
a
l
ro
le
in
d
e
c
id
in
g
t
h
e
p
o
larity
o
f
se
n
ti
m
e
n
ts.
I
f
fe
a
tu
re
s
s
e
lec
ti
o
n
is
p
ro
p
e
r
o
n
e
c
a
n
e
x
trac
t
b
e
tt
e
r
se
n
ti
m
e
n
ts
f
o
r
d
e
c
isio
n
m
a
k
in
g
.
A
d
irec
ted
p
re
p
ro
c
e
ss
in
g
w
il
l
f
e
e
d
f
il
tere
d
i
n
p
u
t
to
a
n
y
m
a
c
h
in
e
lea
rn
in
g
a
p
p
r
o
a
c
h
.
F
e
a
t
u
re
b
a
se
d
c
o
ll
a
b
o
ra
ti
v
e
f
il
terin
g
c
a
n
b
e
u
se
d
f
o
r
b
e
tt
e
r
se
n
ti
m
e
n
t
a
n
a
ly
sis.
Be
tt
e
r
u
se
o
f
p
a
rts
o
f
sp
e
e
c
h
(
P
OS)
f
o
ll
o
w
e
d
b
y
g
u
id
e
d
p
re
p
ro
c
e
ss
in
g
a
n
d
e
v
a
lu
a
ti
o
n
w
il
l
m
in
im
i
z
e
e
rro
r
f
o
r
se
n
ti
m
e
n
t
p
o
larity
a
n
d
h
e
n
c
e
th
e
b
e
tt
e
r
re
c
o
m
m
e
n
d
a
ti
o
n
to
t
h
e
u
se
r
f
o
r
b
u
sin
e
ss
a
n
a
ly
ti
c
s c
a
n
b
e
a
tt
a
in
e
d
.
K
ey
w
o
r
d
s
:
C
o
llab
o
r
ativ
e
f
i
lte
r
in
g
Featu
r
e
e
x
tr
ac
tio
n
Ma
ch
i
n
e
l
ea
r
n
i
n
g
Sen
ti
m
e
n
t
a
n
a
l
y
s
is
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
:
P
r
ak
ash
P
.
R
o
k
ad
e,
Dep
ar
t
m
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
&
E
n
g
in
ee
r
i
n
g
,
KL
E
F De
e
m
ed
U
n
iv
er
s
it
y
Vad
d
es
w
ar
a
m
,
Nala
n
d
a,
Gar
i
m
a
n
a
g
ar
i,N
ea
r
Ma
h
ila
Ma
h
av
id
y
ala
y
a,
Ko
p
ar
g
ao
n
,
Dis
t.
-
Ah
m
ed
n
a
g
a
r
,
Ma
h
ar
ash
tr
a,
I
n
d
ia.
E
m
ail:
p
r
ak
a
s
h
r
o
k
ad
e2
0
0
5
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
Sen
ti
m
e
n
t
a
n
al
y
s
is
(
S
A
)
o
f
b
lo
g
s
is
p
la
y
in
g
a
v
ital
r
o
le
f
o
r
b
u
s
i
n
es
s
d
ec
is
io
n
s
to
p
lan
a
g
o
o
d
b
u
s
i
n
ess
s
tr
ate
g
y
.
S
A
is
a
n
ar
t
if
icial
in
telli
g
en
ce
s
tr
ate
g
y
t
h
a
t
q
u
an
tifie
s
t
h
e
s
en
ti
m
en
t
as
p
o
s
itiv
e,
n
e
g
ati
v
e
o
r
n
eu
tr
al
[
1
]
.
Sen
ti
m
en
ts
ar
e
ex
p
r
ess
ed
at
Do
cu
m
en
t
-
le
v
el,
S
en
ten
ce
-
le
v
el,
an
d
A
s
p
ec
t
-
le
v
el
[
2
]
.
SA
h
a
s
m
a
n
y
ap
p
licatio
n
s
in
v
ar
io
u
s
f
ield
s
l
ik
e
r
an
k
i
n
g
p
r
o
d
u
cts,
s
er
v
ice
s
an
d
m
er
c
h
an
t
s
,
p
r
ed
ictin
g
s
h
ar
e
p
r
ice,
p
r
ed
ictin
g
m
o
v
ie
p
o
p
u
lar
it
y
,
r
ec
o
m
m
e
n
d
atio
n
u
s
i
n
g
b
u
s
i
n
es
s
in
tell
ig
en
ce
.
S
A
ai
m
s
to
p
r
o
v
id
e
th
e
r
ig
h
t
k
n
o
w
led
g
e
to
th
e
r
ig
h
t p
er
s
o
n
at
t
h
e
r
ig
h
t ti
m
e
[
3
]
.
C
u
r
r
en
t
al
g
o
r
ith
m
s
ar
e
b
ei
n
g
u
s
ed
f
o
r
a
s
in
g
le
g
r
o
u
p
o
f
u
s
e
r
s
o
r
p
r
o
d
u
cts,
w
h
ich
i
g
n
o
r
e
t
h
e
i
m
p
ac
t
f
o
r
th
e
o
th
er
g
r
o
u
p
s
[
4
]
.
T
h
er
e,
m
a
y
b
e
f
e
w
f
a
k
e
p
o
s
ts
w
h
ic
h
ar
e
p
o
s
ted
b
y
f
a
k
e
u
s
er
s
,
co
m
p
e
tito
r
s
.
So
i
t
i
s
a
ch
alle
n
g
e
to
f
ilter
t
h
e
p
o
s
ts
w
h
ic
h
ar
e
n
o
t
s
p
ec
i
f
ic
to
th
e
f
ea
tu
r
e
o
f
a
p
r
o
d
u
ct
o
r
s
er
v
ice.
T
r
ad
itio
n
al
S
A
alg
o
r
ith
m
s
d
o
n
o
t
co
n
s
id
er
th
e
f
ac
t
t
h
at
as
ti
m
e
p
ass
e
s
,
th
e
v
alu
e
o
f
d
ata
d
ec
r
ea
s
es
f
o
r
m
ak
in
g
d
ec
is
io
n
s
.
T
h
e
d
ata
co
n
s
id
er
ed
f
o
r
s
h
o
r
t
ten
u
r
e
w
ill
d
ec
r
ea
s
e
th
e
q
u
al
it
y
o
f
r
ec
o
m
m
e
n
d
atio
n
o
r
d
ec
is
io
n
.
B
u
g
s
i
n
b
u
g
s
o
u
t p
r
o
b
lem
s
t
ill r
e
m
ai
n
t
h
er
e
[
5
]
.
C
lu
s
ter
i
n
g
f
o
llo
w
ed
b
y
co
llab
o
r
ativ
e
f
il
ter
in
g
h
as
p
r
o
p
o
s
ed
a
r
e
m
ar
k
ab
le
s
o
l
u
tio
n
to
r
es
o
lv
e
th
e
s
e
is
s
u
es
[
5
]
.
I
n
th
e
f
ir
s
t
s
tep
,
we
p
r
ep
r
o
ce
s
s
th
e
i
n
p
u
t
s
e
n
ti
m
en
ts
a
n
d
id
en
ti
f
y
t
h
e
f
e
at
u
r
es
o
f
th
e
p
r
o
d
u
ct
o
r
s
er
v
ice
d
e
s
cr
ib
ed
in
s
e
n
ti
m
en
ts
.
Us
in
g
c
lu
s
ter
in
g
,
lik
el
y
b
l
o
g
s
ar
e
s
elec
ted
a
n
d
t
h
e
n
f
ee
d
to
co
llab
o
r
ativ
e
f
ilter
i
n
g
al
g
o
r
ith
m
to
f
i
ll
m
i
s
s
i
n
g
g
ap
s
o
f
r
atin
g
f
o
r
s
o
m
e
f
ea
t
u
r
es
[
6
]
.
On
e
o
b
j
ec
tiv
e
o
f
th
e
p
r
o
p
o
s
ed
r
ec
o
m
m
e
n
d
atio
n
s
y
s
te
m
i
s
to
en
h
a
n
ce
tr
ad
itio
n
al
co
n
ten
t
-
b
a
s
ed
f
ilter
i
n
g
b
y
b
u
ild
i
n
g
u
s
er
p
r
o
f
ile
b
ased
o
n
th
e
s
tatic
i
n
f
o
r
m
atio
n
t
h
at
r
ep
r
esen
t th
e
l
ik
el
in
e
s
s
o
f
u
s
er
s
to
t
h
e
f
ea
tu
r
e
s
o
f
th
e
i
te
m
s
o
r
s
er
v
ic
e
[
7
]
.
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
B
u
s
in
ess
r
ec
o
mme
n
d
a
tio
n
b
a
s
ed
o
n
c
o
lla
b
o
r
a
tive
filt
erin
g
a
n
d
fea
tu
r
e
en
g
in
ee
r
in
g
.
.
.
.
(
P
r
a
ka
s
h
P
.
R
o
ka
d
e)
2615
2.
L
I
T
E
R
AT
U
RE
SU
RVE
Y
A
r
e
m
ar
k
ab
le
w
o
r
k
i
s
ca
r
r
ied
o
u
t
in
t
h
e
r
ese
ar
c
h
ar
ea
o
f
s
e
n
ti
m
e
n
t
cla
s
s
i
f
icatio
n
.
T
h
e
m
ai
n
f
o
c
u
s
o
f
th
is
w
o
r
k
is
o
n
cla
s
s
i
f
y
i
n
g
la
r
g
er
p
iece
s
o
f
te
x
t,
l
ik
e
r
ev
ie
w
s
o
f
p
r
o
d
u
ct
o
r
ev
e
n
t
[
8
]
.
T
w
ee
t
s
ar
e
d
i
f
f
er
en
t
f
r
o
m
r
ev
ie
w
s
as
t
h
e
y
h
av
e
d
i
f
f
er
en
t
p
u
r
p
o
s
e.
R
e
v
ie
w
s
ar
e
s
u
m
m
ar
y
o
f
au
t
h
o
r
’
s
th
o
u
g
h
t
s
.
T
w
ee
ts
ar
e
li
m
ited
to
1
4
0
ch
ar
ac
ter
s
o
f
tex
t.T
w
ee
ts
r
ep
r
esen
t
g
en
er
al
m
o
o
d
o
f
p
eo
p
le
th
r
o
u
g
h
v
ar
io
u
s
r
ea
ctio
n
s
b
ased
o
n
ex
p
er
ien
ce
o
r
as
an
i
m
p
r
ess
io
n
f
o
r
n
e
w
s
ar
ticle
s
[
9
]
.
Hu
an
d
L
iu
h
a
v
e
g
iv
e
n
a
tech
n
iq
u
e
f
o
r
Featu
r
e
B
ased
Su
m
m
ar
izatio
n
s
y
s
te
m
(
FB
S)
o
f
cu
s
to
m
er
r
ev
ie
w
s
o
f
p
r
o
d
u
cts.
I
t
al
s
o
g
e
n
er
ates
s
e
n
ti
m
e
n
t
b
ased
s
u
m
m
ar
y
a
s
eith
er
p
o
s
itiv
e
o
r
n
eg
ati
v
e
o
p
in
io
n
u
s
i
n
g
ad
j
ec
tiv
e
w
o
r
d
s
in
r
ev
ie
w
s
[
10
]
.
C
h
ao
v
alit
a
n
d
Z
h
o
u
co
m
p
ar
ed
s
u
p
er
v
i
s
ed
an
d
u
n
s
u
p
er
v
i
s
ed
alg
o
r
ith
m
f
o
r
class
i
f
icat
io
n
an
d
g
o
t
8
3
.
5
4
%
o
f
ac
cu
r
ac
y
f
o
r
s
u
p
er
v
i
s
ed
m
et
h
o
d
an
d
7
7
%
o
f
ac
c
u
r
ac
y
f
o
r
u
n
s
u
p
er
v
is
ed
m
et
h
o
d
[
11
]
.
P
an
g
O
Kee
f
e
an
d
Ko
p
r
in
s
k
a
h
a
v
e
g
i
v
en
tech
n
iq
u
e
to
s
elec
t
f
ea
t
u
r
es
u
s
in
g
a
ttrib
u
te
w
eig
h
t
s
a
n
d
ap
p
lied
Nav
ie
B
ay
e
s
a
n
d
SV
M
cla
s
s
i
f
ier
s
f
o
r
class
i
f
icatio
n
o
f
m
o
o
d
s
[
12
,
1
3
]
.
L
in
g
u
is
tic
f
ea
tu
r
es
ar
e
u
s
ed
to
d
etec
t
t
h
e
t
w
itter
s
e
n
ti
m
e
n
t
u
s
i
n
g
h
a
s
h
ta
g
g
ed
d
ata
s
et
(
HA
S
H)
an
d
e
m
o
tico
n
d
ata
s
et.
R
es
u
lts
ar
e
ev
al
u
ated
b
y
u
s
i
n
g
u
n
ig
r
a
m
s
a
n
d
b
ig
r
a
m
s
[
14
,
15
].
T
h
e
s
tu
d
y
b
y
Has
s
an
s
h
o
w
s
th
at
p
ar
ts
o
f
s
p
ee
ch
f
ea
t
u
r
es
ar
e
n
o
t
p
la
y
in
g
g
o
o
d
r
o
le
i
n
s
en
ti
m
e
n
t
an
al
y
s
is
f
o
r
m
icr
o
-
b
lo
g
g
i
n
g
d
o
m
ai
n
.
A
u
th
o
r
in
tr
o
d
u
ce
s
class
i
f
icatio
n
m
et
h
o
d
f
o
r
q
u
er
y
ter
m
s
e
n
ti
m
e
n
t
an
al
y
s
is
.
Her
e
clas
s
i
f
ier
an
d
f
e
atu
r
e
ex
tr
ac
to
r
ar
e
co
n
s
id
er
ed
as
t
w
o
d
if
f
er
e
n
t
co
m
p
o
n
e
n
t
s
[
1
6
]
.
E
ac
h
to
k
en
i
s
ass
i
g
n
ed
a
s
en
ti
m
en
t
s
co
r
e
ca
lle
d
t
o
tal
s
en
ti
m
en
t
in
d
ex
.
Usi
n
g
c
las
s
i
f
icatio
n
alg
o
r
ith
m
th
e
s
e
n
ti
m
e
n
t
s
ar
e
class
i
f
ied
as
p
o
s
itiv
e
o
r
n
e
g
ativ
e
p
o
lar
it
y
s
e
n
ti
m
en
ts
[
1
7
]
.
Po
liti
ca
l
f
u
t
u
r
e
ca
n
b
e
an
al
y
ze
d
r
ea
l
ti
m
e
m
o
n
ito
r
i
n
g
an
d
a
n
al
y
zi
n
g
p
u
b
lic
co
n
v
er
s
a
tio
n
o
n
s
o
cial
s
i
tes
[
1
8
]
.
Featu
r
e
v
e
cto
r
s
a
n
d
tag
g
ed
co
n
te
n
t
o
f
co
r
p
u
s
ca
n
b
e
u
s
ed
to
m
a
k
e
m
o
d
el
b
y
u
s
i
n
g
m
ac
h
i
n
e
lear
n
in
g
ap
p
r
o
ac
h
.
T
h
is
m
o
d
el
is
u
s
ed
to
clas
s
if
y
o
r
ca
teg
o
r
ies
u
n
ta
g
g
ed
co
r
p
u
s
o
f
tex
t
d
o
cu
m
en
t
[
1
9
]
.
Fo
r
lan
g
u
a
g
e
co
n
s
is
te
n
c
y
t
w
itter
is
m
o
r
e
in
f
o
r
m
al.
E
m
o
tico
n
s
ar
e
u
s
ed
ex
p
r
e
s
s
t
h
e
o
p
in
io
n
.
Ma
n
y
t
w
ee
t
s
ar
e
a
m
b
ig
u
o
u
s
an
d
th
e
s
e
ar
e
m
ax
i
m
izi
n
g
th
e
o
p
in
io
n
f
o
r
r
ea
d
er
s
;
b
u
t
d
e
f
lect
th
e
o
p
in
io
n
to
a
m
ac
h
in
e
lear
n
in
g
al
g
o
r
ith
m
[
20
]
.
Sen
t
i
m
e
n
t
clas
s
if
icatio
n
al
g
o
r
ith
m
(
SC
A
)
an
d
S
VM
ar
e
u
s
ed
to
ev
alu
a
te
th
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
ap
p
r
o
ac
h
u
s
ed
ac
c
u
r
ac
y
,
r
ec
all,
p
r
ec
is
io
n
ar
e
s
o
m
e
p
ar
a
m
eter
s
o
n
w
h
ic
h
s
e
n
ti
m
e
n
t a
n
a
l
y
s
is
p
er
f
o
r
m
an
ce
i
s
ev
alu
ated
[
21
].
3.
P
RO
P
O
SE
D
AP
P
RO
ACH
3
.
1
.
M
a
t
h
e
m
a
t
i
c
a
l
m
o
de
l
L
e
t
S
be
t
h
e
m
o
d
e
l
w
h
i
c
h
d
es
cr
ib
e
s
t
h
e
e
x
t
r
a
c
t
i
o
n
,
p
r
e
p
r
o
c
e
s
s
i
n
g
,
l
e
b
l
i
n
g
a
n
d
e
v
a
l
u
a
t
i
n
g
t
h
e
s
e
n
t
i
m
e
n
t
s
.
S=
{
T
w
,
Pt
,
S
l
,
S
e
}
w
h
e
r
e
T
w
=
T
w
itter
s
en
tim
en
ts
.
P
t
=
P
r
e
p
r
o
c
e
s
s
i
n
g
o
f
T
w
ee
ts
Sl
=L
ab
li
n
g
t
h
e
s
e
n
ti
m
e
n
ts
as p
o
s
iti
v
e,
n
e
g
ati
v
e
o
r
n
eu
tr
al
S
l
=
{
P
v
,
Nv
,
Ne
}
P
v
= {P
1
,
P2
,
…
,
Pn
}
=
P
o
s
itiv
e
C
las
s
N
v
=
{
N
1
,N
2
,
…
,
N
n
}=
Neg
ati
v
e
C
lass
N
e
= {
Ne
1
,
N
e
2
,
…
,
N
e
n
}=
Neu
t
r
al
Se
=
S
en
t
i
m
en
t
e
v
a
l
u
a
t
i
on
3
.
2
.
R
e
s
ea
rc
h d
esig
n
A
p
r
o
p
o
s
e
d
r
esear
ch
d
esig
n
f
o
r
s
en
ti
m
e
n
t
an
al
y
s
is
u
s
i
n
g
co
llab
o
r
ativ
e
f
ilter
i
n
g
a
n
d
f
ea
t
u
r
e
en
g
i
n
ee
r
i
n
g
i
s
g
i
v
e
n
in
Fig
u
r
e
1
.
3
.
2
.
1
.
Da
t
a
c
o
llect
io
n
A
co
r
r
ec
t i
n
p
u
t
m
a
y
lead
s
u
s
t
o
g
et
a
co
r
r
ec
t o
u
tp
u
t.
Se
n
ti
m
en
t
d
ata
i
s
a
v
ail
ab
le
o
n
t
w
it
ter
w
eb
s
ite
o
r
f
r
o
m
k
ag
g
le
d
ataset
.
3
.
2
.
2
.
Da
t
a
p
re
pro
ce
s
s
ing
a.
C
ase
n
o
r
m
aliza
tio
n
T
h
e
t
w
ee
t
s
ar
e
av
a
ilab
le
i
n
c
o
m
b
i
n
ed
ca
s
e
th
at
is
it
m
a
y
c
o
n
tain
u
p
p
er
an
d
lo
w
er
ca
s
e
ch
ar
ac
ter
s
.
I
n
ca
s
e
n
o
r
m
a
lizatio
n
t
h
e
en
tir
e
d
o
cu
m
en
t o
r
s
e
n
ten
ce
i
s
co
n
v
er
ted
in
to
lo
w
er
ca
s
e
p
atter
n
g
en
er
all
y
.
b.
T
o
k
en
izatio
n
A
d
o
cu
m
en
t
i
s
s
p
lit
in
to
s
en
ten
ce
s
.
Se
n
te
n
ce
s
m
a
y
b
e
d
iv
id
ed
in
to
w
o
r
d
s
.
B
y
r
e
m
o
v
i
n
g
ce
r
tai
n
ch
ar
ac
ter
s
li
k
e
p
u
n
c
tu
atio
n
m
a
r
k
s
,
r
e
m
ai
n
i
n
g
w
o
r
d
s
ar
e
n
o
w
to
k
en
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2088
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
4
,
A
u
g
u
s
t
201
9
:
2
6
1
4
-
2619
2616
c.
Sto
p
w
ar
d
r
e
m
o
v
al
A
s
et
o
f
s
to
p
w
o
r
d
s
lis
t is
p
r
o
v
id
ed
to
r
em
o
v
e
t
h
e
m
f
r
o
m
s
e
n
ti
m
e
n
ts
.
T
h
e
f
r
eq
u
e
n
tl
y
u
s
ed
s
to
p
w
o
r
d
s
ar
e
‘
a’
,
’
an
’
,
’
th
e
’
,
’
s
h
all
’
,
’
w
ill
’
,
’
t
h
at’
,
’
a
m
’
,
’
i
s
’
,
’
ar
e
’
,
etc.
.
d.
R
o
o
t ste
m
m
i
n
g
I
n
t
h
is
p
r
o
ce
s
s
d
er
iv
ed
w
o
r
d
s
ar
e
r
ed
u
ce
d
to
t
h
eir
s
te
m
.
Fo
r
ex
a
m
p
le
‘
ca
r
ef
u
l
’
,
‘
ca
r
eless
’
,
‘
ca
r
ef
u
ll
y
’
ar
e
r
ed
u
ce
d
to
‘
ca
r
e’
.
e.
T
r
an
s
f
o
r
m
i
n
g
t
h
e
w
o
r
d
s
A
s
et
o
f
d
ef
in
ed
r
u
le
s
ar
e
u
s
ed
to
tr
an
s
f
o
r
m
th
e
w
o
r
d
to
a
s
p
ec
if
ic
f
o
r
m
.
Fo
r
ex
a
m
p
le
a
w
o
r
d
clar
if
ie
s
ca
n
b
e
r
ep
lace
d
b
y
clar
if
y
.
T
h
e
T
ab
le
1
d
escr
ib
es
h
o
w
th
e
w
o
r
d
s
w
i
th
s
u
f
f
i
x
es
ar
e
co
n
v
er
ted
to
eq
u
iv
ale
n
t
s
te
m
a
f
ter
r
e
m
o
v
al
o
f
s
u
f
f
i
x
es
.
T
h
e
w
o
r
d
s
w
i
th
s
u
f
f
ix
e
s
i
n
clu
m
n
1
ar
e
co
n
v
er
ted
to
eq
u
iv
ale
n
t
s
r
te
m
i
n
co
lu
m
n
2
.
T
ab
le
1
.
W
o
r
d
w
it
h
th
eir
eq
u
i
v
alen
t ste
m
W
o
r
d
w
i
t
h
t
h
e
i
r
e
q
u
i
v
a
l
e
n
t
s
t
e
m
W
o
r
d
s
S
t
e
m
Eq
u
a
l
i
t
y
,
E
q
u
a
l
l
y
Eq
u
a
l
En
g
i
n
e
e
r
i
n
g
,
En
g
i
n
e
e
r
,
En
g
i
n
e
e
r
e
d
En
g
i
n
e
e
r
M
a
n
u
a
l
l
y
,
M
a
n
u
a
l
,
M
a
n
M
a
n
f.
R
e
m
o
v
al
o
f
h
an
d
le
s
lik
e
#
etc.
User
s
i
n
clu
d
e
T
w
it
ter
u
s
er
n
a
m
es
i
n
t
h
eir
t
w
ee
t
s
i
n
o
r
d
er
to
d
ir
ec
t
th
eir
m
e
s
s
a
g
es.
A
d
e
f
ac
to
s
tan
d
ar
d
is
to
in
clu
d
e
th
e
@
s
y
m
b
o
l
b
ef
o
r
e
th
e
u
s
er
n
a
m
e
(
e.
g
.
@
alec
m
g
o
)
.
An
eq
u
iv
alen
ce
class
to
k
e
n
(
USERN
A
ME
)
r
ep
lace
s
all
wo
r
d
s
th
at
s
tar
t
w
it
h
@
s
y
m
b
o
l.
Fig
u
r
e
1
.
Flo
w
o
f
p
r
o
p
o
s
ed
s
en
ti
m
en
t a
n
al
y
s
is
ap
p
r
o
ac
h
3
.
2
.
3
.
T
er
m
f
re
qu
e
ncy
co
un
t
a
nd
f
ea
t
ure
ex
t
ra
ct
io
n
Af
ter
d
o
in
g
p
r
ep
r
o
ce
s
s
in
g
a
li
s
t
o
f
ad
j
ec
tiv
es
in
t
h
e
d
ictio
n
a
r
y
i
s
m
a
tch
ed
w
it
h
ev
er
y
r
ea
m
i
n
g
w
o
r
d
in
th
e
d
ata
s
e
t to
f
i
n
d
o
u
t a
d
j
e
ctiv
es a
n
d
th
u
s
t
h
e
f
ea
tu
r
e
s
,
alo
n
g
w
it
h
t
h
ese
ad
j
ec
tiv
es.
3
.
2
.
4
.
F
ea
t
ure
r
a
t
ing
W
e
w
ill
p
r
o
v
id
e
a
li
s
t
o
f
ad
j
e
ctiv
es
a
lo
n
g
w
it
h
a
cr
is
p
v
al
u
e
s
a
y
0
to
5
s
a
y
in
g
t
h
at
0
s
ta
n
d
f
o
r
th
e
w
o
r
s
t,
5
s
ta
n
d
s
f
o
r
th
e
b
est
an
d
s
o
o
n
.
T
h
u
s
w
e
ca
n
p
r
o
v
id
e
th
e
r
ati
n
g
f
o
r
th
e
f
ea
t
u
r
es
if
t
h
e
u
s
er
h
a
s
co
m
m
e
n
ted
o
n
.
T
h
e
u
n
co
m
m
e
n
ted
f
ea
t
u
r
e
w
ill
n
o
t
h
av
e
a
n
y
r
atin
g
,
r
ath
er
it
w
ill
b
e
e
m
p
t
y
r
atin
g
as
s
h
o
w
n
in
T
ab
le
2
.
T
ab
le
2
.
A
d
j
ec
tiv
e
lis
t
w
it
h
r
atin
g
S
r
.
N
o
.
R
a
t
i
n
g
(
C
r
i
s
p
V
a
l
u
e
)
P
r
o
p
o
se
d
a
d
j
e
c
t
i
v
e
l
i
st
1
0
w
o
r
st
,
v
e
r
y
v
e
r
y
b
a
d
2
1
b
a
d
,
n
o
t
g
o
o
d
3
2
Ok
4
3
G
o
o
d
5
4
v
e
r
y
g
o
o
d
6
5
b
e
st
,
e
x
c
e
l
l
e
n
t
,
marv
a
o
l
o
u
s,fa
b
u
l
o
u
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
B
u
s
in
ess
r
ec
o
mme
n
d
a
tio
n
b
a
s
ed
o
n
c
o
lla
b
o
r
a
tive
filt
erin
g
a
n
d
fea
tu
r
e
en
g
in
ee
r
in
g
.
.
.
.
(
P
r
a
ka
s
h
P
.
R
o
ka
d
e)
2617
3
.
2
.
5
.
Clute
ring
t
he
t
o
p k
us
er
s
W
e
n
ee
d
to
f
in
d
s
i
m
ilar
u
s
er
s
b
ased
o
n
th
eir
in
ter
est
f
o
r
th
e
f
ea
t
u
r
es
o
f
p
r
o
d
u
ct
o
r
s
er
v
ice.
H
er
e
w
e
ar
e
in
ter
ested
to
g
et
to
p
k
u
s
er
s
h
a
v
i
n
g
th
e
s
i
m
ilar
tast
e
f
o
r
t
h
eir
i
m
p
r
es
s
io
n
s
.
W
e
ca
n
p
r
o
v
id
e
th
r
es
h
o
ld
v
al
u
e
to
o
p
tim
ize
t
h
e
r
es
u
lt.
W
h
ile
c
lu
s
ter
i
n
g
u
s
in
g
a
n
ap
p
r
o
p
r
iate
clu
s
ter
i
n
g
alg
o
r
it
h
m
,
s
a
y
k
n
e
ar
est n
ei
g
h
b
o
u
r
.
I
n
th
e
T
ab
le
3
s
h
o
w
n
u
s
er
1
,
3
,
4
ar
e
h
av
in
g
s
i
m
i
lar
tast
e
o
f
i
n
ter
ar
est
f
o
r
f
ea
t
u
r
es.
L
ik
e
w
i
s
e
o
u
t
o
f
P
u
s
er
s
to
p
k
u
s
er
s
w
e
ar
e
f
i
n
d
i
n
g
.
T
h
ese
to
p
k
u
s
er
s
ar
e
n
o
w
th
e
r
ep
r
esen
tati
v
es
o
f
th
e
o
r
ig
in
a
l
d
ata
s
e
t
w
e
h
av
e
co
n
s
id
er
ed
as
an
in
p
u
t.
T
h
e
to
p
k
u
s
er
s
h
av
e
n
o
t
r
ated
f
o
r
all
f
e
atu
r
es.
B
u
t
th
es
e
to
p
k
u
s
er
s
h
av
e
co
m
m
e
n
ted
o
n
s
i
m
ilar
f
ea
t
u
r
e
s
v
er
y
clo
s
el
y
.
T
h
e
m
is
s
i
n
g
g
ap
s
o
f
r
ati
n
g
f
o
r
s
o
m
e
f
ea
t
u
r
es
b
y
t
h
ese
k
u
s
er
s
w
il
l b
e
o
v
er
co
m
e
in
co
llab
o
r
ativ
e
f
ilter
in
g
.
T
ab
le
3
.
User
r
atin
g
f
o
r
d
if
f
er
en
t f
ea
tu
r
es
U
se
r
F
e
a
t
u
r
e
F1
F2
F
3
F4
F5
1
5
4
4
3
2
3
1
2
5
3
3
4
4
4
3
4
5
3
5
4
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
P
(
F
i
n
i
t
e
N
o
.
)
3
2
2
5
5
3
.
2
.
6
.
Co
lla
bo
ra
t
iv
e
f
ilte
ring
f
o
r
re
co
mm
e
nd
a
t
io
n
C
o
llab
o
r
atio
n
m
ea
n
s
r
ec
o
m
m
en
d
atio
n
o
f
ite
m
o
r
s
er
v
ice
b
ased
o
n
f
ea
t
u
r
e
r
ated
in
u
s
er
’
s
c
h
o
ice.
Fil
ter
in
g
i
s
s
ep
ar
atio
n
o
f
s
i
m
i
lar
en
titi
e
s
b
ased
o
n
u
s
er
’
s
li
k
es
o
r
d
is
li
k
e
s
.
T
h
e
m
o
ti
v
atio
n
f
o
r
co
llab
o
r
ativ
e
f
ilter
i
n
g
co
m
e
s
f
r
o
m
t
h
e
id
ea
th
at
o
n
e
p
er
s
o
n
ca
n
g
et
b
est
r
ec
o
m
m
e
n
d
atio
n
f
o
r
a
n
y
b
u
s
in
e
s
s
s
a
y
B
,
f
r
o
m
a
n
o
th
er
p
er
s
o
n
w
h
o
h
a
s
th
e
s
a
m
e
i
n
ter
est
in
B
alr
ea
d
y
.
C
o
llab
o
r
ativ
e
f
ilter
i
n
g
m
eth
o
d
s
ar
e
u
s
ed
f
o
r
m
o
n
ito
r
i
n
g
d
ata
s
u
c
h
as
f
i
n
a
n
cial
d
ata,
s
en
ti
m
e
n
t
b
lo
g
s
f
o
r
p
r
o
d
u
ct
o
r
s
er
v
ices,
an
elec
tr
o
n
ic
co
m
m
er
ce
an
d
w
eb
ap
p
licatio
n
s
.
T
ab
le
4
s
h
o
w
n
e
x
p
lai
n
s
w
or
k
i
n
g
o
f
co
lla
b
o
r
ativ
e
f
i
lter
in
g
.
C
o
n
s
id
er
m
o
v
ie
r
ati
n
g
is
g
iv
e
n
f
o
r
5
f
ea
tu
r
es
f
1
to
f
5
.
R
ati
n
g
f
o
r
f
ea
t
u
r
es
ar
e
i
n
th
e
f
o
r
m
o
f
1
to
5
.
1
s
ta
n
d
s
f
o
r
d
is
li
k
e
a
n
d
5
s
tan
d
s
f
o
r
m
o
s
t
lik
e.
T
ab
le
4
.
C
u
s
to
m
er
r
ati
n
g
f
o
r
f
ea
tu
r
es o
f
m
o
v
ie
C
u
s
t
o
me
r
F
e
a
t
u
r
e
F1
F2
F3
F4
F5
1
5
3
4
4
?
2
3
1
2
3
3
3
4
3
4
3
5
4
3
3
1
5
4
5
1
5
5
2
1
Step
1
:
I
g
n
o
r
e
th
e
m
i
s
s
i
n
g
r
ea
d
in
g
co
lu
m
n
an
d
ca
lc
u
late
t
h
e
av
er
ag
e
o
f
r
e
m
ai
n
in
g
r
o
w
s
.
Av
er
ag
e
o
f
r
o
w
1
=(
5
+3
+4
+4
)
/
4
=4
Av
er
ag
e
o
f
r
o
w
2
=(
3
+1
+2
+3
)
/
4
=2
.
2
5
Av
er
ag
e
o
f
r
o
w
3
=(
4
+3
+4
+3
)
/
4
=3
.
5
Av
er
ag
e
o
f
r
o
w
4
=(
3
+3
+1
+5
)
/
4
=3
Av
er
ag
e
o
f
r
o
w
5
=
(
1
+5
+5
+2
)
/4
=3
.
2
5
Step
2
: Ch
o
o
s
e
2
r
o
w
s
w
h
o
s
e
s
i
m
ilar
it
y
i
s
to
b
e
ca
lcu
lated
u
s
in
g
g
iv
e
n
f
o
r
m
u
la.
w
h
er
e,
Si
m
(
C
i,
C
j
)
=Si
m
ilar
it
y
b
et
wee
n
cu
s
to
m
er
i a
n
d
j
.
r
ip
=P
ar
ticu
lar
r
atin
g
o
f
c
u
s
to
m
er
i.
r
j
p
=
P
ar
ticu
lar
r
atin
g
o
f
cu
s
to
m
er
j
r
iav
g
=
Av
er
ag
e
r
ati
n
g
o
f
cu
s
to
m
er
i
r
j
av
g
=A
v
er
ag
e
r
ati
n
g
o
f
cu
s
to
m
er
j
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2088
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
4
,
A
u
g
u
s
t
201
9
:
2
6
1
4
-
2619
2618
B
y
p
u
tti
n
g
t
h
e
v
a
lu
e
s
in
ab
o
v
e
tab
le
in
to
f
o
r
m
u
la,
w
e
w
ill
g
e
t
Si
m
(
C
1
,
C
2
)
=0
.
8
5
Si
m
(
C
1
,
C
3
)
=0
.
7
Si
m
(
C
1
,
C
4
)
=0
Si
m
(
C
1
,
C
5
)
=0
.
7
9
A
b
o
v
e
r
esu
lts
clea
r
l
y
s
tate
t
h
at
cu
s
to
m
er
1
an
d
c
u
s
to
m
er
2
h
as
h
i
g
h
e
s
t
s
i
m
ilar
i
t
y
in
t
h
eir
r
atin
g
s
.
W
e
m
a
y
co
n
cl
u
d
e
th
at,
r
atin
g
f
o
r
f
ea
tu
r
e
5
f
o
r
cu
s
to
m
er
1
w
ill
b
e
s
a
m
e
as
g
i
v
e
n
b
y
cu
s
to
m
er
2
.
So
,
it
w
il
l
b
e
3
f
o
r
cu
s
to
m
er
1
.
Step
3
:
I
n
th
i
s
s
tep
w
e
ca
n
f
in
d
o
u
t
co
l
u
m
n
av
er
a
g
e
f
o
r
all
cu
s
t
o
m
er
s
f
o
r
all
f
ea
tu
r
e
s
.
T
h
e
T
ab
le
5
ex
ap
lain
s
th
e
co
lu
m
n
a
v
er
ag
e
f
o
r
d
if
f
er
en
t
f
ea
t
u
r
es.
As
th
e
co
lu
n
av
er
ag
e
is
b
et
w
ee
n
1
to
5
,
w
e
ca
n
s
et
th
r
es
h
o
ld
as
p
er
o
u
r
d
em
a
n
d
to
co
m
m
en
t o
n
th
e
q
u
alit
y
o
f
a
f
ea
t
u
r
e
f
o
r
an
y
p
r
o
d
u
ct
o
r
s
er
v
ice.
T
ab
le
5
.
C
o
lu
m
n
a
v
er
ag
e
f
o
r
f
ea
tu
r
es
C
u
s
t
o
me
r
F
e
a
t
u
r
e
F1
F2
F3
F4
F5
1
5
3
4
4
3
2
3
1
2
3
3
3
4
3
4
3
5
4
3
3
1
5
4
5
1
5
5
2
1
C
o
l
u
mn
A
v
e
r
a
g
e
3
.
2
3
3
.
2
3
.
4
3
.
2
No
w
o
n
e
ca
n
u
s
e
ab
o
v
e
s
tatis
t
ics
w
it
h
s
o
m
e
t
h
r
es
h
o
ld
f
o
r
e
v
er
y
f
ea
tu
r
e
f
o
r
f
ea
t
u
r
e
b
ased
r
ec
o
m
m
e
n
d
atio
n
o
f
th
e
m
o
v
ie.
4.
CO
NCLUS
I
O
N
W
e
h
av
e
th
o
r
o
u
g
h
l
y
s
tu
d
ie
d
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
u
s
in
g
co
llab
o
r
ativ
e
f
i
lter
in
g
an
d
f
ea
t
u
r
e
en
g
i
n
ee
r
i
n
g
f
o
r
b
u
s
i
n
es
s
r
ec
o
m
m
en
d
atio
n
.
T
h
e
p
r
ep
r
o
ce
s
s
in
g
o
n
i
n
p
u
t
d
ata
s
e
t
w
ill
d
ef
i
n
atel
y
i
m
p
r
o
v
es
th
e
q
u
alit
y
o
f
t
h
e
co
r
p
u
s
.
W
e
w
il
l
g
et
a
p
r
o
p
er
s
et
o
f
f
ea
tu
r
es
u
s
i
n
g
f
r
eq
u
e
n
tl
y
o
cc
u
r
r
ed
ad
jectiv
es.
C
lu
s
ter
in
g
alg
o
r
ith
m
li
k
e
k
n
ea
r
est
n
eig
h
b
o
u
r
w
ill
p
r
o
v
id
e
u
s
to
p
k
s
i
m
ilar
u
s
er
s
w
h
ic
h
ca
n
g
iv
e
t
h
e
r
ec
o
m
m
e
n
d
atio
n
f
o
r
an
y
p
r
o
d
u
ct
o
f
s
er
v
ice
u
s
i
n
g
co
llab
o
r
ativ
e
f
ilter
in
g
.
W
e
ca
n
p
r
o
v
id
e
th
r
esh
o
ld
v
alu
e
f
o
r
in
d
iv
id
u
al
f
ea
t
u
r
e
s
o
th
at
p
r
o
d
u
ct
o
r
s
er
v
ice
ca
n
b
e
r
ec
o
m
m
e
n
d
ed
b
ased
o
n
th
a
t s
p
ec
if
ic
f
ea
tu
r
e
o
n
l
y
.
Fo
r
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
in
th
is
p
ap
er
,
w
e
w
ill
p
r
o
v
id
e
th
r
es
h
o
ld
v
alu
e
to
all
f
ea
t
u
r
es
co
n
s
id
er
in
g
a
s
a
s
y
s
te
m
,
w
h
ich
w
ill
g
iv
e
u
s
t
h
e
r
ec
o
m
m
e
n
d
atio
n
f
o
r
an
y
p
r
o
d
u
ct
o
r
s
er
v
ice.
I
n
th
e
f
u
tu
r
e,
o
n
e
ca
n
d
ir
ec
tl
y
co
n
s
id
er
th
e
Ka
g
g
le
d
ata
s
et,
w
h
ic
h
p
r
o
v
id
es
th
e
r
ati
n
g
o
f
a
n
y
p
r
o
d
u
ct
o
r
s
er
v
ice
b
y
m
n
u
m
b
er
o
f
u
s
er
s
f
o
r
f
n
u
m
b
er
o
f
f
ea
t
u
r
es.
I
t
w
il
l
r
ed
u
ce
th
e
r
o
le
o
f
p
r
ep
r
o
ce
s
s
i
n
g
a
n
d
w
e
ca
n
co
m
p
ar
e
t
h
e
m
ac
h
i
n
e
lear
n
i
n
g
tech
n
iq
u
es
f
o
r
b
etter
o
u
tco
m
e
s
.
RE
F
E
R
E
NC
E
S
[1
]
P
.
S
.
P
riy
a
a
n
d
T
.
V
.S.
Ra
o
,
“
A
n
a
l
y
sin
g
Ev
e
n
t
-
Re
late
d
S
e
n
ti
m
e
n
t
s
o
n
S
o
c
ial
M
e
d
ia
w
it
h
Ne
u
ra
l
N
e
tw
o
rk
s
,
”
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
ol
/i
ss
u
e
:
7
(
3
)
,
p
p
.
1
1
9
-
1
2
4
,
2
0
1
8
.
[2
]
M
.
A
.
F
a
u
z
i,
e
t
a
l
.
,
“
I
m
p
ro
v
in
g
S
e
n
ti
m
e
n
t
A
n
a
l
y
sis
o
f
S
h
o
rt
In
f
o
rm
a
l
In
d
o
n
e
sia
n
P
r
o
d
u
c
t
Re
v
ie
w
s
u
sin
g
S
y
n
o
n
y
m
Ba
se
d
F
e
a
tu
re
Ex
p
a
n
sio
n
,
”
T
EL
KOM
NIK
A
T
e
lec
o
mm
u
n
ica
ti
o
n
Co
mp
u
t
in
g
El
e
c
tro
n
ics
a
n
d
Co
n
tro
l
,
v
ol
/
issu
e
:
16
(
3
)
,
p
p
.
1
3
4
5
-
1
3
5
0
,
2
0
1
8
.
[3
]
Z
.
Z
.
G
a
o
,
e
t
a
l
.
,
“
T
i
m
e
-
W
e
ig
h
te
d
Un
c
e
rtain
Ne
a
re
st
Ne
ig
h
b
o
r
Co
ll
a
b
o
ra
ti
v
e
F
il
terin
g
A
lg
o
rit
h
m
,
”
T
EL
KOM
NIKA
In
d
o
n
e
sia
n
J
o
u
rn
a
l
o
f
El
e
c
trica
l
En
g
i
n
e
e
rin
g
,
v
ol
/i
ss
u
e
:
12
(
8
)
,
p
p
.
6
3
9
3
-
6
4
0
2
,
2
0
1
4
.
[4
]
M
.
W
.
Ch
u
g
h
tai
,
e
t
a
l
.
,
“
G
o
a
l
-
b
a
se
d
H
y
b
rid
F
il
terin
g
f
o
r
Us
e
r
-
to
-
us
e
r
P
e
rso
n
a
li
z
e
d
Re
c
o
m
m
e
n
d
a
ti
o
n
,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
El
e
c
trica
l
a
n
d
Co
m
p
u
ter
En
g
i
n
e
e
rin
g
(
IJ
ECE
)
,
v
ol
/i
ss
u
e
:
3
(
3
)
,
p
p
.
329
-
3
3
6
,
2
0
1
3
.
[5
]
P
.
A
ro
ra
,
e
t
a
l
.
,
“
A
n
A
p
p
ro
a
c
h
f
o
r
Big
Da
ta
to
Ev
o
lv
e
th
e
A
u
sp
icio
u
s
In
f
o
rm
a
ti
o
n
f
ro
m
C
ro
ss
-
Do
m
a
in
s
,
”
In
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
o
f
E
lec
trica
l
a
n
d
C
o
mp
u
ter
En
g
in
e
e
rin
g
(
IJ
ECE
)
,
v
ol
/i
ss
u
e
:
7
(
2
)
,
p
p
.
9
6
7
-
9
7
4
,
2
0
1
7
.
[6
]
M
.
R
.
M
a
’a
rif
a
n
d
A
.
M
u
l
y
a
n
to
,
“
I
m
p
ro
v
in
g
Re
c
o
m
m
e
n
d
e
r
S
y
s
tem
Ba
s
e
d
o
n
Item
’s
S
tru
c
tu
ra
l
In
f
o
rm
a
ti
o
n
i
n
Aff
in
it
y
Ne
t
w
o
rk
,
”
Pro
c
e
e
d
in
g
o
f
I
n
ter
n
a
t
io
n
a
l
Co
n
fer
e
n
c
e
o
n
El
e
c
trica
l
En
g
in
e
e
rin
g
,
C
o
mp
u
t
e
r
S
c
ien
c
e
a
n
d
In
fo
rm
a
t
ics
(
EE
CS
I
2
0
1
4
),
Y
o
g
y
a
k
a
rta
,
I
n
d
o
n
e
sia
,
2
0
1
4
.
[7
]
A
.
El
-
Ko
ra
n
y
a
n
d
S
.
M
.
Kh
a
tab
,
“
On
to
lo
g
y
-
b
a
se
d
S
o
c
ial
Re
c
o
m
m
e
n
d
e
r
S
y
ste
m
,
”
IAE
S
In
ter
n
a
ti
o
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
ol
/i
ss
u
e
:
1
(
3
)
,
p
p
.
1
2
7
-
1
3
8
,
2
0
1
2
.
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
B
u
s
in
ess
r
ec
o
mme
n
d
a
tio
n
b
a
s
ed
o
n
c
o
lla
b
o
r
a
tive
filt
erin
g
a
n
d
fea
tu
r
e
en
g
in
ee
r
in
g
.
.
.
.
(
P
r
a
ka
s
h
P
.
R
o
ka
d
e)
2619
[8
]
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
c
las
sifc
a
ti
o
n
u
si
n
g
m
a
c
h
in
e
lea
rn
i
n
g
tec
h
n
i
q
u
e
s
,
”
Pro
c
e
e
d
in
g
s
o
f
t
h
e
Co
n
fer
e
n
c
e
o
n
Emp
iric
a
l
M
e
th
o
d
s in
Na
tu
ra
l
L
a
n
g
u
a
g
e
Pr
o
c
e
ss
in
g
(
EM
NL
P)
, p
p.
7
9
,
2
0
0
2
.
[9
]
P.
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
tat
io
n
A
p
p
li
e
d
t
o
Un
su
p
e
rv
ise
d
C
las
sif
ic
a
ti
o
n
o
f
Re
v
ie
w
s
,”
Pro
c
e
e
d
in
g
s o
f
t
h
e
Asso
c
ia
ti
o
n
fo
r C
o
mp
u
ta
ti
o
n
a
l
L
in
g
u
isti
c
s
,
2
0
0
2
.
[1
0
]
M
.
Hu
a
n
d
B.
L
iu
,
“
M
in
in
g
a
n
d
S
u
m
m
a
rizin
g
Cu
sto
m
e
r
Re
v
ie
w
s
,
”
Pro
c
e
e
d
in
g
s
o
f
th
e
1
0
th
A
CM
S
IGKD
D,
In
ter
n
a
t
io
n
a
l
C
o
n
f
e
re
n
c
e
o
n
K
n
o
wled
g
e
Disc
o
v
e
ry
a
n
d
Da
t
a
M
in
i
n
g
,
2
0
0
4
.
[1
1
]
P
.
Ch
a
o
v
a
li
t
a
n
d
L
.
Zh
o
u
,
“
M
o
v
ie
Re
v
ie
w
M
in
in
g
:
A
Co
m
p
a
riso
n
b
e
twe
e
n
S
u
p
e
rv
ise
d
a
n
d
Un
su
p
e
rv
ise
d
Clas
sif
ic
a
ti
o
n
A
p
p
ro
a
c
h
e
s,
”
S
y
ste
m
S
c
ien
c
e
s,
HICS
S
'0
5
,
Pro
c
e
e
d
in
g
s
o
f
th
e
3
8
th
A
n
n
u
a
l
Ha
wa
i
i
In
ter
n
a
t
io
n
a
l
C
o
n
fer
e
n
c
e
o
n
IEE
E
,
p
p
.
1
1
2
c
-
1
1
2
c
,
2
0
0
5
.
[1
2
]
T
.
O‟
Ke
e
fe
a
n
d
I.
K
o
p
ri
n
sk
a
,
“
F
e
a
tu
re
S
e
lec
ti
o
n
a
n
d
W
e
ig
h
ti
n
g
in
S
e
n
ti
m
e
n
t
A
n
a
l
y
sis
,
”
Pro
c
e
e
d
in
g
s
o
f
th
e
1
4
t
h
Au
stra
l
a
sia
n
Do
c
u
me
n
t
Co
m
p
u
ti
n
g
S
y
mp
o
si
u
m,
S
y
d
n
e
y
,
A
u
stra
li
a
,
2
0
0
9
.
[1
3
]
A
.
P
a
k
a
n
d
P
.
P
a
ro
u
b
e
k
,
“
Tw
it
ter
a
s
a
c
o
rp
u
s
f
o
r
se
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
,
”
Pro
c
e
e
d
in
g
s
o
f
th
e
S
e
v
e
n
th
I
n
ter
n
a
t
io
n
a
l
C
o
n
fer
e
n
c
e
o
n
L
a
n
g
u
a
g
e
Res
o
u
rc
e
s a
n
d
Eva
lu
a
ti
o
n
(
L
RE
C’1
0
)
,
p
p
.
1
3
2
0
-
1
3
2
6
,
2
0
1
0
.
[1
4
]
E
.
Ko
u
lo
m
p
is,
e
t
a
l
.
,
“
Tw
it
ter
S
e
n
ti
m
e
n
t
A
n
a
l
y
sis
:
T
h
e
G
o
o
d
t
h
e
Ba
d
a
n
d
th
e
OMG
!
,”
Pr
o
c
e
e
d
i
n
g
o
f
t
h
e
Fi
f
t
h
In
ter
n
a
t
io
n
a
l
AA
AI
C
o
n
fer
e
n
c
e
o
n
W
e
b
lo
g
s a
n
d
S
o
c
i
a
l
M
e
d
i
a
,
2
0
1
1
.
[1
5
]
F
.
M
.
F
.
W
o
n
g
,
e
t
a
l
.
,
“
W
h
y
Watc
h
in
g
M
o
v
ie
Tw
e
e
ts
Wo
n
’t
T
e
ll
th
e
W
h
o
le
S
to
ry
?
,
”
A
r
x
iv
p
re
p
rin
t
a
rX
iv
:1
2
0
3
.
4
6
4
2
,
p
p
.
6
,
2
0
1
2
.
[1
6
]
H
.
S
a
if
,
e
t
a
l
.
,
“
S
e
m
a
n
ti
c
S
e
n
ti
m
e
n
t
A
n
a
l
y
si
s
o
f
Tw
it
ter
,”
Pro
c
e
e
d
in
g
s
o
f
th
e
1
1
t
h
In
ter
n
a
ti
o
n
a
l
S
e
ma
n
ti
c
W
e
b
Co
n
fer
e
n
c
e
,
2
0
1
2
.
[1
7
]
G
a
n
n
W
.
J
.
K
.
,
e
t
a
l
.
,
“
Tw
it
ter
a
n
a
l
y
ti
c
s
f
o
r
in
sid
e
r
trad
in
g
f
ra
u
d
d
e
tec
ti
o
n
sy
st
e
m
,”
Pre
s
e
n
ted
a
t
se
c
o
n
d
AS
E
in
ter
n
a
t
io
n
a
l
c
o
n
fer
e
n
c
e
o
n
Bi
g
Da
ta
,
2
0
1
4
.
[1
8
]
Je
n
se
n
M
.
J
.
,
e
t
a
l
.
,
“
In
tr
o
d
u
c
ti
o
n
,”
E.
A
n
d
u
iza
,
M
.
Je
n
se
n
,
&
L
.
Jo
rb
a
(Ed
s.),
“
Dig
it
a
l
m
e
d
ia
a
n
d
p
o
li
ti
c
a
l
e
n
g
a
g
e
m
e
n
t
w
o
rld
w
id
e
:
A
c
o
m
p
a
ra
ti
v
e
stu
d
y
,”
Ne
w
Yo
rk
,
NY
,
Ca
m
b
rid
g
e
Un
iv
e
rsit
y
P
re
ss
,
pp.
1
-
15
,
2
0
1
2
.
[1
9
]
A
.
A
.
Ko
th
a
ri
a
n
d
W
.
D.
P
a
tel,
“
A
No
v
e
l
A
p
p
ro
a
c
h
t
o
w
a
rd
s
C
o
n
tex
t
Ba
se
d
Re
c
o
m
m
e
n
d
a
ti
o
n
s
Us
in
g
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
in
e
M
e
t
h
o
d
o
l
o
g
y
,
”
Pro
c
e
d
ia
C
o
mp
u
ter
S
c
ien
c
e
,
v
o
l.
57,
p
p
.
1
1
7
1
-
1
1
7
8
,
2
0
1
5
.
[2
0
]
A
.
T
rip
a
th
y
,
e
t
a
l
.
,
“
Clas
sif
ica
ti
o
n
o
f
S
e
n
ti
m
e
n
tal
Re
v
iew
s
Us
in
g
M
a
c
h
in
e
L
e
a
rn
in
g
T
e
c
h
n
iq
u
e
s,”
3
r
d
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
Rec
e
n
t
T
re
n
d
s in
Co
mp
u
t
in
g
,
2
0
1
5
.
[2
1
]
V
.
S
a
h
a
y
a
k
,
e
t
a
l
.
,
“
S
e
n
ti
m
e
n
t
A
n
a
l
y
sis o
n
Tw
it
ter Da
ta,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
I
n
n
o
v
a
ti
v
e
Res
e
a
rc
h
in
A
d
v
a
n
c
e
d
En
g
i
n
e
e
rin
g
(
IJ
IRA
E)
,
v
o
l
/i
ss
u
e
:
2
(1
)
,
2
0
1
5
.
B
I
O
G
RAP
H
I
E
S
O
F
AUTH
O
RS
Pra
k
a
s
h
P.
Ro
k
a
d
e
h
a
s
r
e
c
e
iv
e
d
h
isB.E
.
d
e
g
r
e
e
in
Co
m
p
u
ter
f
ro
m
P
u
n
e
Un
iv
e
rsit
y
,
M
a
h
a
ra
sh
tra
;
In
d
ia
in
2
0
0
5
.
He
h
a
s
re
c
e
iv
e
d
h
is
M
.
T
e
c
h
.
d
e
g
re
e
in
Co
m
p
u
ter
En
g
in
e
e
rin
g
f
ro
m
Bh
a
rti
V
id
y
a
p
e
e
rth
,
P
u
n
e
,
M
a
h
a
ra
sh
tra
,
In
d
ia
in
2
0
1
1
a
n
d
p
re
se
n
tl
y
p
u
rsu
i
n
g
h
is
P
h
.
D.
i
n
C
o
m
p
u
ter
S
c
ien
c
e
a
n
d
En
g
in
e
e
rin
g
f
ro
m
Ko
n
e
ru
L
a
k
sh
m
a
iah
Ed
u
c
a
ti
o
n
F
o
u
n
d
a
ti
o
n
,
f
o
rm
e
rl
y
K
L
Un
iv
e
rsit
y
,
V
a
d
d
e
sw
a
r
a
m
,
A
n
d
h
ra
P
ra
d
e
sh
,
In
d
ia.His
re
se
a
rc
h
in
tere
st
in
c
lu
d
e
s
S
e
n
ti
m
e
n
t
A
n
a
l
y
sis,
Op
in
io
n
M
in
i
n
g
,
a
n
d
M
a
c
h
in
e
L
e
a
rn
in
g
.
Ar
u
n
a
K
u
m
a
r
i
D
h
a
s
re
c
e
iv
e
d
h
e
r
P
h
.
D.
d
e
g
re
e
in
Co
m
p
u
ter
S
c
i
e
n
c
e
a
n
d
En
g
in
e
e
rin
g
f
ro
m
th
e
K
L
Un
iv
e
rsit
y
,
V
a
d
d
e
sw
a
ra
m
,
A
n
d
h
ra
P
ra
d
e
sh
,
In
d
ia
.
Cu
rre
n
tl
y
,
S
h
e
is P
ro
f
e
ss
o
r
Ko
n
e
ru
L
a
k
sh
m
a
iah
Ed
u
c
a
ti
o
n
F
o
u
n
d
a
ti
o
n
,
f
o
rm
e
rl
y
K
L
Un
iv
e
rsit
y
.
He
r
tea
c
h
i
n
g
a
n
d
re
se
a
rc
h
a
re
a
s
in
c
lu
d
e
s
D
a
ta
M
in
i
n
g
,
M
a
c
h
in
e
L
e
a
rn
in
g
a
n
d
h
a
s
p
u
b
l
ish
e
d
m
o
re
th
a
n
5
0
p
a
p
e
r
s
in
m
a
n
y
Na
ti
o
n
a
l,
In
tern
a
ti
o
n
a
l
jo
u
r
n
a
ls.
S
h
e
is h
o
n
o
u
re
d
b
y
DST
Yo
u
n
g
S
c
ien
ti
st A
wa
rd
(G
o
v
t.
o
f
In
d
ia).
Evaluation Warning : The document was created with Spire.PDF for Python.