T
E
L
KO
M
N
I
KA
T
e
lec
om
m
u
n
icat
ion
,
Com
p
u
t
i
n
g,
E
lec
t
r
on
ics
an
d
Cont
r
ol
Vol.
18
,
No.
3
,
J
une
2020
,
pp.
13
4
3
~
13
5
3
I
S
S
N:
1693
-
6930
,
a
c
c
r
e
dit
e
d
F
ir
s
t
G
r
a
de
by
Ke
me
nr
is
tekdikti
,
De
c
r
e
e
No:
21/E
/KP
T
/2018
DO
I
:
10.
12928/
T
E
L
KO
M
NI
KA
.
v18i3.
14778
1343
Jou
r
n
al
h
omepage
:
ht
tp:
//
jour
nal.
uad
.
ac
.
id/
index
.
php/T
E
L
K
OM
N
I
K
A
A
p
p
li
c
a
t
io
n
of
r
e
c
om
m
e
n
d
at
i
on
syste
m
w
ith
A
HP m
e
t
h
od
a
n
d
s
e
n
t
i
m
e
n
t
a
n
al
ysi
s
I
r
a
P
r
as
e
t
yan
in
gr
u
m
1
,
Kh
oli
d
F
at
h
on
i
2
,
T
r
i
T
an
ggu
h
Jiwo
P
r
iyan
t
or
o
3
1,
3
Co
mp
u
t
er
an
d
I
n
fo
rma
t
i
c
E
n
g
i
n
eer
i
n
g
D
e
p
art
me
n
t
,
Po
l
i
t
ek
n
i
k
E
l
e
k
t
r
o
n
i
k
a
N
eg
er
i
Su
rab
a
y
a,
In
d
o
n
es
i
a
2
Creat
i
v
e
Mu
l
t
i
me
d
i
a
T
ech
n
o
l
o
g
y
D
e
p
art
m
en
t
,
Po
l
i
t
e
k
n
i
k
E
l
ek
t
ro
n
i
k
a
N
e
g
eri
Su
ra
b
ay
a,
In
d
o
n
es
i
a
Ar
t
icle
I
n
f
o
AB
S
T
RA
CT
A
r
ti
c
le
h
is
tor
y
:
R
e
c
e
ived
S
e
p
1
,
2019
R
e
vis
e
d
J
a
n
19
,
2020
Ac
c
e
pted
F
e
b
25
,
2020
O
v
er
t
i
me,
p
e
o
p
l
e
n
eed
s
are
i
n
creas
i
n
g
.
N
eed
s
t
h
at
mu
s
t
b
e
met
,
o
ft
en
cau
s
e
p
ro
b
l
em
s
i
n
t
h
e
p
eo
p
l
e
i
n
d
et
ermi
n
i
n
g
a
ch
o
i
ce.
Peo
p
l
e
mu
s
t
mak
e
t
h
e
ri
g
h
t
ch
o
i
ce
an
d
acco
r
d
i
n
g
t
o
t
h
e
i
r
n
ee
d
s
.
N
o
t
an
ea
s
y
t
h
i
n
g
fo
r
p
eo
p
l
e
t
o
ma
k
e
t
h
e
s
e
ch
o
i
ce
s
.
T
h
erefo
re,
a
rec
o
mmen
d
at
i
o
n
s
y
s
t
em
i
s
n
ee
d
ed
t
o
s
u
p
p
o
rt
p
eo
p
l
e
i
n
ma
k
i
n
g
d
ec
i
s
i
o
n
s
t
h
a
t
f
i
t
t
h
e
i
r
cri
t
er
i
a.
T
h
i
s
re
s
earch
p
ro
v
i
d
e
s
a
s
y
s
t
em
t
h
at
can
p
ro
v
i
d
e
reco
mmen
d
at
i
o
n
s
f
o
r
d
eci
s
i
o
n
s
u
p
p
o
r
t
p
eo
p
l
e
acco
rd
i
n
g
t
o
t
h
ei
r
cr
i
t
er
i
a,
w
h
i
ch
are
w
e
b
-
b
a
s
ed
.
T
h
e
d
e
ci
s
i
o
n
-
mak
i
n
g
s
y
s
t
em
i
n
t
h
i
s
res
earc
h
u
s
e
s
t
h
e
an
al
y
t
i
cal
h
i
erarch
y
p
ro
ce
s
s
(A
H
P)
met
h
o
d
.
A
H
P
i
s
a
mu
l
t
i
-
cr
i
t
er
i
a
d
ec
i
s
i
o
n
-
mak
i
n
g
me
t
h
o
d
,
w
h
i
c
h
i
n
t
h
i
s
res
earc
h
o
n
e
o
f
t
h
e
cri
t
er
i
a
i
s
u
s
i
n
g
s
en
t
i
me
n
t
an
a
l
y
s
i
s
.
Sen
t
i
me
n
t
a
n
al
y
s
i
s
i
s
t
h
e
p
ro
ces
s
o
f
u
n
d
ers
t
an
d
i
n
g
,
ex
t
rac
t
i
n
g
,
an
d
p
ro
ces
s
i
n
g
t
ex
t
u
a
l
d
a
t
a
t
o
g
et
s
en
t
i
me
n
t
i
n
f
o
rmat
i
o
n
fro
m
an
o
p
i
n
i
o
n
s
e
n
t
e
n
ce.
T
h
e
o
p
i
n
i
o
n
s
en
t
i
me
n
t
v
a
l
u
e
o
f
eac
h
al
t
er
n
at
i
v
e
w
i
l
l
b
e
i
n
c
l
u
d
ed
i
n
t
h
e
A
H
P
ca
l
cu
l
at
i
o
n
t
o
g
et
t
h
e
b
es
t
al
t
ern
a
t
i
v
e
reco
mmen
d
at
i
o
n
s
acco
r
d
i
n
g
t
o
t
h
e
cri
t
e
ri
a
o
f
p
e
o
p
l
e
.
T
h
e
res
u
l
t
o
f
t
h
i
s
res
earch
i
s
t
h
at
t
h
e
s
y
s
t
em
can
p
ro
v
i
d
e
reco
mme
n
d
a
t
i
o
n
s
t
o
p
e
o
p
l
e
o
r
u
s
ers
acco
rd
i
n
g
t
o
t
h
ei
r
cr
i
t
er
i
a
an
d
al
t
er
n
at
i
v
e
s
as
w
el
l
as
p
u
b
l
i
c
o
p
i
n
i
o
n
ab
o
u
t
each
al
t
er
n
at
i
v
e.
K
e
y
w
o
r
d
s
:
Ana
lyt
ica
l
hier
a
r
c
hy
pr
oc
e
s
s
(
AH
P
)
De
c
is
ion
-
making
s
y
s
t
em
S
e
nti
ment
a
na
lys
is
Th
i
s
i
s
a
n
o
p
en
a
c
ces
s
a
r
t
i
c
l
e
u
n
d
e
r
t
h
e
CC
B
Y
-
SA
l
i
ce
n
s
e
.
C
or
r
e
s
pon
din
g
A
u
th
or
:
I
r
a
P
r
a
s
e
tyaningr
um,
De
pa
r
tm
e
nt
of
C
omput
e
r
a
nd
I
nf
or
matic
E
nginee
r
i
ng,
E
lec
tr
onic
E
nginee
r
ing
P
oly
tec
hnic
I
ns
ti
tut
e
o
f
S
u
r
a
ba
ya
,
I
T
S
R
oa
d,
Ke
puti
h,
S
ur
a
ba
ya
60111
,
I
ndone
s
ia.
E
mail:
i
r
a
@pe
ns
.
a
c
.
id
1.
I
NT
RODU
C
T
I
ON
T
he
ne
e
ds
of
pe
ople
incr
e
a
s
e
with
ti
me.
T
o
mee
t
their
ne
e
ds
,
pe
ople
a
r
e
of
ten
f
a
c
e
d
with
c
ondit
ions
whe
r
e
pe
ople
ha
ve
to
make
a
de
c
is
ion.
I
n
m
a
king
a
de
c
is
ion,
the
c
omm
unit
y
ne
e
ds
s
uppor
ter
s
to
make
de
c
is
ions
be
yond
their
a
bil
it
y
to
thi
nk.
On
e
that
c
a
n
be
us
e
d
by
the
c
omm
unit
y
to
s
uppor
t
de
c
is
ion
making
is
a
r
e
c
omm
e
nda
ti
on
.
R
e
c
omm
e
nda
ti
ons
c
a
n
be
s
im
ply
r
e
f
e
r
r
e
d
to
a
s
s
ugge
s
ti
ons
that
e
n
c
our
a
ge
,
jus
ti
f
y,
or
r
e
inf
or
c
e
s
omething
or
s
omeone
.
T
hi
s
r
e
c
omm
e
nda
ti
on
is
ve
r
y
im
por
tant
to
c
onvince
other
s
that
s
omet
hing
or
s
omeone
is
r
ight
.
R
e
c
omm
e
nda
ti
ons
a
r
e
the
r
e
s
ult
of
de
c
is
ion
making.
De
c
is
ion
making
is
a
s
e
lec
ti
on
pr
oc
e
s
s
f
r
om
va
r
ious
a
lt
e
r
na
ti
ve
a
c
ti
ons
that
mi
ght
be
c
hos
e
n
by
mec
ha
nis
m
c
e
r
ta
in,
with
the
a
im
to
p
r
oduc
e
the
be
s
t
de
c
is
ion
[
1]
.
Ano
ther
de
f
ini
t
ion
that
is
a
lm
os
t
the
s
a
me
a
ls
o
s
tate
s
that
de
c
is
ion
making
is
a
n
a
c
t
of
c
hoos
ing
f
r
om
the
a
lt
e
r
na
ti
ve
s
f
a
c
e
d
whic
h
a
r
e
ba
s
e
d
on
f
a
c
ts
a
nd
c
a
r
r
ied
out
thr
ough
a
mathe
matica
l
a
ppr
oa
c
h
that
c
a
n
pr
ovide
the
be
s
t
s
olut
ions
made
by
mana
ge
r
s
[
2]
.
De
c
is
ion
s
upp
or
t
s
ys
tem
is
a
c
omput
e
r
-
ba
s
e
d
inf
or
mation
s
ys
tem
that
do
a
ppr
oa
c
h
to
pr
oduc
e
va
r
ious
a
lt
e
r
na
ti
ve
s
de
c
is
ion
to
he
lp
c
e
r
tain
pa
r
ti
e
s
in
de
a
li
ng
with
pr
oblems
us
ing
da
ta
a
nd
model
[
3]
.
De
c
is
ion
s
uppor
t
s
ys
t
e
ms
a
r
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
3
,
J
une
2020:
13
4
3
-
13
5
3
1344
made
by
a
pplyi
ng
high
c
ompete
nc
y
a
da
ptati
ons
s
o
that
they
c
a
n
be
us
e
d
a
s
a
lt
e
r
na
ti
ve
s
in
making
a
de
c
is
ion
[
4]
.
M
ode
l
that
de
s
c
r
ibes
the
pr
oc
e
s
s
de
c
is
ion
making
c
ons
is
ts
of
f
our
pha
s
e
s
[
5]
,
na
mely:
int
e
ll
igenc
e
,
de
s
ign,
s
e
lec
ti
on
a
nd
im
pleme
ntati
on.
S
o
me
s
tudi
e
s
r
e
late
d
to
de
c
is
ion
s
uppor
t
s
ys
te
ms
in
ge
ne
r
a
l
include
[
6]
that
e
xa
mi
ne
s
a
s
ys
tem
a
n
a
lys
is
de
s
ign
f
or
the
a
s
s
e
s
s
ment
of
voc
a
ti
ona
l
s
tudents
'
s
of
t
s
kil
ls
a
s
pr
os
pe
c
ti
ve
wor
ke
r
s
.
R
e
s
e
a
r
c
h
[
7]
a
ppli
e
s
a
de
c
is
ion
s
uppor
t
s
ys
t
e
m
to
pr
e
dict
s
tock
of
goods
.
M
e
a
nwhile
r
e
s
e
a
r
c
h
[
8]
us
e
s
DSS
f
or
hous
ing
s
e
le
c
ti
on
a
nd
r
e
s
e
a
r
c
h
[
9]
a
ppli
e
s
a
n
e
xpe
r
t
s
ys
tem
to
diagnos
e
tele
c
omm
unica
ti
on
s
e
r
vice
dis
r
upti
ons
on
a
n
I
ndo
ne
s
ian
tele
c
omm
unica
ti
ons
ope
r
a
tor
.
W
he
r
e
a
s
r
e
s
e
a
r
c
h
[
10]
pr
oduc
e
d
a
s
ys
tem
to
de
a
l
with
a
ddit
ional
tea
c
hing
a
nd
lea
r
ning
s
pe
c
ializa
ti
on
in
a
c
c
or
da
nc
e
with
the
a
c
ti
vit
y
of
lea
r
ning
ne
e
de
d.
M
a
ny
r
e
c
omm
e
nda
ti
on
s
ys
tem
s
ha
ve
be
e
n
de
ve
lo
pe
d
to
s
uppor
t
de
c
is
ion
making
in
c
e
r
tain
c
a
s
e
s
.
How
e
ve
r
,
pe
ople
ha
ve
d
if
f
e
r
e
nt
ne
e
ds
o
r
c
a
s
e
s
,
s
o
the
ne
e
ds
in
making
a
de
c
is
ion
a
r
e
dif
f
e
r
e
nt.
M
a
ny
pe
ople's
ne
e
ds
or
c
a
s
e
s
a
r
e
dif
f
e
r
e
nt
f
r
om
the
e
xis
ti
ng
r
e
c
omm
e
nda
ti
on
s
ys
tem
c
a
s
e
s
,
s
o
they
d
o
not
ge
t
r
e
c
omm
e
nda
ti
ons
f
or
de
c
is
ion
s
uppor
t
in
thei
r
c
a
s
e
.
P
e
ople
a
ls
o
ne
e
d
a
r
e
c
omm
e
nda
ti
on
f
r
om
the
t
opic
that
they
ha
ve
de
ter
mi
ne
d
ba
s
e
d
on
the
opini
ons
of
pe
ople.
I
nf
o
r
mation
de
r
ived
f
r
om
pe
ople
(
in
the
f
or
m
o
f
opini
ons
)
c
a
n
a
ls
o
be
us
e
d
a
s
a
de
c
is
ion
s
upp
or
t
with
s
e
ve
r
a
l
o
ther
c
ons
ider
a
ti
ons
,
includ
ing
c
r
it
e
r
ia
a
c
c
or
ding
to
the
pe
r
s
on
him
s
e
lf
.
F
r
om
s
ome
of
the
pr
oblems
de
s
c
r
ibed
a
bove
,
it
is
ve
r
y
s
uppor
ti
ve
to
make
the
a
ppli
c
a
ti
on
of
a
dyna
mi
c
r
e
c
omm
e
nda
ti
on
s
ys
tem
a
s
a
s
uppor
ter
of
pe
ople's
de
c
is
ions
ba
s
e
d
on
their
c
a
s
e
.
I
n
thi
s
s
tudy,
the
a
uthor
s
a
r
e
int
e
r
e
s
ted
in
making
a
dyna
m
ic
r
e
c
omm
e
nda
ti
on
s
ys
tem
a
ppli
c
a
ti
on
us
ing
a
na
lyt
ica
l
hier
a
r
c
hy
pr
oc
e
s
s
(
AH
P
)
c
a
lcula
ti
ons
c
ombi
ne
d
w
it
h
s
e
nti
ment
a
na
lys
is
c
a
lcul
a
ti
ons
.
Ana
lyt
ica
l
hier
a
r
c
hica
l
pr
o
c
e
s
s
(
AH
P
)
is
a
hier
a
r
c
hy
with
the
main
input
in
the
f
or
m
of
human
view
s
,
d
e
ve
loped
by
P
r
o
f
.
T
homas
L
or
ie
S
a
a
ty
f
r
om
W
ha
r
ton
B
us
ines
s
S
c
hool
in
th
e
e
a
r
ly
1970s
a
nd
t
his
method
is
u
s
e
d
to
f
ind
the
or
de
r
or
r
a
nking
of
pr
io
r
it
ies
f
r
om
va
r
ious
a
lt
e
r
na
ti
ve
s
in
p
r
oblem
s
olvi
ng
[
11]
.
S
e
nti
ment
a
na
lys
is
va
lue
wi
l
l
be
one
of
the
c
r
it
e
r
ia
in
AH
P
c
a
lcula
ti
on
to
ge
t
the
be
s
t
r
e
c
omm
e
nda
ti
on
a
c
c
or
ding
to
us
e
r
c
r
it
e
r
ia.
P
r
e
viou
s
s
tudi
e
s
r
e
ga
r
ding
the
r
e
c
omm
e
nda
ti
on
s
ys
tem
a
s
a
r
e
f
e
r
e
nc
e
f
or
thi
s
r
e
s
e
a
r
c
h
include
r
e
s
e
a
r
c
h
f
r
om
khol
id
F
a
thoni
,
I
r
a
P
r
a
s
e
t
ya
ningr
um
a
nd
c
ha
ir
unis
a
.
T
his
r
e
s
e
a
r
c
h
is
a
r
e
s
e
a
r
c
h
on
a
r
e
c
omm
e
nda
ti
on
s
ys
tem
f
or
c
hil
d
c
a
r
e
in
E
a
s
t
S
ur
a
ba
ya
by
us
ing
the
AH
P
method
.
T
he
r
e
s
ult
s
of
th
is
s
tudy
c
a
n
p
r
ovide
in
f
or
mation
a
bo
ut
whe
r
e
da
yc
a
r
e
in
E
a
s
t
S
ur
a
ba
ya
a
nd
pr
ovide
r
e
c
omm
e
nda
ti
ons
r
e
g
a
r
ding
the
s
e
lec
ti
on
of
da
yc
a
r
e
[
12
]
.
S
e
ve
r
a
l
other
s
tudi
e
s
r
e
late
d
to
AH
P
-
ba
s
e
d
de
c
is
ion
s
up
por
t
s
ys
tem
(
DSS
)
include
:
DSS
f
or
e
mpl
oye
e
s
of
t
s
kil
l
c
ompete
nc
y
a
s
s
e
s
s
m
e
nt
[
13]
,
DS
S
f
o
r
de
ter
mi
na
ti
on
of
land
us
e
c
onf
or
m
it
y
(
c
a
s
e
s
tudy
:
S
e
mar
a
ng
R
e
ge
nc
y)
[
14
]
,
D
S
S
f
or
mot
or
c
yc
le
s
e
lec
ti
on
[
15
]
,
DSS
f
or
e
mpl
oye
e
pe
r
f
or
manc
e
a
ppr
a
is
a
l
a
t
S
e
r
a
ng
Hos
pit
a
l
[
16]
,
a
nd
DSS
f
or
the
s
e
lec
ti
on
of
outs
t
a
nding
s
tudents
[
17]
.
W
hil
e
r
e
s
e
a
r
c
h
r
e
late
d
to
s
e
nti
ment
a
na
lys
is
methods
,
na
mely
r
e
s
e
a
r
c
h
c
onduc
ted
by
Ga
ur
a
v
Kuma
r
,
t
h
is
s
tudy
a
im
s
to
r
e
c
omm
e
nd
the
be
s
t
pr
oduc
ts
ba
s
e
d
on
the
opini
ons
e
xpr
e
s
s
e
d
in
c
us
tom
e
r
r
e
view
s
.
R
e
s
e
a
r
c
he
r
s
us
e
methods
s
uc
h
a
s
mac
hine
lea
r
ning
a
nd
s
e
nti
ment
c
las
s
if
ica
ti
on
to
a
na
lyze
c
us
tom
e
r
r
e
view
s
to
s
umm
a
r
ize
them
.
R
e
s
e
a
r
c
he
r
s
a
na
lyze
c
us
tom
e
r
r
e
v
iew
s
f
r
om
va
r
ious
onli
ne
platf
o
r
ms
a
nd
us
e
a
n
e
f
f
e
c
ti
ve
mul
ti
-
c
r
it
e
r
ia
de
c
is
ion
-
making
a
ppr
oa
c
h
to
e
va
luate
a
nd
r
e
c
omm
e
nd
the
mos
t
s
uit
a
ble
pr
od
uc
ts
[
18
]
.
T
he
r
e
s
ult
s
of
thi
s
s
tudy
a
r
e
that
c
a
n
he
lp
us
e
r
s
in
the
de
c
is
ion
-
making
pr
oc
e
s
s
.
R
e
s
e
a
r
c
h
f
r
om
De
s
he
ng
Da
s
h
W
u
,
L
ij
ua
n
Z
he
ng,
a
nd
Da
vid
L
.
Ols
on,
thi
s
s
tudy
de
ve
loped
a
ne
w
s
e
nti
ment
ontol
ogy
to
c
onduc
t
c
ontext
-
s
e
ns
it
ive
s
e
nti
ment
a
na
lys
is
of
onli
ne
opini
on
pos
ts
on
t
he
s
tock
mar
ke
t.
T
he
methodology
in
thi
s
s
tudy
int
e
gr
a
tes
popular
s
e
nti
ment
a
na
lys
i
s
int
o
a
mac
hine
lea
r
ning
a
ppr
oa
c
h
ba
s
e
d
on
s
uppor
t
ve
c
tor
mac
hin
e
s
a
nd
modeling
ge
ne
r
a
l
a
uto
r
e
gr
e
s
s
ive
c
ondit
ional
he
ter
os
c
e
da
s
ti
c
it
y
.
[
19
]
S
e
nti
ment
a
na
lys
is
of
onli
ne
opini
on
pos
ts
c
a
n
f
a
c
il
it
a
te
inves
tor
inves
tm
e
nt
de
c
is
ion
making
a
nd
s
tock
c
ompany
r
is
k
pe
r
c
e
pti
on.
R
e
s
e
a
r
c
h
f
r
o
m
Z
a
oli
Ya
ng,
Gua
nmi
ng
Xiong
,
Z
e
ho
ng
C
a
o,
Yuc
he
n
L
i,
a
nd
L
uc
he
ng
Hua
ng.
T
his
r
e
s
e
a
r
c
h
p
r
opos
e
s
a
method
ba
s
e
d
on
the
c
las
s
if
ica
ti
on
of
s
e
nti
ment
or
ienta
ti
on
a
nd
dis
c
r
e
te
DI
F
W
A
ope
r
a
tor
s
(
DD
I
F
W
A)
f
or
onli
ne
pu
r
c
ha
s
ing
de
c
is
ions
by
c
ons
ider
ing
dyna
mi
c
inf
or
mation
pr
e
f
e
r
e
nc
e
s
[
20
]
.
T
he
r
e
s
ult
s
of
thi
s
s
tudy
a
r
e
a
ble
to
obtain
a
lt
e
r
na
ti
ve
pr
oduc
t
r
a
ti
ngs
a
nd
f
e
a
tur
e
s
to
s
uppor
t
c
ons
umer
pur
c
ha
s
ing
de
c
is
ions
us
ing
int
uit
ioni
s
ti
c
f
uz
z
y
va
luation
f
unc
ti
ons
a
nd
the
"
ve
r
ti
c
a
l
pr
o
jec
ti
on
di
s
tanc
e
"
method.
R
e
s
e
a
r
c
h
f
r
o
m
J
us
ti
n
E
ko
S
oe
li
s
to.
T
e
xt
mi
ning
c
a
n
be
a
ppli
e
d
to
va
r
ious
f
ields
.
One
a
ppli
c
a
ti
on
us
e
s
text
mi
n
ing
in
digi
tal
ne
ws
pa
pe
r
s
to
do
poli
ti
c
a
l
s
e
nti
ment
a
na
lys
is
.
I
n
thi
s
s
e
nti
ment
a
na
lys
is
is
a
ppli
e
d
to
obtain
i
nf
or
mation
f
r
om
ne
ws
a
r
ti
c
les
a
bout
digi
tal
pos
it
ive
or
ne
ga
ti
ve
s
e
nti
ments
a
bout
c
e
r
tain
poli
ti
c
ians
[
21
]
.
R
e
s
e
a
r
c
h
f
r
om
Ana
s
tas
ia
De
wa
nti
P
r
a
tama
P
ut
r
i
whic
h
dis
c
us
s
e
s
the
de
s
ign
of
a
r
e
c
omm
e
nda
ti
on
s
ys
tem
f
or
E
a
s
t
J
a
va
tou
r
is
m
objec
ts
.
Da
ta
c
oll
e
c
ted
in
t
his
s
tudy
c
ons
is
ted
of
E
a
s
t
J
a
va
tour
is
t
a
tt
r
a
c
ti
on
da
ta
a
nd
publi
c
opini
on
a
bout
e
a
c
h
tour
is
t
a
tt
r
a
c
ti
on
.
I
n
th
is
s
tudy
us
ing
S
e
nti
ment
Ana
lys
is
f
or
the
pr
oc
e
s
s
of
e
va
l
ua
ti
ng
publi
c
opini
on
a
bout
E
a
s
t
J
a
va
tour
is
m
objec
ts
on
s
oc
ial
media
.
T
his
r
e
s
e
a
r
c
h
pr
opos
e
s
a
ne
w
mod
e
li
ng
by
pr
ov
idi
ng
r
e
c
omm
e
nda
ti
ons
on
tour
is
m
in
E
a
s
t
J
a
va
f
r
om
e
xis
ti
ng
c
omm
e
nts
to
be
a
ble
to
he
lp
tour
is
ts
f
ind
inf
o
r
mation
a
bout
a
tt
r
a
c
ti
ons
in
E
a
s
t
J
a
va
that
a
r
e
e
xpe
c
ted
to
r
e
a
li
z
e
the
S
a
pta
P
e
s
ona
W
is
a
ta
in
E
a
s
t
J
a
va
[
22
]
.
S
ome
other
s
tudi
e
s
us
ing
s
e
nti
ment
a
na
lys
is
a
r
e
s
e
nti
ment
a
na
lys
is
o
f
tele
vis
ion
s
hows
ba
s
e
d
o
n
o
p
i
n
i
o
n
t
h
e
p
u
b
l
i
c
u
s
e
s
t
h
e
k
-
n
e
a
r
e
s
t
n
e
i
g
h
b
o
r
m
e
t
h
o
d
[
2
3
]
,
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
A
ppli
c
ati
on
of
r
e
c
omm
e
ndati
on
s
y
s
tem
w
it
h
A
HP
me
thod
and
s
e
nti
me
nt
analys
is
(
I
r
a
P
r
as
e
tyaningr
um
)
1345
s
e
nti
ment
a
na
lys
is
s
ys
tem
of
c
us
tom
e
r
c
omm
e
nts
of
the
xyz
hotel
c
us
tom
e
r
us
ing
the
na
ive
ba
ye
s
c
las
s
if
ier
method
[
24]
,
a
nd
hot
e
l
r
e
view
s
e
nti
ment
a
na
lys
is
us
ing
a
pa
r
ti
c
le
s
wa
r
m
opti
mi
z
a
ti
on
s
uppor
t
a
lgor
it
hm
[
25]
.
T
his
r
e
s
e
a
r
c
h
p
r
ovides
a
s
ys
tem
that
c
a
n
p
r
ovide
r
e
c
omm
e
nda
ti
ons
f
or
de
c
is
ion
s
uppor
t
pe
ople
a
c
c
o
r
ding
to
their
c
r
i
ter
ia,
whic
h
a
r
e
we
b
-
ba
s
e
d.
T
he
de
c
is
ion
-
making
s
ys
t
e
m
in
thi
s
r
e
s
e
a
r
c
h
us
e
s
the
a
na
lyt
ica
l
hier
a
r
c
hy
pr
oc
e
s
s
(
AH
P
)
method.
AH
P
is
a
mul
t
i
-
c
r
it
e
r
ia
d
e
c
is
ion
-
making
method,
whic
h
in
thi
s
r
e
s
e
a
r
c
h
o
ne
of
the
c
r
it
e
r
ia
is
us
ing
s
e
nti
ment
a
na
lys
is
.
2.
RE
S
E
AR
CH
M
E
T
HO
D
T
he
s
ys
tem
pr
oc
e
s
s
f
low
in
th
is
s
tudy
is
s
how
n
in
F
igur
e
1.
F
r
om
F
igur
e
1.
us
e
r
s
c
a
n
input
c
r
it
e
r
ia
a
nd
a
lt
e
r
na
ti
ve
da
ta
a
nd
pa
ir
wis
e
c
ompar
is
on
va
lues
a
c
c
or
ding
to
their
c
a
s
e
to
ge
t
r
e
c
omm
e
nda
ti
ons
.
Admin
ha
s
the
a
uthor
it
y
to
pr
oc
e
s
s
the
s
ys
tem
da
taba
s
e
.
S
tep
1
.
AH
P
d
e
s
ign
Af
ter
us
e
r
s
de
c
ide
goa
l,
we
c
a
n
go
to
de
s
ign
AH
P
:
a)
De
ter
mi
ne
the
AH
P
Va
r
iable
:
in
the
AH
P
c
a
lcula
ti
on
pr
oc
e
s
s
,
ther
e
a
r
e
two
im
por
tant
va
r
iable
s
that
mus
t
be
pr
e
s
e
nt,
na
mely
c
r
it
e
r
ia
a
nd
a
lt
e
r
na
ti
ve
s
.
C
r
it
e
r
ia
va
r
iable
is
c
a
ndidate
c
r
it
e
r
ia
e
xpe
c
ted
by
us
e
r
.
Alter
na
ti
ve
va
r
ia
bles
a
r
e
a
lt
e
r
na
ti
ve
s
that
will
be
c
a
lcula
ted
in
the
AH
P
pr
oc
e
s
s
to
ge
t
the
be
s
t
a
lt
e
r
na
ti
ve
a
s
a
r
e
c
omm
e
nda
ti
on.
b)
Hie
r
a
r
c
hy
d
e
s
ign:
thi
s
hier
a
r
c
hy
of
AH
P
c
ontains
c
r
it
e
r
ia
a
nd
a
lt
e
r
na
ti
ve
s
that
will
be
take
n
in
to
a
c
c
ount
in
the
AH
P
p
r
oc
e
s
s
.
I
n
thi
s
s
tudy
,
s
e
nti
ment
a
n
a
lys
is
be
c
omes
a
manda
tor
y
c
r
it
e
r
ion
that
mus
t
be
pr
e
s
e
nt
in
AH
P
c
a
lcula
ti
ons
.
T
his
s
e
nti
ment
a
na
lys
is
is
the
va
lue
of
s
e
nti
ment
a
na
lys
is
of
publi
c
opini
on
on
s
oc
ial
media
T
witt
e
r
a
bout
e
a
c
h
a
lt
e
r
na
ti
ve
.
A
f
ter
us
e
r
s
de
c
ide
goa
l,
than
they
mus
t
de
c
ide
c
r
it
e
r
ia
a
nd
a
lt
e
r
na
ti
ve
.
T
he
hier
a
r
c
hy
in
the
AH
P
p
r
oc
e
s
s
in
thi
s
s
tudy
is
a
s
F
igu
r
e
2.
F
igur
e
1.
De
s
ign
s
ys
tem
F
igur
e
2.
Hie
r
a
r
c
hy
AH
P
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
3
,
J
une
2020:
13
4
3
-
13
5
3
1346
S
tep
2
.
Opinion
s
c
r
a
ping
f
r
om
T
witt
e
r
T
his
r
e
s
e
a
r
c
h
uti
li
z
e
s
T
witt
e
r
's
ope
n
API
to
ge
t
publi
c
opini
on
a
bout
e
a
c
h
a
lt
e
r
na
ti
ve
e
nter
e
d
by
the
us
e
r
.
W
it
h
the
T
wit
ter
API
,
the
s
ys
tem
c
a
n
take
p
*
ubli
c
opini
on
a
bout
a
lt
e
r
na
ti
ve
s
ba
s
e
d
on
a
ke
ywor
d.
Ke
ywor
d
to
ge
t
a
n
opini
on
f
r
om
T
witt
e
r
in
the
f
or
m
of
a
n
a
lt
e
r
na
ti
ve
na
me.
E
xa
mpl
e
s
of
opini
on
s
c
r
a
ping
with
the
ke
ywor
d
"
KFC
M
ulyos
a
r
i"
us
ing
the
T
wit
ter
API
a
r
e
a
s
F
igu
r
e
3.
F
igur
e
3.
Opinion
s
c
r
a
ping
S
tep
3
.
S
e
nti
ment
a
na
lys
is
T
he
r
e
a
r
e
f
ive
s
tage
s
in
the
text
mi
ning
pr
oc
e
s
s
,
na
mely
tokeniz
ing,
f
il
ter
ing
,
s
temmi
ng,
tagging
,
a
nd
a
na
lyzing
.
I
n
thi
s
s
tudy,
the
a
na
lyzing
pr
oc
e
s
s
is
not
ne
e
de
d
be
c
a
us
e
the
text
mi
ning
pr
oc
e
s
s
is
o
nly
us
e
d
to
s
e
a
r
c
h
f
or
im
por
tant
wor
ds
that
c
a
n
r
e
pr
e
s
e
nt
th
e
text.
a)
T
oke
nizing
is
the
pr
oc
e
s
s
of
c
utt
ing
text
int
o
c
oll
e
c
ti
ons
of
wor
ds
.
T
he
pr
oc
e
s
s
of
c
utt
ing
text
is
d
one
ba
s
e
d
on
s
pa
c
e
s
in
e
a
c
h
wor
d.
b)
F
il
ter
ing
is
the
pr
oc
e
s
s
of
r
e
movi
ng
unne
c
e
s
s
a
r
y
wor
ds
or
s
topl
is
ts
.
W
o
r
ds
not
r
e
p
r
e
s
e
nti
ng
c
a
n
be
in
the
f
o
r
m
of
c
onjunctions
,
a
uxil
iar
y
wor
ds
,
c
onjunctions
,
e
tc.
c)
S
temmi
ng
is
the
p
r
oc
e
s
s
of
r
e
movi
ng
a
f
f
ixes
to
a
wor
d
s
o
that
the
c
ontent
c
ontains
ba
s
ic
wor
ds
that
c
a
n
r
e
pr
e
s
e
nt
the
c
ontent.
d)
T
a
gging
is
the
pr
oc
e
s
s
of
jus
ti
f
ying
wor
ds
that
a
r
e
not
pr
ope
r
ly
wr
it
ten.
I
n
a
ddit
ion,
the
tagging
pr
oc
e
s
s
is
a
ls
o
us
e
d
to
r
e
plac
e
non
-
s
tanda
r
d
wor
ds
int
o
s
ta
nda
r
d.
e)
Ke
ywor
ds
a
r
e
im
por
tant
wor
ds
that
c
a
n
r
e
pr
e
s
e
nt
the
text,
s
o
dur
ing
the
opini
on
a
na
lys
is
pr
oc
e
s
s
,
thes
e
ke
ywor
ds
will
be
us
e
d
a
s
a
r
e
f
e
r
e
nc
e
.
Af
ter
opini
ons
a
r
e
pr
oc
e
s
s
e
d
in
text
mi
ning
,
they
c
a
n
be
pr
oc
e
s
s
e
d
in
s
e
nti
ment
a
na
lys
i
s
.
S
e
nti
ment
a
na
lys
is
r
e
f
e
r
s
to
the
c
las
s
if
ica
ti
on
of
opini
ons
int
o
3
c
las
s
e
s
,
na
mely
pos
it
ive,
ne
ga
ti
ve
,
a
nd
ne
ut
r
a
l.
T
o
do
s
e
nti
ment
a
na
lys
is
,
it
is
ne
c
e
s
s
a
r
y
to
ha
ve
wor
d
a
nd
r
ule
de
gr
e
e
da
ta.
T
he
de
gr
e
e
of
the
wo
r
d
is
us
e
d
a
s
a
r
e
f
e
r
e
nc
e
f
o
r
wor
d
s
e
nti
ment
a
nd
the
r
ule
us
e
d
a
s
a
s
e
nti
ment
c
a
lcula
ti
on
tec
hnique.
a)
De
gr
e
e
of
wor
d
t
he
de
gr
e
e
of
the
wor
d
is
us
e
d
to
give
va
lue
to
e
a
c
h
wor
d.
T
his
va
lue
is
in
the
f
or
m
of
number
s
1,
-
1,
a
nd
0
whe
r
e
1
is
pos
it
ive
,
-
1
is
ne
ga
ti
ve
,
a
nd
0
is
ne
utr
a
l
.
b)
R
ule
is
us
e
d
to
pr
ovide
r
ules
f
or
c
omm
e
nti
ng
s
e
nti
ment
a
s
s
e
s
s
ment.
T
his
pr
oc
e
s
s
doe
s
not
us
e
a
s
pe
c
ial
a
lgor
it
hm,
but
r
a
ther
a
n
im
pr
e
s
s
ion
tec
hnique.
I
mpr
e
s
s
ion
tec
hniques
a
r
e
mor
e
inclined
to
a
na
lyze
the
wor
ding
of
a
s
e
ntenc
e
.
T
his
tec
hnique
a
na
lyze
s
the
loca
ti
on
of
a
djec
ti
ve
s
,
ve
r
bs
,
a
nd
pr
o
pos
it
ions
in
a
s
e
ntenc
e
.
P
r
opos
it
ions
a
r
e
s
tate
men
ts
a
bout
thi
ngs
that
c
a
n
be
judged
r
ight
or
wr
ong,
f
or
e
xa
m
ple
not.
T
he
s
e
r
ules
include
:
a)
S
ingl
e
a
djec
ti
ve
:
i
f
a
s
e
ntenc
e
only
c
ons
is
ts
of
a
dj
e
c
ti
ve
s
,
without
a
ny
ve
r
bs
a
nd
pr
opos
it
ions
a
r
oun
d
it
,
then
the
s
e
ntenc
e
va
lue
f
oll
ows
the
a
djec
ti
ve
.
b)
S
ingl
e
v
e
r
b
:
i
f
a
s
e
ntenc
e
c
ons
is
ts
only
of
ve
r
b
s
,
without
a
djec
ti
ve
s
a
nd
pr
opos
it
ions
a
r
ound
th
e
m,
the
s
e
ntenc
e
va
lue
f
oll
o
ws
the
ve
r
b
.
c)
P
r
opos
it
ion
be
f
or
e
a
djec
ti
ve
:
i
f
ther
e
is
a
pr
opos
it
ion
be
f
or
e
the
a
djec
ti
ve
in
a
s
e
ntenc
e
,
the
s
e
nti
m
e
nt
va
lue
in
the
s
e
ntenc
e
f
oll
ows
the
XO
R
logi
c
r
ules
E
xa
mpl
e
:
“
s
ay
a
ti
dak
mar
ah
(
i
a
m
not
a
ngr
y
)
”
“
T
idak
(
No)
”
is
a
ne
ga
ti
ve
pr
opos
it
ion
a
nd
a
n
ge
r
is
a
ne
ga
ti
ve
a
djec
ti
ve
,
s
o
the
s
e
nti
ment
va
lue
in
the
s
e
ntenc
e
is
1
(
pos
it
ive)
.
d)
P
r
opos
it
ion
be
f
o
r
e
Ve
r
b
:
a
s
in
the
pr
e
vious
r
ule,
i
f
ther
e
is
a
p
r
opos
it
ion
be
f
o
r
e
the
ve
r
b
in
a
s
e
nten
c
e
,
the
s
e
nti
ment
va
lue
in
the
s
e
ntenc
e
f
oll
ows
the
XO
R
logi
c
r
u
les
.
E
xa
mpl
e
:
“
Say
a
ti
dak
me
nolak
me
mbuk
a
r
e
s
to
r
an
bar
u
(
I
do
not
r
e
f
us
e
to
ope
n
ne
w
r
e
s
taur
a
nt)
”
(
va
l
ue
=
-
1)
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
A
ppli
c
ati
on
of
r
e
c
omm
e
ndati
on
s
y
s
tem
w
it
h
A
HP
me
thod
and
s
e
nti
me
nt
analys
is
(
I
r
a
P
r
as
e
tyaningr
um
)
1347
e)
Ve
r
b
be
f
or
e
Adje
c
ti
ve
:
b
e
s
ides
pr
opos
it
ions
,
a
djec
ti
ve
s
c
a
n
a
ls
o
c
ha
nge
the
s
e
nti
ment
va
lue
of
ve
r
bs
.
As
s
e
s
s
ment
of
s
e
nti
ment
f
r
om
the
mee
ti
ng
be
twe
e
n
the
ve
r
bs
be
f
or
e
the
a
djec
ti
ve
a
ppa
r
e
ntl
y
doe
s
not
e
xis
t
in
the
logi
c
of
logi
c
,
bu
t
the
oppos
it
e
of
XO
R
.
E
xa
mpl
e
“
Dia
paham
de
ngan
baik
(
He
/s
he
unde
r
s
tand
we
ll
)
”
(
va
lue
=
1)
f)
P
r
opos
it
ion
+
Ve
r
b
+
Adje
c
ti
ve
:
o
f
ten
pr
opos
it
io
ns
,
ve
r
bs
,
a
nd
a
djec
ti
ve
s
a
ppe
a
r
togethe
r
,
s
o
ther
e
is
a
ne
e
d
f
o
r
r
ules
to
c
a
ptur
e
the
s
e
ntenc
e
.
T
he
pr
oc
e
s
s
of
c
a
lcula
ti
ng
s
e
nti
ment
va
lue
s
tar
ts
f
r
om
c
a
lcula
ti
ng
the
c
ombi
na
ti
on
of
ve
r
bs
a
nd
a
djec
ti
ve
s
,
that
is
,
in
r
ule
c
.
E
xa
mpl
e
:
“
Dia
t
idak
me
maham
i
de
ngan
baik
(
He
/
s
he
did
not
unde
r
s
tand
we
ll
)
”
(
va
lue
=
-
1)
g)
All
s
ymbol
s
in
c
omm
e
nts
r
e
moved
,
e
xc
e
pt
P
e
r
iods
a
nd
C
omm
a
s
:
a
ll
c
ha
r
a
c
ter
s
or
s
ymbol
s
that
a
r
e
not
ne
e
de
d
will
be
de
lete
d
e
xc
e
pt
f
o
r
pe
r
iods
a
nd
c
om
mas
.
h)
S
pe
ll
ing
I
ndone
s
ian
W
or
ds
:
e
ve
r
y
ke
ywor
d
that
h
a
s
be
e
n
obtaine
d
mus
t
be
matc
he
d
with
a
n
I
ndone
s
ian
dictionar
y
to
c
o
r
r
e
c
t
wr
i
ti
ng
e
r
r
or
s
.
i)
P
e
r
iods
a
nd
c
omm
a
s
br
e
a
k
s
e
ntenc
e
:
i
f
ther
e
i
s
a
c
omm
a
or
pe
r
iod
s
ymbol
in
a
c
omm
e
nt,
t
he
n
the
c
omm
e
nt
is
c
ons
ider
e
d
to
be
a
f
e
w
s
e
ntenc
e
s
a
nd
s
e
nte
nc
e
s
a
r
e
c
a
lcula
ted
f
or
e
a
c
h
s
e
nten
c
e
.
T
he
c
omm
e
nt
va
lue
is
obtaine
d
f
r
om
the
s
um
of
th
e
s
e
nti
ment
s
e
ntenc
e
s
.
j)
S
ome
E
ngli
s
h
wor
ds
e
nte
r
e
d
in
the
da
taba
s
e
:
r
e
gional
langua
ge
s
will
be
a
utom
a
ti
c
a
ll
y
ignor
e
d,
but
s
ome
wor
ds
in
E
ngl
is
h
c
a
n
be
de
tec
ted
s
uc
h
a
s
:
good,
gr
e
a
t,
nice
,
wonde
r
f
ul,
be
a
uti
f
ul
,
a
nd
s
e
ve
r
a
l
other
wor
ds
in
E
ngli
s
h.
S
tep
4
.
T
he
opini
on
s
e
nti
ment
va
lue
of
e
a
c
h
a
lt
e
r
na
t
ive
will
be
include
d
in
the
AH
P
c
a
lcula
ti
on
to
ge
t
the
be
s
t
a
lt
e
r
na
ti
ve
r
e
c
omm
e
nda
ti
ons
a
c
c
or
ding
to
the
c
r
it
e
r
ia
of
pe
ople
.
C
ompar
a
ti
ve
J
udge
ment
is
the
c
or
e
of
AH
P
,
be
c
a
us
e
it
will
in
f
luenc
e
the
pr
ior
it
y
of
th
e
e
leme
nts
.
I
n
a
s
s
e
s
s
ing
the
r
e
lative
im
por
tanc
e
of
the
two
e
leme
nts
a
r
e
c
ipr
oc
a
l
a
xiom
mea
ns
that
if
e
leme
nt
i
is
va
lued
twice
a
s
im
po
r
tant
a
s
j,
then
e
leme
nt
j
mus
t
be
e
qua
l
to
½
ti
mes
a
s
im
por
tant
a
s
e
leme
nt
i.
I
n
a
ddit
ion,
the
c
ompar
is
on
of
the
s
a
me
two
e
lem
e
nts
will
pr
oduc
e
the
number
1
,
mea
ning
that
it
is
e
qua
ll
y
im
por
tant
.
T
wo
dif
f
e
r
e
nt
e
leme
nts
c
a
n
be
judge
d
e
qua
ll
y
im
por
tant.
F
r
om
e
a
c
h
pa
ir
wis
e
c
ompar
is
on
matr
i
x
t
he
n
the
e
igen
ve
c
tor
is
s
e
a
r
c
he
d
to
ge
t
loca
l
pr
ior
it
y
.
B
e
c
a
us
e
the
pa
ir
wis
e
c
ompar
is
on
matr
ix
e
xis
ts
a
t
e
a
c
h
leve
l,
then
to
ge
t
a
g
lobal
p
r
ior
it
y
mus
t
be
s
y
nthes
ize
d
be
twe
e
n
loca
l
pr
ior
it
y
.
S
tep
5
.
L
ogica
l
c
ons
is
tenc
y
s
tate
s
the
mea
s
ur
e
of
c
on
s
is
tenc
y
in
judgm
e
nt
o
r
we
ight
ing
in
pa
ir
wis
e
c
ompar
is
ons
.
T
his
tes
t
is
ne
c
e
s
s
a
r
y,
be
c
a
us
e
in
a
c
tual
c
ir
c
ums
tanc
e
s
ther
e
mi
ght
be
s
ome
de
viations
f
r
om
the
r
e
lations
hip
s
o
that
the
c
ompar
is
on
mat
r
ix
is
n
ot
c
ons
is
tently
p
e
r
f
e
c
t.
T
he
incons
is
tenc
y
index
is
ba
s
e
d
on
maximum
E
igenva
lue,
whic
h
is
c
a
lcula
ted
by
s
umm
ing
the
pr
oduc
t
of
e
a
c
h
e
leme
nt
in
the
E
ige
nve
c
tor
.
T
he
c
a
lcula
ti
on
of
the
c
ons
is
tenc
y
index
(
S
a
a
ty
,
20
05)
is
given
by
the
(
1)
.
=
−
−
1
(
1)
W
he
r
e
C
I
is
the
c
ons
is
t
e
nc
y
index
a
nd
n
is
the
number
of
e
va
luate
d
c
r
it
e
r
ia.
I
n
or
de
r
to
ve
r
if
y
whe
ther
the
c
ons
is
tenc
y
index
(
C
I
)
is
a
de
qua
te,
S
a
a
ty
(
S
a
a
ty,
2005)
s
ugge
s
ts
wha
t
ha
s
be
e
n
c
a
ll
e
d
the
c
ons
is
tenc
y
r
a
te
(
C
R
)
,
whic
h
is
de
ter
mi
ne
d
by
the
r
a
ti
o
b
e
twe
e
n
th
e
c
ons
is
tenc
y
index
a
nd
the
r
a
ndom
c
ons
is
tenc
y
in
de
x
(
R
I
)
.
T
he
matr
ix
will
be
c
ons
ider
e
d
c
ons
is
tent
if
the
r
e
s
ult
ing
r
a
ti
o
is
les
s
than
10%
.
=
/
(
2)
3.
RE
S
UL
T
S
A
ND
AN
AL
YSI
S
T
e
s
ts
c
a
r
r
ied
out
to
de
ter
mi
ne
the
s
uit
a
bil
it
y
o
f
th
e
s
ys
tem
with
the
e
xpe
c
ted
output
.
T
he
c
a
s
e
us
e
d
in
thi
s
tes
t
is
f
indi
ng
a
p
lac
e
to
e
a
t
o
r
r
e
s
taur
a
nt.
3.
1.
Cr
it
e
r
ia
we
igh
t
in
g
B
e
f
or
e
us
e
r
input
ti
ng
the
c
r
it
e
r
ion
,
c
ompar
is
on
va
lues
a
nd
a
lt
e
r
na
ti
ve
,
it
is
ne
c
e
s
s
a
r
y
to
input
the
e
xpe
c
ted
f
ood
c
r
it
e
r
ia
da
ta
f
ir
s
t
.
A
f
ter
input
ti
n
g
the
c
r
it
e
r
ia
da
ta,
the
we
ight
ing
of
the
c
r
it
e
r
ia
c
a
n
then
be
done
.
C
r
it
e
r
ia
a
r
e
we
ight
e
d
be
twe
e
n
c
r
it
e
r
ia
in
T
a
ble
1.
T
he
r
e
s
ult
s
of
c
ompar
ing
manua
l
c
a
lcula
ti
ons
with
the
AH
P
s
ys
tem
ha
ve
the
s
a
me
va
lue
s
hown
in
F
ig
ur
e
4.
T
he
tes
t
wa
s
c
onduc
ted
us
ing
30
s
e
nt
e
nc
e
s
da
ta
s
e
t
with
pos
it
ive,
ne
ga
ti
ve
,
a
nd
ne
utr
a
l
s
e
nti
ments
.
Da
tas
e
t
s
e
nti
ment
s
e
ntenc
e
s
s
uc
h
a
s
S
a
tur
da
y
nigh
t
a
nd
a
ya
m
ne
longs
o
(
M
a
lam
mi
nggu
da
n
a
ya
m
ne
longs
o)
,
a
ya
m
ne
longs
o
is
r
e
a
ll
y
f
a
mous
in
S
ur
a
ba
ya
(
ter
ke
na
l
ba
nge
t
ini
a
ya
m
ne
longs
o
di
S
ur
a
ba
ya
)
,
J
us
t
s
mell
the
c
ooking,
it
make
s
you
hungr
y,
e
s
pe
c
ially
it
tas
tes
good
(
C
ium
ar
oma
mas
ak
anny
a
aja
biki
n
lapar
,
apalagi
r
as
any
a
pas
ti
e
nak
)
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
3
,
J
une
2020:
13
4
3
-
13
5
3
1348
T
a
ble
1
.
C
r
it
e
r
ia
w
e
ight
ing
F
igur
e
4
.
C
r
it
e
r
ia
we
ight
ing
3.
2.
Re
s
p
on
s
e
t
im
e
p
r
os
e
s
s
c
r
ap
in
g
op
in
ion
d
an
s
e
n
t
im
e
n
t
an
alys
is
S
c
r
a
ping
is
done
ba
s
e
d
on
a
lt
e
r
na
ti
ve
na
me
k
e
ywor
ds
.
S
e
nti
ment
a
na
lys
is
pr
oc
e
s
s
r
uns
a
f
ter
the
opini
on
s
c
r
a
ping
pr
oc
e
s
s
is
c
ompl
e
te.
T
he
r
e
s
ult
s
of
the
r
e
s
pons
e
ti
me
tr
ial
opini
on
s
c
r
a
ping
pr
oc
e
s
s
a
nd
s
e
nti
ment
a
na
lys
is
with
s
e
ve
r
a
l
va
r
iable
s
a
s
in
T
a
ble
2
.
F
r
om
T
a
ble
2
,
the
number
of
li
ne
s
o
f
wor
ds
or
opini
ons
a
bout
a
lt
e
r
na
ti
ve
s
c
a
n
a
f
f
e
c
t
the
s
pe
e
d
of
r
e
s
pons
e
ti
me.
T
he
mor
e
opini
ons
,
the
longer
the
opini
on
e
xtr
a
c
ti
on
pr
oc
e
s
s
in
the
s
e
nti
ment
a
na
lys
is
pr
oc
e
s
s
.
I
nter
ne
t
s
pe
e
d
ha
s
a
n
e
f
f
e
c
t
on
the
p
r
oc
e
s
s
of
s
c
r
a
ping
opini
ons
f
r
om
s
oc
ial
media
T
witt
e
r
be
c
a
us
e
the
s
ys
tem
a
c
c
e
s
s
e
s
the
T
witt
e
r
API
dir
e
c
tl
y
.
T
a
ble
2
.
R
e
s
u
l
t
of
r
e
s
pons
e
ti
me
s
c
r
a
ping
pr
oc
e
s
s
a
nd
s
e
nti
ment
a
na
lys
is
tes
ti
ng
K
e
yw
or
ds
S
c
r
a
pi
ng opini
ons
r
e
s
ul
t
R
e
s
pons
e
t
im
e
D
e
pot
S
la
me
t
7
li
ne
1.90 s
B
e
be
k P
a
lu
pi
5
li
ne
1.44 s
S
ot
o C
a
k H
a
r
13
li
ne
2.35 s
I
ka
n B
a
ka
r
K
e
put
ih
4
li
ne
1.32 s
A
ya
m Ne
lo
ngs
o
7
li
ne
1.57 s
3.
3.
Alt
e
r
n
a
t
ive
we
igh
t
i
n
g
Alter
na
ti
ve
da
ta
input
"
plac
e
s
to
e
a
t"
whic
h
in
c
ludes
na
mes
a
nd
a
ddit
ional
inf
or
mation
a
bout
a
lt
e
r
na
ti
ve
s
a
r
e
a
ls
o
ne
e
de
d
be
f
or
e
we
ight
ing
the
a
lt
e
r
na
ti
ve
s
.
E
a
c
h
a
lt
e
r
na
ti
ve
"
plac
e
to
e
a
t"
that
ha
s
be
e
n
e
nter
e
d,
the
us
e
r
c
a
n
s
e
e
the
de
t
a
il
s
a
long
with
s
o
me
opini
ons
obtaine
d
f
r
om
the
s
ys
tem
s
c
r
a
ping
f
r
om
s
oc
ial
media
T
w
it
ter
.
An
a
lt
e
r
na
ti
ve
c
ompar
is
on
va
lue
input
is
pe
r
f
o
r
med
on
e
a
c
h
c
r
it
e
r
ion
e
xc
e
pt
the
s
pe
c
if
i
c
c
r
it
e
r
ion,
s
e
nti
ment
a
na
lys
is
a
s
s
hown
F
igur
e
s
5,
6
,
a
nd
7.
F
igur
e
5
.
Alter
na
ti
ve
we
ight
of
f
ood
’
s
pr
ice
a
nd
a
lt
e
r
na
ti
ve
we
ight
of
f
ood’
s
menu
P
r
ic
e
M
e
nu
D
is
ta
nc
e
L
oc
a
ti
on
P
a
r
ki
ng A
r
e
a
S
e
nt
im
e
nt
P
r
ic
e
(
har
ga m
ak
anan
)
1
0,333333
0,333333
3
5
M
e
nu
(
m
e
nu m
ak
anan
)
3
1
0,333333
3
5
D
is
ta
nc
e
L
oc
a
ti
on
(
ja
r
ak
t
e
m
pat
m
a
k
an
)
3
3
1
3,000003
7
P
a
r
ki
ng A
r
e
a
(
lu
as
par
k
e
r
)
0,333333
0,333333
0,333333
1
3
S
e
nt
im
e
nt
(
s
e
nt
ime
nt
anali
s
a
)
0,2
0,2
0,142857
0,333333
1
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
A
ppli
c
ati
on
of
r
e
c
omm
e
ndati
on
s
y
s
tem
w
it
h
A
HP
me
thod
and
s
e
nti
me
nt
analys
is
(
I
r
a
P
r
as
e
tyaningr
um
)
1349
F
igur
e
6
.
Alter
na
ti
ve
we
ight
of
dis
tanc
e
loca
ti
on
a
nd
a
lt
e
r
na
ti
ve
we
ight
o
f
c
r
it
e
r
ia
F
igur
e
7
.
S
e
nti
ment
we
ight
f
o
r
e
a
c
h
a
lt
e
r
na
ti
ve
a
n
d
f
inel
r
e
s
ult
T
he
S
e
nti
ment
Ana
lys
is
Ac
c
ur
a
ti
on
tes
t
wa
s
c
on
duc
ted
us
ing
30
s
e
ntenc
e
s
da
tas
e
t
with
pos
it
ive,
ne
ga
ti
ve
,
a
nd
ne
utr
a
l
s
e
nti
ments
.
T
o
c
a
lcula
te
the
a
c
c
ur
a
c
y
of
s
e
nti
ment
a
na
lys
is
,
a
f
or
mul
a
li
ke
the
f
oll
owing
is
us
e
d:
=
∑
100%
(
3)
f
r
om
thi
s
tes
t,
the
numbe
r
of
c
o
r
r
e
c
t
pr
e
diction
s
in
s
e
nti
ment
s
e
ntenc
e
s
is
26.
C
a
lcula
ti
on
of
s
e
nti
ment
a
na
lys
is
a
c
c
ur
a
c
y
is
a
s
f
oll
ows
:
=
26
30
100%
=
86
.
7%
t
he
a
c
c
ur
a
c
y
of
s
e
nti
ment
a
na
lys
is
r
e
a
c
he
s
86.
7%
i
n
tes
ti
ng
with
30
s
e
nti
ment
s
e
ntenc
e
s
da
tas
e
t.
3.
4.
Re
s
p
on
s
e
t
im
e
c
alcul
at
ion
p
r
oc
e
s
s
R
e
s
pons
e
ti
me
tes
ti
ng
of
the
c
a
lcula
ti
on
pr
oc
e
s
s
is
done
with
2
va
r
iable
s
,
na
mely
the
number
o
f
c
r
it
e
r
ia
a
nd
the
number
of
a
lt
e
r
na
ti
ve
s
.
T
he
r
e
s
ult
s
of
the
r
e
s
pons
e
ti
me
tes
ti
ng
of
the
c
a
lcula
ti
on
pr
oc
e
s
s
with
thes
e
2
va
r
iable
s
a
r
e
a
s
in
T
a
ble
3.
F
r
o
m
the
tes
t
r
e
s
ult
s
in
T
a
ble
3
,
the
a
ve
r
a
ge
va
lue
of
r
e
s
pons
e
ti
me
is
dir
e
c
tl
y
pr
opor
ti
ona
l
to
the
number
of
a
lt
e
r
na
ti
ve
s
.
I
n
tes
ti
ng
with
a
f
ixed
number
of
a
lt
e
r
na
t
ives
a
nd
the
number
of
dif
f
e
r
e
nt
c
r
it
e
r
ia,
r
e
s
pons
e
ti
me
is
dir
e
c
tl
y
pr
opor
ti
ona
l
to
the
number
o
f
c
r
it
e
r
ia
.
I
t
c
a
n
be
c
onc
luded
that
the
s
pe
e
d
of
r
e
s
pons
e
ti
me
c
a
lc
ulation
pr
oc
e
s
s
de
pe
nds
on
the
number
of
c
r
it
e
r
ia
a
nd
a
lt
e
r
na
ti
ve
s
.
T
he
mo
r
e
the
number
o
f
c
r
it
e
r
ia
a
nd
a
lt
e
r
na
ti
ve
s
,
the
g
r
e
a
ter
the
r
e
s
pons
e
ti
me.
3.
5.
App
li
c
a
t
ion
t
e
s
t
in
g
on
s
e
ve
r
al
we
b
b
r
ows
e
r
T
e
s
ti
ng
the
s
ys
tem
on
s
e
ve
r
a
l
we
b
br
ows
e
r
s
is
ne
e
de
d
to
de
ter
mi
ne
the
s
uit
a
bil
it
y
a
nd
r
e
s
pons
ibi
li
ty
of
the
s
ys
tem
in
the
br
ows
e
r
Google
C
hr
ome
,
M
oz
il
la
F
ir
e
f
ox
,
I
nter
ne
t
E
xplor
e
r
,
a
nd
M
icr
os
of
t
E
dge
.
3.
6.
S
ys
t
e
m
t
e
s
t
in
g
S
ys
tem
tes
ti
ng
is
done
to
va
li
da
t
e
e
a
c
h
a
ppl
ica
ti
on's
f
e
a
tur
e
s
in
a
c
c
or
da
nc
e
with
the
e
xpe
c
ted
output
.
T
he
r
e
s
ult
s
of
blac
kbox
tes
ti
ng
c
a
n
be
s
e
e
n
in
T
a
ble
4.
I
nter
f
a
c
e
tes
ti
ng
is
a
ls
o
pe
r
f
or
med
with
the
us
a
bil
it
y
pr
inciples
of
10
he
ur
is
ti
c
e
va
luat
ions
de
ve
loped
by
J
a
c
ob
Ne
il
s
e
n
r
e
s
ult
i
ng
in
s
e
ve
r
a
l
e
va
luations
a
s
in
T
a
ble
5.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
3
,
J
une
2020:
13
4
3
-
13
5
3
1350
T
a
ble
3
.
R
e
s
ult
of
r
e
s
pons
e
ti
me
c
a
lcula
ti
on
pr
oc
e
s
s
tes
ti
ng
A
lt
e
r
na
ti
ve
A
mount
C
r
it
e
r
ia
A
mount
R
e
s
pons
e
T
im
e
A
ve
r
a
ge
R
e
s
pons
e
T
im
e
2
2
831s
874
.
5s
3
836s
4
890s
5
957s
5
2
882s
982s
3
954s
4
992s
5
1100s
7
2
931s
1042
.
75s
3
1020s
4
1100s
5
1120s
T
a
b
le
4
.
R
e
s
ult
of
va
li
da
ti
on
tes
ti
ng
N
o.
C
a
s
e
t
e
s
t
C
ondi
ti
on
R
e
s
ul
t
1.
R
e
gi
s
tr
a
ti
on
U
s
e
r
s
c
a
n
r
e
gi
s
te
r
a
n
a
c
c
ount
w
it
h
a
n
e
ma
il
a
nd
us
e
r
na
me
th
a
t
ha
s
ne
ve
r
be
e
n
r
e
gi
s
te
r
e
d.
V
a
li
d
2.
L
ogi
n
U
s
e
r
s
c
a
n l
og i
n w
it
h a
r
e
gi
s
te
r
e
d us
e
r
na
me
a
nd pa
s
s
w
or
d.
V
a
li
d
3.
C
R
U
D
G
oa
l
U
s
e
r
s
c
a
n
c
ha
nge
a
nd d
e
le
te
goa
ls
.
V
a
li
d
4.
L
is
t
C
r
it
e
r
ia
D
is
pl
a
ys
c
r
it
e
r
ia
ba
s
e
d on the
us
e
r
w
ho i
s
l
ogge
d i
n.
V
a
li
d
5.
C
R
U
D
C
r
it
e
r
ia
U
s
e
r
s
c
a
n
a
dd, c
ha
nge
a
nd de
le
te
t
he
ir
c
r
it
e
r
ia
da
ta
.
V
a
li
d
6.
C
ompa
r
is
on va
lu
e
of
c
r
it
e
r
ia
S
how
in
g t
he
va
lu
e
pa
ir
w
is
e
c
ompa
r
is
on be
tw
e
e
n us
e
r
c
r
it
e
r
ia
.
V
a
li
d
7.
U
pda
te
c
ompa
r
is
on va
lu
e
c
r
it
e
r
ia
U
s
e
r
da
p
a
t
me
nguba
h va
lu
e
p
a
ir
w
is
e
c
ompa
r
is
on a
nt
a
r
kr
it
e
r
ia
me
r
e
ka
.
V
a
li
d
8.
L
is
t
A
lt
e
r
na
ti
ve
S
how
in
g a
lt
e
r
na
ti
ve
s
ba
s
e
d on the
u
s
e
r
w
ho i
s
l
ogge
d i
n.
V
a
li
d
9.
C
R
U
D
A
lt
e
r
na
ti
ve
U
s
e
r
s
c
a
n
a
dd, c
ha
nge
a
nd de
le
te
t
he
ir
a
lt
e
r
na
ti
ve
da
ta
.
V
a
li
d
10.
C
ompa
r
is
on va
lu
e
of
a
lt
e
r
na
ti
ve
D
is
pl
a
y pa
ir
w
is
e
c
ompa
r
is
on v
a
lu
e
s
be
tw
e
e
n a
lt
e
r
na
ti
ve
s
on e
a
c
h us
e
r
c
r
it
e
r
ia
.
V
a
li
d
11.
U
pda
te
c
ompa
r
is
on va
lu
e
a
lt
e
r
na
ti
ve
U
s
e
r
s
c
a
n c
ha
nge
t
he
va
lu
e
of
pa
ir
w
is
e
c
ompa
r
is
on be
twe
e
n a
lt
e
r
na
ti
ve
s
on e
a
c
h of
th
e
ir
c
r
it
e
r
ia
.
V
a
li
d
12.
C
a
lc
ul
a
ti
on
S
how
in
g
th
e
r
e
s
ul
ts
of
c
a
lc
ul
a
ti
on
s
in
th
e
f
or
m
of
a
lt
e
r
na
ti
ve
r
e
c
omm
e
nda
ti
ons
f
r
om t
he
s
ys
te
m pr
oc
e
s
s
.
V
a
li
d
13.
H
is
to
r
y
D
is
pl
a
ys
a
hi
s
to
r
y of
t
he
r
e
s
ul
ts
of
pr
e
vi
ous
c
a
lc
ul
a
ti
ons
.
V
a
li
d
T
a
ble
5
.
He
ur
is
ti
c
t
e
s
ti
ng
N
o.
P
r
in
c
ip
le
s
I
s
t
he
r
e
?
H
ow
?
1.
V
is
ib
il
it
y of
s
ys
te
m s
ta
tu
s
Y
e
s
T
he
s
ys
t
e
m
ha
s
gi
ve
n
not
if
ic
a
ti
on
w
he
n
s
uc
c
e
s
s
f
ul
ly
c
r
e
a
ti
ng,
upda
ti
ng,
a
nd
de
le
ti
ng.
2.
M
a
tc
h
be
twe
e
n
s
ys
t
e
m
a
nd
th
e
r
e
a
l
w
or
ld
Y
e
s
T
he
s
ys
te
m
us
e
s
a
gl
oba
l
la
ngua
ge
(
E
ngl
is
h)
th
a
t
is
f
a
mi
l
ia
r
a
nd
e
a
s
y
to
unde
r
s
ta
nd.
3.
U
s
e
r
c
ont
r
ol
a
nd f
r
e
e
dom
Y
e
s
U
s
e
r
s
c
a
n a
dd, c
ha
nge
, a
nd de
le
te
da
ta
. B
ut
t
he
r
e
i
s
no s
ys
te
m t
o r
e
s
to
r
e
de
le
te
d
da
ta
.
4.
C
ons
is
te
nc
y
a
nd s
ta
nd
a
r
ds
Y
e
s
E
ve
r
y butt
on, t
e
xt
, c
ol
or
, a
ppe
a
r
a
nc
e
on one
pa
ge
i
s
t
he
s
a
me
a
s
a
not
he
r
pa
ge
.
5.
E
r
r
or
p
r
e
ve
nt
io
n
Y
e
s
T
he
r
e
is
a
n
e
r
r
or
me
s
s
a
ge
on
e
ve
r
y
us
e
r
th
a
t
r
e
qui
r
e
s
doi
ng
th
e
pr
e
vi
ous
s
te
ps
,
s
uc
h a
s
f
il
li
ng goa
ls
, c
omp
a
r
is
on c
r
it
e
r
ia
, a
lt
e
r
na
ti
ve
c
ompa
r
is
o
n.
6.
R
e
c
ogni
ti
on r
a
th
e
r
t
ha
n r
e
c
a
ll
Y
e
s
U
s
e
r
s
do
not
ne
e
d
to
r
e
me
mbe
r
th
e
a
pp
e
a
r
a
nc
e
of
th
e
p
a
ge
me
nu,
be
c
a
us
e
th
e
r
e
a
r
e
but
to
ns
t
ha
t
ma
tc
h t
he
ir
f
unc
ti
ons
.
7.
F
le
xi
bi
li
ty
a
nd e
f
f
ic
ie
nc
y of
us
e
Y
e
s
I
n
c
hoos
in
g
a
pa
ir
w
is
e
c
ompa
r
is
on
va
lu
e
,
th
e
r
e
i
s
in
f
or
ma
ti
on
a
bout
th
e
me
a
ni
ng of
e
a
c
h va
lu
e
i
n pa
ir
w
is
e
c
ompa
r
is
on.
8.
A
e
s
th
e
ti
c
a
nd mi
ni
ma
li
s
t
de
s
ig
n
Y
e
s
S
im
pl
e
de
s
ig
n
not
to
o
ma
ny
c
ont
r
a
s
ti
ng
c
ol
or
s
.
T
he
te
xt
a
nd
a
ppe
a
r
a
nc
e
of
th
e
pa
ge
l
ook s
moot
h, not t
oo f
la
s
hy or
t
hi
c
k.
9.
H
e
lp
us
e
r
s
r
e
c
ogni
z
e
,
di
a
gno
s
e
,
a
nd r
e
c
ove
r
f
r
om e
r
r
or
s
Y
e
s
A
n
e
r
r
or
me
s
s
a
ge
a
ppe
a
r
s
w
he
n
c
onf
ir
mi
ng
th
e
pa
s
s
w
or
d
is
i
nc
or
r
e
c
t
a
nd
th
e
lo
gi
n i
s
i
nc
or
r
e
c
t.
E
r
r
or
s
out
s
id
e
t
he
s
ys
te
m a
r
e
di
r
e
c
te
d a
t
th
e
e
r
r
or
pa
ge
.
10.
H
e
lp
a
nd doc
ume
nt
a
ti
on
No
T
he
r
e
is
no
he
lp
me
nu
a
nd
doc
ume
nt
a
ti
on.
D
oc
ume
nt
a
ti
on
is
onl
y
in
th
e
f
or
m
of
a
book.
3.
7.
Us
e
r
t
e
s
t
in
g
T
e
s
ti
ng
is
done
to
the
us
e
r
to
f
ind
out
whe
ther
the
s
ys
tem
ha
s
pr
ovided
output
that
is
in
a
c
c
or
da
nc
e
with
the
us
e
r
.
T
he
tes
t
wa
s
c
onduc
ted
on
s
tudents
of
s
e
ve
r
a
l
P
E
NS
2016
in
f
or
matics
e
nginee
r
ing
s
tu
de
nts
a
nd
ge
ne
r
a
l
us
e
r
s
.
Us
e
r
s
who
do
tes
ti
ng
to
taling
10
p
e
ople
including
8
s
tudents
of
inf
or
matics
e
nginee
r
ing
a
nd
2
ge
ne
r
a
l
us
e
r
s
.
T
he
r
e
s
ult
s
f
r
om
r
e
s
ponde
nts
a
r
e
a
s
f
oll
ows
:
o
f
the
10
us
e
r
s
who
tr
ied
thi
s
a
ppli
c
a
ti
on,
a
ll
o
f
them
c
a
n
us
e
the
a
ppli
c
a
ti
on
e
a
s
il
y.
B
a
s
e
d
on
th
e
r
e
s
ult
s
of
tes
ti
ng
with
us
e
r
s
in
F
igur
e
8
,
they
c
a
n
us
e
the
a
ppli
c
a
ti
on
e
a
s
il
y.
T
a
king
the
opini
on
by
the
s
ys
tem
r
e
ga
r
ding
e
a
c
h
a
lt
e
r
na
ti
ve
is
les
s
than
the
m
a
xim
um.
T
he
r
e
a
r
e
a
lt
e
r
na
ti
ve
s
that
a
r
e
not
dis
c
us
s
e
d
o
n
T
witt
e
r
s
oc
ial
media
,
s
o
s
e
nti
ment
a
na
lys
is
on
thes
e
a
lt
e
r
na
ti
ve
s
is
z
e
r
o.
I
t
ne
e
ds
to
be
e
xpa
nde
d
r
e
ga
r
d
ing
s
oc
ial
media
a
s
a
n
a
lt
e
r
na
ti
ve
s
our
c
e
o
f
opin
io
n.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
A
ppli
c
ati
on
of
r
e
c
omm
e
ndati
on
s
y
s
tem
w
it
h
A
HP
me
thod
and
s
e
nti
me
nt
analys
is
(
I
r
a
P
r
as
e
tyaningr
um
)
1351
F
igur
e
8.
Us
e
r
s
ur
ve
y
4.
CONC
L
USI
ON
Af
ter
c
onduc
ti
ng
va
r
ious
s
tage
s
s
tar
ti
ng
f
r
om
the
de
s
ign
s
tage
,
making
the
s
ys
tem
then
pr
oc
e
e
d
to
the
tes
ti
ng
a
nd
a
na
lys
is
pha
s
e
,
the
f
oll
owing
c
onc
lus
ions
c
a
n
be
dr
a
wn
that
t
he
r
e
s
ult
s
of
the
c
a
lcul
a
ti
on
of
the
AH
P
pr
oc
e
s
s
on
the
s
ys
tem
with
manua
l
c
a
lc
ulations
a
r
e
the
s
a
me,
s
o
that
the
c
a
lcula
ti
on
of
the
AH
P
pr
oc
e
s
s
is
a
c
c
ur
a
te.
C
a
lcula
ti
on
s
pe
e
d
is
inf
luenc
e
d
by
th
e
number
o
f
c
r
it
e
r
ia
a
nd
a
lt
e
r
na
ti
ve
us
e
r
s
.
Applica
ti
ons
c
a
n
pr
ovide
r
e
c
omm
e
nda
ti
ons
a
c
c
or
ding
to
us
e
r
ne
e
ds
.
S
e
nti
ment
A
n
a
l
y
s
i
s
o
f
o
p
i
n
i
o
n
s
c
a
n
h
e
l
p
i
n
u
s
e
r
d
e
c
i
s
i
o
n
m
a
k
i
n
g
.
T
h
e
c
o
n
t
r
i
b
u
t
i
o
n
o
f
t
h
i
s
r
e
s
e
a
r
c
h
i
s
t
o
p
r
o
v
i
d
e
o
n
e
o
f
t
h
e
m
u
l
t
i
-
c
r
i
t
e
r
i
a
d
e
c
i
s
i
o
n
-
m
a
k
i
n
g
m
e
t
h
o
d
s
with
one
of
the
c
r
it
e
r
ia
be
ing
the
r
e
s
ult
s
of
opini
ons
f
r
om
s
oc
ial
media
a
bout
thes
e
a
lt
e
r
na
t
ives
.
T
he
c
ombi
na
ti
on
of
AH
P
a
nd
s
e
nti
ment
a
na
lys
is
pr
ovides
us
e
r
s
a
ti
s
f
a
c
ti
on.
Us
e
r
s
ge
t
the
be
s
t
r
e
c
omm
e
nda
ti
on
f
r
om
the
AH
P
c
a
lcula
ti
ons
with
s
ome
c
r
it
e
r
ia
f
r
om
us
e
r
s
a
nd
publi
c
opini
on
a
bout
e
a
c
h
a
lt
e
r
na
ti
ve
.
RE
F
E
RE
NC
E
S
[1
]
Su
ry
a
d
i
K
.
an
d
Ramad
h
an
i
M.
A
.
“
D
eci
s
i
o
n
S
u
p
p
o
rt
S
y
s
t
em:
A
St
r
u
ct
u
ral
D
i
s
c
o
u
r
s
e
o
n
Id
ea
l
i
z
at
i
o
n
an
d
Imp
l
eme
n
t
a
t
i
o
n
o
f
D
ec
i
s
i
o
n
Mak
i
n
g
(
i
n
Bah
a
s
a:
Si
s
t
em
Pen
d
u
k
u
n
g
K
e
p
u
t
u
s
an
:
Su
at
u
W
aca
n
a
St
ru
k
t
u
ral
Id
eal
i
s
as
i
d
an
Imp
l
emen
t
as
i
Pen
g
amb
i
l
a
n
K
ep
u
t
u
s
a
n
),
”
Ban
d
u
n
g
:
Remaj
a
Ro
s
d
a
k
ar
y
a
O
ffs
e
t
,
1
9
9
8
.
[2
]
N
.
K
o
mal
a
s
ar
i
,
“
D
eci
s
i
o
n
Su
p
p
o
rt
Su
p
p
o
rt
S
y
s
t
em
(SPK
2
T
)
(
i
n
Bah
a
s
a:
S
i
s
t
em
Pen
d
u
k
u
n
g
K
e
p
u
t
u
s
an
K
e
l
ai
k
a
n
T
erb
a
n
g
(SPK
2
T
)
)
,”
Ju
r
n
a
l
I
n
d
u
s
t
r
i
E
l
ek
t
r
o
d
a
n
P
e
n
er
b
a
n
g
a
n
,
v
o
l
.
4
,
n
o
.
1
,
2
0
2
0
.
[3
]
D
ai
h
an
i
D
.
“Co
mp
u
t
er
i
ze
d
D
eci
s
i
o
n
Ma
k
i
n
g
(
i
n
Bah
a
s
a:
K
o
m
p
u
t
eri
s
as
i
Pen
g
amb
i
l
a
n
K
e
p
u
t
u
s
an
)
,”
Bo
g
o
r
:
G
h
al
i
a
In
d
o
n
e
s
i
a,
2
0
0
1
.
[4
]
Ch
ami
d
A
.
A
.
an
d
B.
Su
rars
o
,
an
d
“
Imp
l
eme
n
t
a
t
i
o
n
o
f
A
H
P
an
d
Pro
me
t
h
e
e
Met
h
o
d
s
fo
r
Su
p
p
l
i
er
Sel
ec
t
i
o
n
(
i
n
Ba
h
as
a
:
Imp
l
eme
n
t
a
s
i
Me
t
o
d
e
A
H
P
d
an
Pr
o
met
h
ee
u
n
t
u
k
Pemi
l
i
h
an
S
u
p
p
l
i
er
)
,”
Ju
r
n
a
l
S
i
s
t
em
I
n
f
o
r
m
a
s
i
B
i
s
n
i
s
,
v
o
l
.
5
,
n
o
.
2
,
p
p
.
1
2
8
–
1
3
6
,
M
ay
2
0
1
5
.
[5
]
O
s
i
s
P.
K
.
“D
eci
s
i
o
n
Su
p
p
o
r
t
Sy
s
t
em
(
i
n
Bah
as
a
:
Si
s
t
em
Pen
u
n
j
an
g
K
ep
u
t
u
s
a
n
)
,”
U
n
i
v
ers
i
t
a
s
Mercu
Bu
an
a
Y
o
g
y
a
k
art
a
,
2
0
0
2
.
[6
]
S.
U
t
ami
n
i
n
g
s
i
h
,
“
Caree
r
-
b
a
s
ed
S
o
ft
S
k
i
l
l
D
e
v
e
l
o
p
m
en
t
at
V
o
ca
t
i
o
n
a
l
Sch
o
o
l
s
i
n
Semara
n
g
C
i
t
y
(
i
n
Bah
as
a
:
Pen
g
em
b
an
g
an
So
f
t
Sk
i
l
l
Berb
as
i
s
K
ar
i
e
r
p
ad
a
SMK
d
i
K
o
t
a
Semaran
g
,
”
D
i
n
.
P
en
d
i
d
i
k
,
v
o
l
.
6
,
n
o
.
2
,
p
p
.
1
1
9
–
1
3
3
,
2
0
1
1
.
[7
]
R.
Men
g
g
u
n
a
k
an
,
et
al
.
,
“
N
a
t
i
o
n
a
l
Sy
mp
o
s
i
u
m
o
n
A
p
p
l
i
e
d
T
ech
n
o
l
o
g
y
(S
N
T
T
)
(
i
n
Ba
h
as
a
:
Si
mp
o
s
i
u
m
N
as
i
o
n
a
l
T
ek
n
o
l
o
g
i
T
era
p
an
(S
N
T
T
)
,”
5
2
0
1
7
ISS
N
:
2
3
3
9
-
0
2
8
X
,
”
p
p
.
2
6
7
–
2
7
1
,
2
0
1
7
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
3
,
J
une
2020:
13
4
3
-
13
5
3
1352
[8
]
H
.
Su
p
ri
y
o
n
o
,
“
Sel
ec
t
i
o
n
o
f
Res
i
d
en
t
i
a
l
U
s
i
n
g
t
h
e
W
e
i
g
h
t
e
d
Pro
d
u
ct
Met
h
o
d
(
i
n
Bah
as
a
:
Pemi
l
i
h
an
Ru
ma
h
T
i
n
g
g
a
l
Men
g
g
u
n
ak
a
n
Met
o
d
e
W
e
i
g
h
t
e
d
Pro
d
u
c
t
,
”
Kh
a
z
a
n
a
h
I
n
f
o
r
m
a
t
i
k
a
:
Ju
r
n
a
l
I
l
m
u
K
o
m
p
u
t
er
d
a
n
In
f
o
r
m
a
t
i
k
a
,
v
o
l
.
1
,
n
o
.
1
,
p
p
.
2
3
–
2
8
,
2
0
1
5
.
[9
]
E
.
L
es
t
ari
an
d
E
.
U
.
A
r
t
h
a,
“
E
x
p
er
t
Sy
s
t
em
w
i
t
h
D
em
p
s
t
er
Sh
afer
Met
h
o
d
fo
r
D
i
ag
n
o
s
i
n
g
IN
D
I
H
O
M
E
Serv
i
ce
D
i
s
o
r
d
ers
at
PT
T
E
L
K
O
M
Mag
el
an
g
(
i
n
Bah
as
a:
Si
s
t
e
m
Pak
ar
d
en
g
a
n
Met
o
d
e
D
emp
s
t
er
Sh
afer
u
n
t
u
k
D
i
ag
n
o
s
i
s
G
an
g
g
u
an
L
ay
an
an
IN
D
I
H
O
ME
d
i
PT
T
E
L
K
O
M
Mag
el
an
g
)
,”
Kh
a
z
a
n
a
h
In
f
o
r
m
a
t
i
k
a
:
Ju
r
n
a
l
Il
m
u
Ko
m
p
u
t
e
r
d
a
n
In
f
o
r
m
a
t
i
k
a
,
v
o
l
.
3
,
n
o
.
1
,
p
p
.
1
6
–
2
4
,
2
0
1
7
.
[1
0
]
O
.
J
.
F.
W
as
s
al
am,
R.
U
mar,
an
d
A
.
Y
u
d
h
a
n
a,
“
Imp
l
emen
t
at
i
o
n
an
d
D
ev
e
l
o
p
men
t
o
f
W
e
b
-
Bas
e
d
E
-
L
earn
i
n
g
Sy
s
t
em
at
ST
IMI
K
Mu
h
ammad
i
y
a
h
Pag
u
y
an
g
an
(
i
n
Bah
as
a
:
Imp
l
emen
t
as
i
d
an
Pen
g
em
b
a
n
g
a
n
Si
s
t
em
E
-
L
earn
i
n
g
Berb
as
i
s
W
e
b
p
a
d
a
ST
IMIK
M
u
h
amma
d
i
y
ah
Pag
u
y
a
n
g
an
)
,”
P
r
o
ceed
i
n
g
S
E
N
D
I
,
p
p
.
1
0
4
–
1
0
7
,
2
0
1
7
.
[1
1
]
N
u
g
ro
h
o
an
d
S.
H
art
a
t
i
,
“
A
H
P
(A
n
al
y
t
i
cal
H
i
erarc
h
y
Pro
ces
s
)
Bas
ed
D
eci
s
i
o
n
S
u
p
p
o
r
t
Sy
s
t
em
fo
r
D
e
t
erm
i
n
i
n
g
Co
n
fo
rmi
t
y
o
f
L
an
d
U
s
e
(Cas
e
St
u
d
y
:
Semaran
g
Reg
en
cy
)
(
i
n
Bah
a
s
a:
Si
s
t
em
Pen
d
u
k
u
n
g
K
ep
u
t
u
s
a
n
Berb
as
i
s
A
H
P
(A
n
al
y
t
i
cal
H
i
erarc
h
y
Pr
o
ces
s
)
u
n
t
u
k
Pen
e
n
t
u
an
K
es
es
u
ai
a
n
Pen
g
g
u
n
aa
n
L
ah
an
(St
u
d
i
K
a
s
u
s
:
K
ab
u
p
at
e
n
Semaran
g
),
”
Ju
r
n
a
l
In
f
o
r
m
a
t
i
ka
,
v
o
l
.
6
,
n
o
.
2
,
p
p
.
6
3
0
–
6
4
1
,
2
0
1
2
.
[1
2
]
Fat
h
o
n
i
,
K
.
Pra
s
e
t
y
a
n
i
n
g
r
u
m,
I.
H
ar
i
y
a
t
i
,
C.
L
.
“
D
ec
i
s
i
o
n
Su
p
p
o
rt
S
y
s
t
em
f
o
r
Ch
o
o
s
i
n
g
D
a
y
care
i
n
S
u
rab
a
y
a
C
i
t
y
U
s
i
n
g
A
n
a
l
y
t
i
ca
l
H
i
erarc
h
y
Pro
c
es
s
(
A
H
P
)
,”
IE
E
E
In
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
er
e
n
ce
o
n
A
p
p
l
i
e
d
S
c
i
en
ce
a
n
d
Tech
n
o
l
o
g
y
(i
CA
S
T)
,
p
p
.
5
4
4
-
5
5
0
,
2
0
1
8
.
[1
3
]
U
mar
R,
Fad
l
i
l
A
,
an
d
Y
u
m
i
n
a
h
.
“
D
eci
s
i
o
n
Su
p
p
o
rt
Sy
s
t
em
w
i
t
h
A
H
P
Met
h
o
d
fo
r
E
m
p
l
o
y
ee
S
o
ft
S
k
i
l
l
Co
mp
e
t
en
c
y
A
s
s
es
s
men
t
.
(
i
n
Bah
as
a:
Si
s
t
em
Pen
d
u
k
u
n
g
K
e
p
u
t
u
s
an
d
e
n
g
a
n
Met
o
d
e
A
H
P
u
n
t
u
k
Pen
i
l
a
i
a
n
K
o
mp
et
e
n
s
i
So
ft
Sk
i
l
l
K
ar
y
aw
a
n
),
”
Kh
a
z
a
n
a
h
In
f
o
r
m
a
t
i
k
a
:
Ju
r
n
a
l
Il
m
u
Ko
m
p
u
t
e
r
d
a
n
In
f
o
r
m
a
t
i
ka
,
v
o
l
.
4
n
o
.
1
,
p
p
.
2
7
-
3
4
,
J
u
n
2
0
1
8
.
[1
4
]
S.
H
art
at
i
an
d
N
u
g
r
o
h
o
A,
“
A
H
P
(A
n
a
l
y
t
i
ca
l
H
i
erarc
h
y
Pro
ces
s
)
Bas
ed
D
ec
i
s
i
o
n
Su
p
p
o
r
t
Sy
s
t
em
fo
r
D
e
t
ermi
n
i
n
g
Co
n
fo
rmi
t
y
o
f
L
an
d
U
s
e
(Cas
e
St
u
d
y
:
Semaran
g
Reg
en
cy
)
i
In
Bah
a
s
a:
Si
s
t
em
Pen
d
u
k
u
n
g
K
ep
u
t
u
s
a
n
Berb
as
i
s
A
H
P
(A
n
al
y
t
i
cal
H
i
erarc
h
y
Pr
o
ces
s
)
u
n
t
u
k
Pen
e
n
t
u
an
K
es
es
u
ai
a
n
Pen
g
g
u
n
aa
n
L
ah
an
(St
u
d
i
K
a
s
u
s
:
K
ab
u
p
at
e
n
Semaran
g
),
”
Ju
r
n
a
l
In
f
o
r
m
a
t
i
ka
V
o
l
6
,
N
o
.
2
,
J
u
l
i
2
0
1
2
.
[1
5
]
Su
p
ri
y
o
n
o
,
“D
ec
i
s
i
o
n
Su
p
p
o
r
t
Sy
s
t
em
(SPK
)
Mo
t
o
rc
y
cl
e
Sel
ect
i
o
n
U
s
i
n
g
t
h
e
A
H
P
Met
h
o
d
(
i
n
Bah
as
a
:
Si
s
t
em
Pen
u
n
j
a
n
g
K
ep
u
t
u
s
a
n
(SPK
)
Pemi
l
i
h
an
Se
p
e
d
a
Mo
t
o
r
Men
g
g
u
n
ak
a
n
Me
t
o
d
e
A
H
P),
”
S
i
m
e
t
r
i
s
:
Ju
r
n
a
l
Tek
n
i
k
M
es
i
n
,
E
l
ekt
r
o
d
a
n
I
l
m
u
Ko
m
p
u
t
er
,
v
ol
.
1
,
n
o
.
1
,
p
p
.
5
5
-
58,
2
0
1
2
.
[1
6
]
Saefu
d
i
n
S.
an
d
W
ah
y
u
n
i
n
g
s
i
h
S
,
“D
eci
s
i
o
n
Su
p
p
o
rt
Sy
s
t
em
f
o
r
E
mp
l
o
y
ee
Perfo
rman
ce
E
v
al
u
at
i
o
n
U
s
i
n
g
A
n
a
l
y
t
i
ca
l
H
i
erarch
y
Pro
ce
s
s
(
A
h
p
)
Me
t
h
o
d
i
n
Sera
n
g
D
i
s
t
ri
ct
H
o
s
p
i
t
a
l
(
i
n
Bah
a
s
a:
S
i
s
t
em
Pe
n
d
u
k
u
n
g
K
e
p
u
t
u
s
a
n
U
n
t
u
k
Pe
n
i
l
ai
a
n
K
i
n
erj
a
Peg
a
w
ai
Men
g
g
u
n
a
k
an
Me
t
o
d
e
A
n
al
y
t
i
cal
H
i
erarch
y
Pro
ces
s
(A
h
p
)
Pad
a
RSU
D
Seran
g
),
”
Ju
r
n
a
l
S
i
s
t
e
m
In
f
o
r
m
a
s
i
(JS
i
I)
,
v
o
l
.
1
,
2
0
1
7
.
[1
7
]
Mu
n
t
h
afa,
A
.
E
.
,
Mu
b
aro
k
,
H
.
,
“A
p
p
l
i
ca
t
i
o
n
o
f
A
n
a
l
y
t
i
cal
H
i
erarch
y
Pr
o
ces
s
(A
H
P)
Me
t
h
o
d
i
n
D
eci
s
i
o
n
Su
p
p
o
r
t
Sy
s
t
em
f
o
r
St
u
d
en
t
A
ch
i
ev
eme
n
t
D
et
erm
i
n
a
t
i
o
n
(
i
n
Bah
as
a
:
Pen
erap
a
n
Met
o
d
e
A
n
al
y
t
i
cal
H
i
erarch
y
Pro
ces
s
(A
H
P)
d
al
am
S
i
s
t
em
Pen
d
u
k
u
n
g
K
ep
u
t
u
s
a
n
Pen
e
n
t
u
an
Mah
as
i
s
w
a
Berp
re
s
t
a
s
i
),
”
J
u
r
n
a
l
S
i
l
i
wa
n
g
i
S
e
r
i
S
a
i
n
s
d
a
n
Tekn
o
l
o
g
i
, v
ol.
3
.
n
o.
2
,
2
0
1
7
.
[1
8
]
G
au
rav
K
u
mar,
“A
Mu
l
t
i
-
Cr
i
t
e
ri
a
D
ec
i
s
i
o
n
-
Mak
i
n
g
A
p
p
ro
ac
h
fo
r
Reco
mmen
d
i
n
g
a
Pro
d
u
ct
u
s
i
n
g
Sen
t
i
men
t
A
n
a
l
y
s
i
s
”,
2
0
1
8
1
2
th
In
t
e
r
n
a
t
i
o
n
a
l
Co
n
f
er
e
n
ce
o
n
R
e
s
ea
r
ch
Ch
a
l
l
en
g
es
i
n
In
f
o
r
m
a
t
i
o
n
S
c
i
en
ce
(R
CI
S
)
,
2
0
1
8
.
[1
9
]
D
es
h
en
g
D
as
h
W
u
,
L
i
j
u
a
n
Z
h
en
g
,
d
an
D
a
v
i
d
L
.
O
l
s
o
n
,
“A
D
ec
i
s
i
o
n
Su
p
p
o
rt
A
p
p
r
o
ach
f
o
r
O
n
l
i
n
e
St
o
c
k
Fo
ru
m
S
en
t
i
me
n
t
A
n
al
y
s
i
s
”,
IE
E
E
T
r
a
n
s
a
ct
i
o
n
s
o
n
S
ys
t
e
m
s
,
M
a
n
,
a
n
d
Cyb
e
r
n
e
t
i
c
s
:
S
y
s
t
e
m
s
,
v
o
l
.
4
4
,
n
o
.
8
,
p
p
.
1
0
7
7
-
1
0
8
7
,
2
0
1
4
.
[2
0
]
Z
ao
l
i
Y
e
s
n
g
,
et
al
.
,
“A
D
eci
s
i
o
n
Me
t
h
o
d
fo
r
O
n
l
i
n
e
Pu
r
ch
as
e
s
Co
n
s
i
d
eri
n
g
D
y
n
am
i
c
In
f
o
rmat
i
o
n
Preferen
ce
Ba
s
ed
o
n
Sen
t
i
me
n
t
O
ri
e
n
t
a
t
i
o
n
Cl
a
s
s
i
fi
ca
t
i
o
n
a
n
d
D
i
s
cre
t
e
D
I
FW
A
O
p
erat
o
rs
”,
IE
E
E
A
cces
s
,
p
p
.
7
7
0
08
-
7
7
0
2
6
,
2
0
1
7
.
[2
1
]
Y
.
E
.
So
el
i
s
t
i
o
,
"
Si
mp
l
e
T
ex
t
Mi
n
i
n
g
fo
r
Sen
t
i
men
t
A
n
al
y
s
i
s
o
n
Po
l
i
t
i
cal
Fi
g
u
res
U
s
i
n
g
t
h
e
N
a
ï
v
e
Bay
e
s
Cl
as
s
i
fi
e
r
Met
h
o
d
(
i
n
Bah
a
s
a:
T
e
x
t
Mi
n
i
n
g
Sed
er
h
an
a
u
n
t
u
k
Sen
t
i
me
n
t
A
n
a
l
y
s
i
s
p
ad
a
Fi
g
u
r
Po
l
i
t
i
k
Men
g
g
u
n
ak
a
n
Met
o
d
e
N
aï
v
e
Bay
es
C
l
as
s
i
f
i
er
)
,
"
Th
e
P
r
o
ceed
i
n
g
s
o
f
t
h
e
7
th
ICTS
,
B
a
l
i
,
M
a
y
1
5
th
-
16
th
,
2
0
1
3
(ISS
N
:
9
7
7
2
3
3
8
1
8
5
0
0
1
)
.
[2
2
]
A
n
a
s
t
a
s
i
a
D
e
w
an
t
i
Prat
ama
Pu
t
ri
,
“
T
em
p
o
ra
l
Sen
t
i
men
t
Bas
ed
E
as
t
J
av
a
T
o
u
r
i
s
m
O
b
j
ec
t
Reco
mmen
d
at
i
o
n
Sy
s
t
em
(
i
n
Bah
a
s
a:
Si
s
t
em
Re
k
o
me
n
d
a
s
i
O
b
j
e
k
W
i
s
a
t
a
J
aw
a
T
i
mu
r
Berb
as
i
s
T
em
p
o
r
al
Sen
t
i
me
n
t
),
Po
l
i
t
ek
n
i
k
E
l
e
k
t
r
o
n
i
k
a
N
eg
er
i
Su
rab
a
y
a,
2
0
1
8
.
”
[2
3
]
N
u
r
j
an
a
h
,
W
.
E
.
,
Perd
an
a,
R.
S.
,
Fau
zi
,
M.
A
.
“Sen
t
i
men
t
A
n
a
l
y
s
i
s
o
f
T
el
e
v
i
s
i
o
n
Sh
o
w
s
Bas
ed
o
n
Pu
b
l
i
c
O
p
i
n
i
o
n
o
n
T
w
i
t
t
er
So
ci
a
l
Med
i
a
u
s
i
n
g
t
h
e
K
-
N
eares
t
N
ei
g
h
b
o
r
M
et
h
o
d
an
d
W
ei
g
h
t
i
n
g
t
h
e
N
u
mb
er
o
f
Ret
w
eet
s
(
i
n
Ba
h
as
a:
A
n
a
l
i
s
i
s
Se
n
t
i
men
T
er
h
ad
a
p
T
a
y
an
g
an
T
e
l
ev
i
s
i
Berd
as
ar
k
an
O
p
i
n
i
Mas
y
arak
a
t
p
a
d
a
Med
i
a
So
s
i
a
l
T
w
i
t
t
e
r
men
g
g
u
n
ak
a
n
Met
o
d
e
K
-
N
eare
s
t
N
e
i
g
h
b
o
r
d
an
Pemb
o
b
o
t
an
J
u
ml
a
h
Ret
w
ee
t
),
”
Ju
r
n
a
l
P
en
g
em
b
a
n
g
a
n
Tekn
o
l
o
g
i
In
f
o
r
m
a
s
i
d
a
n
Il
m
u
Ko
m
p
u
t
er
e
-
I
S
S
N
,
v
o
l
.
1
,
n
o
.
1
2
,
p
p
.
1
7
5
0
-
1
7
5
7
,
D
es
em
b
er
2
0
1
7
.
[2
4
]
Si
p
ay
u
n
g
,
E
.
M.
,
Mah
ara
n
i
an
d
H
.
,
Z
efan
y
a,
I.
“
D
es
i
g
n
i
n
g
Cu
s
t
o
mer
Co
mme
n
t
Se
n
t
i
men
t
A
n
a
l
y
s
i
s
Sy
s
t
em
U
s
i
n
g
t
h
e
N
ai
v
e
Bay
e
s
Cl
a
s
s
i
fi
er
Me
t
h
o
d
(
i
n
Ba
h
as
a
:
Pera
n
can
g
an
S
i
s
t
em
A
n
al
i
s
i
s
Se
n
t
i
men
K
o
me
n
t
ar
Pe
l
an
g
g
an
Men
g
g
u
n
ak
a
n
Met
o
d
e
N
a
i
v
e
Bay
e
s
Cl
a
s
s
i
fi
er),
”
Ju
r
n
a
l
S
i
s
t
e
m
In
f
o
r
m
a
s
i
,
v
o
l
.
8
,
n
o
.
1
,
A
p
r
i
l
2
0
1
6
.
[2
5
]
In
d
ra
y
u
n
i
,
E
.
“
H
o
t
el
Rev
i
ew
Se
n
t
i
men
t
A
n
al
y
s
i
s
U
s
i
n
g
t
h
e
V
ect
o
r
Mach
i
n
e
Su
p
p
o
rt
A
l
g
o
r
i
t
h
m
(
i
n
Ba
h
as
a
:
A
n
a
l
i
s
a
Sen
t
i
m
en
Rev
i
e
w
H
o
t
el
Men
g
g
u
n
a
k
an
A
l
g
o
ri
t
ma
Su
p
p
o
r
t
V
ect
o
r
Mach
i
n
e
),
”
E
V
O
L
U
S
I:
Ju
r
n
a
l
S
a
i
n
s
d
a
n
M
a
n
a
j
em
e
n
,
v
o
l
.
4
,
n
o
.
2
,
2
0
1
6
.
Evaluation Warning : The document was created with Spire.PDF for Python.