Int
ern
at
i
onal
Journ
al of Ele
ctrical
an
d
Co
mput
er
En
gin
eeri
ng
(IJ
E
C
E)
Vo
l.
11
,
No.
1
,
Febr
uar
y
2021
, pp.
4
8
9
~
497
IS
S
N: 20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v11
i
1
.
pp
489
-
497
489
Journ
al h
om
e
page
:
http:
//
ij
ece.i
aesc
or
e.c
om
Su
stainab
le governanc
e in
smart citie
s and
use of su
pervis
ed
learnin
g based
op
ini
on mini
ng
Hena Iqb
al
1
, S
uj
ni
Pa
ul
2
,
K
ha
li
quz
z
ama
n
Kh
an
3
1
School
of Engi
nee
ring
and
Tec
hnolog
y
,
Al
Dar
Univer
sit
y
Co
ll
e
ge,
Uni
te
d
Arab Em
ira
te
s
2
Depa
rt
m
ent
of
Com
pute
r
Infor
m
at
ion
Sci
ence,
Higher
Col
le
ges
of
Technol
og
y
,
Unite
d
Arab
Em
ira
t
es
3
School
of
Busin
ess Adm
ini
strat
i
on
,
Al
Dar
Univ
ersity
Co
ll
eg
e,
Unite
d
Arab Em
ira
te
s
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
May
7,
2020
Re
vised Ju
n 3
0, 20
20
Accepte
d Aug
4,
202
0
Eva
lu
at
ion
is
an
ana
l
y
t
ical
and
orga
nized
proc
e
ss
to
figure
out
the
pre
sen
t
positi
ve
infl
u
en
ce
s,
f
avour
ab
le
future
prospe
ct
s
,
ex
isti
ng
short
c
om
ings
and
ult
eri
o
r
complex
it
ie
s of
an
y
p
la
n,
progra
m
,
pra
cti
ce
or
a
pol
icy
.
E
val
ua
ti
on
of
poli
c
y
is
an
essenti
al
and
vital
pr
oce
ss
req
uire
d
t
o
m
ea
sure
the
p
erf
orm
anc
e
or
progre
ss
ion
of
the
s
che
m
e.
The
m
ai
n
purp
ose
of
polic
y
e
val
ua
ti
on
is
to
empow
er
v
ari
ous
stak
eholders
and
enhance
the
ir
soci
o
-
ec
onom
i
c
envi
ronm
ent
.
A
la
rg
e
num
ber
of
policie
s
or
sche
m
es
in
diff
ere
n
t
ar
ea
s
are
la
un
che
d
b
y
gover
nm
ent
in
vie
w
of
ci
tizen
welf
are.
Although,
the
gover
nm
enta
l
poli
c
ie
s
intend
to
bet
t
er
shape
u
p
the
life
qua
li
t
y
of
people
but
m
a
y
a
lso
i
m
pac
t
their
ever
y
da
y
’s
l
ife.
A
la
te
st
gov
ern
m
ent
a
l
sche
m
e
Saubhag
y
a
la
un
che
d
b
y
India
n
gover
nm
ent
in
2017
has
be
en
sele
c
te
d
for
eva
lu
at
ion
b
y
a
ppl
y
ing
opini
on
m
ini
ng
te
chni
q
ues.
The
da
ta
s
et
of
public
opini
on
associate
d
with
th
is
sche
m
e
has
be
en
ca
p
ture
d
b
y
Twi
tt
er
.
The
primar
y
inte
nt
is
to
offe
r
op
i
nion
m
ini
ng
as
a
sm
art
ci
t
y
t
ec
hn
olog
y
th
a
t
har
ness
the
use
r
-
gene
ra
te
d
big
dat
a
and
anal
y
s
e
it
to
off
er
a
sus
ta
ina
bl
e
gover
nance
m
od
el
.
Ke
yw
or
d
s
:
Op
i
nion m
ining
Po
li
cy
ev
al
uation
Schem
e
Twitt
er
This
is an
open
acc
ess arti
cl
e
un
der
the
CC
B
Y
-
SA
l
ic
ense
.
Corres
pond
in
g
Aut
h
or
:
Hen
a
Iq
bal,
School
of E
ng
i
neer
i
ng and T
e
chnolo
gy,
Al D
a
r Un
i
versi
ty
Colle
ge,
Near G
GI
C
O M
et
ro
, Ga
rho
ud, D
ub
ai
,
U
nited
Ar
a
b
Em
irat
es
.
Em
a
il
:
hen
a@
al
dar
.ac
.ae
1.
INTROD
U
CTION
Po
li
cy
is
a
se
t
of
la
ws
an
d
pr
i
nciples,
coll
ect
i
on
of
ru
le
s
and
re
gu
la
ti
ons,
an
a
rt
of
c
od
e
c
onduct
m
anifested
by
any
authe
ntic
governin
g
body.
They
ar
e
desig
ne
d
to
e
nr
ic
h
the
deci
sion
m
aking
proces
s
resu
lt
in
g
in
fa
vourable
outc
om
es
fo
r
the
be
tt
er
m
ent
of
any
so
ci
et
y
or
com
m
un
it
y.
T
he
pro
gr
e
ss
cu
rv
e
of
a
po
li
cy
repre
sents
the
ga
p
betwee
n
act
ual
and
e
xp
e
ct
ed
end
resu
lt
s
w
hich
ca
n
be
de
te
rm
ined
by
po
li
cy
evaluati
on.
I
nd
eed,
the
eval
ua
ti
on
of
a
poli
cy
is
a
crit
ic
al
pr
oces
s
and
ref
l
ect
s
the
pu
blic
respo
ns
e
w
hich
m
a
y
var
ie
s
f
ro
m
p
e
o
p
l
e
t
o
p
e
o
p
l
e
,
c
o
m
m
u
n
i
t
y
o
r
s
o
c
i
e
t
y
.
T
h
e
r
e
f
o
r
e
,
t
h
e
r
e
i
s
p
r
e
s
s
i
n
g
n
e
e
d
f
o
r
c
o
n
s
u
m
i
n
g
i
n
d
i
v
i
d
u
a
l
’
s
o
p
i
n
i
o
n
[
1
]
in t
he
e
valuati
on
proces
s whic
h
l
eads t
he gover
nm
ent to f
inali
ze the
fu
t
ur
e
ac
ti
on
s
for
a
poli
cy
.
More
ov
e
r,
the
s
m
art
city
-
s
m
art
nation
init
ia
ti
ves
in
recen
t
ye
ars
by
var
iou
s
gove
rn
m
ent
agen
ci
es
acro
s
s
the
gl
obe
ai
m
at
design
i
ng
sm
art
ci
ty
te
chnolo
gies
(S
CT
s)
w
hic
h
e
nhance
the
ci
ti
es’
sm
artness
an
d
i
m
pr
oves
s
us
ta
inabili
ty
.
S
ocia
l
web
[
2
]
play
s
a
vital
ro
le
i
n
order
t
o
s
uppor
t
so
ci
al
interact
ion
s
am
ong
pe
op
le
.
It
facil
it
at
es
t
he
public
interc
omm
un
ic
at
ion
about
a
ny
gove
rn
m
ental
po
li
cy
with
a
vie
w
to
get
t
heir
fe
edb
a
c
k
by
the
us
e
of
var
io
us
s
ocia
l
network
i
ng
sit
es
and
c
hannels.
P
ro
ces
sing
of
s
uc
h
a
huge
am
ou
nt
of
data
gen
e
rated
ove
r
s
uch
m
edia
te
nds
to
bet
te
r
analy
se
t
he
decisi
on
m
akin
g
process
.
The
pr
ocessi
ng
is
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol. 11,
No.
1,
Febr
uar
y
2021 :
48
9
-
49
7
490
accom
plished
with
the
ai
d
of
op
i
nion
m
ining
f
or
t
he
sake
of
e
xtracti
ng
public
opini
on
f
ro
m
this
big
da
ta
in
the
directi
on
of
good
poli
cy
m
aking
.
One
of
the
rece
nt
governm
ental
sc
hem
es’
"Pradh
an
Ma
ntri
Sa
ha
j
Bi
j
li
Har
G
ha
r
Y
oj
a
na,
or
Saub
ha
gy
a"
[
3
]
has
bee
n
co
ntem
plated
in
this
wor
k
for
inco
r
porati
ng
the
te
ch
niques
of
op
i
nion
m
ining
for
the
sa
ke
of
im
pr
ov
in
g
t
he
process
of
po
li
cy
ev
al
uation.
T
he
sc
hem
e
has
bee
n
la
unch
e
d
("Saub
ha
gya
schem
e:
All
yo
u
nee
d
t
o
kn
ow",
2017
[
3
]
)
f
or
the
el
ect
rif
ic
at
ion
of
ho
use
holds
in
r
ural
an
d
urba
n
area
s
of
India
by D
ece
m
ber
2
01
8
[
4
]
.
This
pa
per
util
iz
es
on
e
of
th
e
fam
ou
s
so
ci
al
network
i
ng
sit
es,
Twitt
er
[
5
]
,
as
a
source
for
datase
t
colle
ct
ion
due
to foll
owin
g
re
aso
ns
:
-
Associ
at
ion
wi
th volum
e a
nd
var
ia
nce
of
pe
op
le
-
Coll
ect
ion
of num
ero
us
div
e
rs
ifie
d
to
pics
-
Pr
ovi
ding a
great
o
pp
or
t
un
it
y f
or
resea
rch
e
r
s to
e
xploit
the
ir interests
-
Plac
ing
a
nd s
ha
rin
g of
t
houg
hts at a lar
ge
s
cal
e to get
a c
om
m
on
conclusi
on for a s
pecifi
c sub
j
ect
.
-
Most wi
dely
use
d
t
oo
l
by
gove
rn
m
ent to a
ppro
ac
h pe
op
le
-
Act as a
n
ea
rly
w
a
rn
i
ng syst
em
f
or
go
vernm
ental
bod
ie
s
du
e to its
qu
ic
k n
at
ur
e
-
On
ly
foru
m
which
w
or
l
dw
i
de
sh
a
re
op
i
nion
or f
ee
dback
aga
inst g
over
nm
e
ntal exe
rcises
Data
has
been
colle
ct
ed
ove
r
a
per
i
od
of
t
wo
m
on
ths
a
fter
the
im
ple
mentat
ion
of
sc
hem
e
and
is
furthe
r
div
i
ded
into
4
phases
as
li
ste
d
in
T
able
1.
This
c
la
ssific
at
ion
of
data
colle
ct
ion
per
i
od
inte
nds
to
pro
vid
e a
n
a
pp
roxim
at
e cov
er
age
of pu
blic p
erceptio
n re
gardin
g
the
sc
h
em
e
[
4
]
.
Table
1
.
Ph
ase
s alo
ng w
it
h t
he
ir co
rr
es
pondi
ng
d
urat
ion
Ph
ase No
Du
ration
Ph
ase I
No
v
9, 20
1
7
--
No
v
15
,
2
0
1
7
Ph
ase II
No
v
1
6
,
2
0
1
7
--
No
v
22
,
2
0
1
7
Ph
ase III
No
v
23
,
2
0
1
7
–
No
v
29
,
2
0
1
7
Ph
ase IV
No
v
30
,
2
0
1
7
--
Dec 9, 20
1
7
This
wor
k
proffers
an
op
i
ni
on
m
ining
ba
sed
s
us
ta
ina
bl
e
gov
e
r
nan
ce
m
od
el
,
Sm
art
Ele
ct
ricGov,
as
a
s
m
art
ci
t
y
te
chn
ol
ogy
(SC
T)
m
od
el
fo
r
s
m
art
ci
t
y
-
s
m
art
nation
init
i
at
ive.
It
dem
onstrat
es
the
sm
a
rt
ci
t
y
no
ti
on
of
en
ga
ging
m
or
e
eff
e
ct
ively
and
act
ively
with
it
s
c
it
iz
ens
fo
r
gove
rn
m
ent
po
li
cy
trackin
g,
pri
m
aril
y
its
evaluati
on.
It
al
so
e
xp
la
i
ns
the
c
once
pt
of
opini
on
m
ining
an
d
pr
es
ents
ho
w
opin
ion
m
ining
ha
s
bee
n
introd
uced
for
the
optim
iz
at
io
n
of
poli
cy
eva
luati
on
process
.
Sect
ion
3
perform
s
the
evaluati
on
of
Saub
hagy
a
schem
e
(a
go
ve
rn
m
ent
init
ia
ti
ve
to
pr
ov
i
de
el
e
ct
rici
ty
to
ever
y
house
within
the
natio
n)
us
in
g
opinio
n
m
ining
te
chn
iq
ues
wh
i
ch
com
pr
ise
s
of
data
acq
uisit
ion
,
a
naly
ti
cal
and
c
on
ce
ptua
l
rep
rese
ntati
on
of
data
an
d
app
ly
op
i
nion
m
ining
process
to
de
te
rm
ine
the
pola
rity
of
publ
ic
opinio
n
with
t
heir
t
weets.
Sect
io
n
4
em
pirical
ly
analy
ses
the
resu
lt
s
and
fi
ndin
gs
of
al
l
t
he
dif
fer
e
nt
m
achine
le
arni
ng
al
gorithm
s
in
te
r
m
s
of
eff
ic
acy
m
easur
es al
ong wit
h
thei
r gr
a
ph
ic
al
represe
nt
at
ion
s.
La
stl
y, Sect
ion 5
conc
lud
es
the
pa
per.
2.
SMART
CITI
ES, POLI
C
Y E
VA
L
UA
TI
O
N AND
OPI
N
ION
MI
NI
NG
The
gove
rn
m
ent
age
ncies
a
r
e
ta
king
ste
ps
su
c
h
that
t
he
s
m
art
ci
ty
init
i
at
ives
be
ne
fit
ever
y
on
e
-
it
s
reside
nts,
busi
ness
pe
rs
on
s
a
nd
the
go
vern
m
ent.
A
po
li
cy
is
a
po
werfu
l
too
l
co
nf
i
gure
d
in
respo
ns
e
to
address
the
so
ci
al
chal
le
ng
es
a
nd
to
ac
com
plish
the
basic
esse
ntial
s
of
a
c
omm
on
m
an.
It
helps
gove
rn
m
ent
or
orga
nizat
ion
in
i
m
pr
ovin
g
the
process
of
decisi
on
m
aki
ng
a
nd
re
sour
ce
op
ti
m
iz
at
ion
("P
olicy
",
20
18
)
[
4
]
.
They
ar
e
dev
is
ed
to
ass
ur
e
a
fa
vour
a
ble
ou
tc
om
e
wh
ic
h
l
ifts
the
s
ocio
-
e
conom
ic
l
evel
of
a
c
omm
un
ity
or
so
ci
et
y.
If
th
e
gove
rn
m
ent
poli
ci
es
are
favo
ur
a
ble,
m
any
l
ocal
&
gl
ob
al
com
pan
ie
s
will
inv
est
in
the
fu
t
ure
sm
art
citie
s
proj
ec
ts
.
T
he
process
of
poli
c
y
m
aking
is
a
ne
ver
en
ding
effo
rt
pe
rform
ed
with
a
vi
ew
to
form
ulate
the
req
ui
rem
e
nts
of
people
into
go
vernm
ental
cou
rse
of
act
io
n.
Figure
1
repres
ents
the
abstra
ct
and
gen
e
ric
view
of
a
poli
cy
li
fe
cy
cl
e.
The
cy
c
le
segr
e
gates
t
he
process
of
po
li
cy
int
o
a
s
equ
e
nce
of
phases
:
po
li
cy
p
la
nnin
g,
poli
cy
an
al
ysi
s & devel
opm
ent, polic
y m
on
it
or
in
g
a
nd
poli
cy
ev
al
uatio
n.
a.
Po
li
cy
Plan
nin
g:
Be
i
ng
the
first
ph
a
se
of
the
cy
cl
e,
the
m
ai
n
fo
c
us
is
to
i
den
ti
fy
the
pro
blem
.
These
pr
ob
le
m
s
m
a
y
arise
ei
ther
from
e
xisti
ng
c
om
ple
xiti
es
in
the
syst
e
m
or
to
a
chieve
the
f
ut
ur
e
obj
ect
ives
f
or
the
w
el
far
e
of
so
ci
et
y.
Var
i
ou
s
c
on
tr
over
s
ia
l
issues
or
di
sp
ute
d
po
i
nts
w
hich
seek
a
n
i
m
m
ediat
e
at
ten
ti
on
f
r
om
go
ve
rn
m
ent
or
org
anizat
ion
an
d
f
os
te
r
them
to
pro
vid
e
a
feasible
so
luti
on
ar
e
recog
nized
an
d
reco
r
de
d
f
or
t
he
age
nd
a
.
The
natur
e
of
eac
h
and
eve
ry
pro
blem
is
then
d
escribe
d
in
orde
r
to g
et
dee
per a
nd b
et
te
r
un
derst
and abili
ty
o
f
the situati
on.
b.
Po
li
cy
A
naly
sis
&
De
vel
opm
ent:
An
org
anized
a
nd
c
om
ple
te
interpret
at
ion
of
ide
ntifie
d
pro
blem
is
ano
t
her
c
riti
cal
ste
p
in
the
po
l
ic
y
li
fe
cy
cl
e
m
od
el
.
It
is
the
process
of
dec
o
m
po
sin
g
the
entan
gled
iss
ue
s
into
sm
al
le
r
and
at
om
ic
com
p
on
e
nts
in
or
der
to
bette
r
un
de
rstan
d
the
co
rrel
at
ion
bet
wee
n
them
wh
ic
h
i
n
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
Su
st
aina
ble go
vern
an
ce
in
s
m
ar
t ci
ti
es and u
se o
f s
upervi
se
d
le
arni
ng bas
ed op
i
nion
mini
ng
(
Hena Iq
ba
l
)
491
tur
n
adep
ts
th
e
pr
oce
ss
of
pro
blem
so
lving
a
nd
decisi
on
m
aking
.
T
echnoc
racy,
burea
ucr
acy
an
d
dem
ocr
acy
are
the
three
e
ssenti
al
par
am
et
ers
ke
pt
un
der
c
onsiderat
ion
wh
il
e
ana
ly
sing
a
poli
cy
.
It
assesses
the
po
li
ti
cal
an
d
te
chn
ic
al
c
om
pli
cat
ion
s
a
nd
prov
i
de
valua
ble
inf
or
m
at
ion
to
decisi
on
m
aker
s
or
po
li
cy
anal
yst
s
about
the
aug
m
entat
ion
of
a
poli
cy
and
it
s
ef
fect
on
s
ocio
-
ec
onom
i
c,
po
li
ti
c
al
,
te
chn
ic
al
a
nd
le
gal
fact
or
s
[
6
-
8
]
.
P
olicy
anal
ysi
s
is
fo
ll
ow
e
d
by
the
pro
ces
s
of
po
li
cy
devel
op
m
ent
w
hic
h
include
s
it
s
form
ulati
on
and
plan
of
im
plem
entat
ion
.
T
his
requires
ef
fe
ct
ive
wr
it
in
g
s
kill
s
in
orde
r
t
o
dr
a
ft
the
po
li
cy
,
it
s
m
od
e
of
co
m
m
un
ic
at
ion
to
dif
fer
e
nt
sta
keholde
r
s
and
to
get
it
pu
blishe
d
on
a
unive
rsal f
orum
.
c.
Po
li
cy
Mon
it
ori
ng:
The
pri
m
e
fo
c
us
of
this
ph
ase
is
on
the
co
ntin
uous
assessm
ent
of
i
m
ple
m
entat
io
n
pr
act
ic
es
car
ri
ed
out
f
or
poli
cy
dev
el
opm
ent
[
9
]
.
It
helps
t
o
rec
ognise
t
he
pr
om
isi
ng
ga
ps
i
n
the
exec
ution
of
poli
ci
es
and
t
races
the
bl
ueprint
f
or
their
e
nri
chm
ent.
In
-
house
strat
egies and
p
r
oce
dures
are
de
velo
pe
d
to
ens
ur
e
the
tr
ackin
g
a
nd
repor
ti
ng
a
bout
th
e
po
li
cy
pro
gre
ss.
T
he
pr
oces
s
of
m
on
it
ori
ng
sh
oul
d be
per
i
od
ic
i
n n
at
ure t
hat m
eans,
pe
rfor
m
ed
at
r
e
gu
l
ar in
te
r
vals
of t
i
m
e.
d.
Po
li
cy
Eval
uation:
P
olicy
ev
al
uation
is
one
of
the
cr
ucial
ph
a
se
in
the
li
fe
s
pan
of
a
poli
cy
.
It
a
naly
ses
var
i
ou
s
inter
na
l
and
exter
na
l
factor
s
w
hi
ch
m
easur
e
the
poli
cy
pe
rfor
m
ance
bas
ed
on
pe
op
le
adm
inist
rati
on
in
al
l
the
po
te
ntial
aspects.
I
t
helps
in
determ
ining
whet
he
r
the
poli
cy
outc
om
e
is
the
as
per
the
requi
sit
e
exp
ect
at
io
ns
or
f
or
the
purpose
f
or
wh
ic
h
it
is
i
nten
ded
t
o
be
[10
-
12]
.
B
oth
the
qual
it
at
ive
an
d
qua
ntit
at
i
ve
m
easur
es
a
re
ta
ke
n
i
nt
o
c
on
si
der
at
io
n
f
or
t
he
re
sourc
e
optim
iz
ation
of
te
chn
ic
al
,
s
oci
al
,
fina
ncial
,
et
hical
an
d
oth
e
r
aspects
with
a
view
to
pro
vid
e
a
n
ov
e
rall
r
at
ing
of
a
poli
cy
su
ccess
or
fail
ur
e
.
Figure
1.
Ma
jo
r
p
hase
s
of
a
p
olicy
l
ife
c
yc
le
m
od
el
2.1.
Opinion
mini
ng
The
nom
enclatu
re
of
the
te
r
m
op
inion
m
ining
is
com
pr
is
ed
of
tw
o
te
r
m
s
op
inio
n
a
nd
mining
t
hat
i
m
plies
m
inin
g
of
op
i
nion.
Th
us
,
op
i
nion
m
ining
[
13
-
15
]
rep
re
sents
a
sy
m
bo
li
c
m
ea
ning
f
or
ext
rac
ti
on
of
public
sentim
e
nts
ov
e
r
va
rio
us
sub
j
ect
m
a
tt
ers
of
di
ver
s
ifie
d
dom
ai
ns
.
It
is
a
pr
ocess
of
capt
ur
i
ng
an
d
cat
egorizi
ng
th
e
per
ce
ption
of
public
about
a
par
ti
cular
pr
oduct,
pro
j
ect
,
phen
om
eno
n
or
a
pr
op
os
al
.
It
play
s
a
vital
r
ole
wi
th
a
view
to
e
ns
ure
ci
ti
zen
par
ti
ci
patio
n
(
a)
for
t
he
pro
gr
ess
of
pu
blic
act
ivit
ie
s
w
hi
ch
are
directl
y
or
in
di
rectl
y
i
m
pact
i
ng
t
heir
qu
al
it
y
of
li
fe
an
d
(
b)
t
o
receive
t
he
fee
dback
re
gardin
g
a
ny
product
,
top
ic
or
eve
nt
for
im
pr
ov
isi
ng
their
fu
t
ur
e
c
ourse
of
act
io
n.
[
1,
13
-
16
]
represents
th
e
thre
e
dim
ension
s
(
ta
sk
s
,
te
chn
iq
ues
an
d
ap
plic
at
ion
s
)
of
opini
on
m
in
ing
.
T
he
s
pecial
iz
at
ion
of
al
l
three
dim
ension
s
il
lustrate
s
that
al
l
of
t
hem
are
interrelat
ed
a
nd
c
an
be
us
e
d
in
any
com
bin
at
ion
t
o
achie
ve
op
i
nion
m
ining
.
F
or
exam
ple,
out
of
var
i
ou
s
avail
ab
le
te
chn
iq
ues
{
m
achine
le
ar
nin
g
(ML
),
le
xic
on
base
d
(LB)
,
hybri
d
te
c
hn
i
ques
(
HT
),
onto
log
y
base
d
(
OB),
c
on
te
xt
base
d
(
CB
)}
any
of
them
cou
ld
be
us
e
d
to
ex
plor
e
op
i
nion/senti
m
ents
by
perf
or
m
ing
var
i
ou
s
tasks
(
su
bject
ivit
y
cl
assifi
cat
ion
,
se
nti
m
ent
cl
assifi
cat
ion
,
rev
ie
w
m
easur
es,
s
pa
m
detect
ion
,
le
x
ic
on
form
ation
,
featur
e
sel
ect
io
n
a
nd
so
on
)
for
any
avail
able/
pro
bab
le
applicati
on
area
(Gov
e
r
nm
ent,
Ma
rk
et
,
Business
,
Sm
art
So
ci
et
y
Se
r
vices,
Inf
or
m
a
ti
on
Sec
uri
ty
&
A
naly
sis,
S
ub
Com
ponent
Tech
no
l
og
y
et
c).
To
acc
om
plish
the
goal
of
thi
s
pa
per,
c
oupli
ng
of
t
wo
dim
ensio
ns
i.e
.
te
c
hn
i
qu
e
s
a
nd
a
ppli
cations
of
opini
on
m
ini
ng
has
be
en
done
as
in
dex
e
d
i
n
T
a
bl
e
2.
It
util
iz
es
m
achine
le
arn
in
g
base
d
a
ppr
oac
hes
for
opinio
n
m
ining
in
t
he
i
m
ple
m
entat
ion
sphere of
go
ve
rn
m
ental
pr
ac
ti
ces
with
the
s
pecial
iz
at
ion
of
sup
e
rv
ise
d
m
achin
e
le
arn
in
g
te
c
hn
i
qu
e
s in
poli
cy
ev
al
uatio
n
[
17
-
2
0
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol. 11,
No.
1,
Febr
uar
y
2021 :
48
9
-
49
7
492
Table
2
.
C
oher
ence
of
t
w
o
d
im
ension
s
w
it
h
their
s
pecial
iz
at
ion
OM
Techn
iq
u
e(s)
OM
Ap
p
licatio
n
(s)
Machin
e L
e
arnin
g
Based
Go
v
ern
m
en
t
Su
p
ervis
ed
M
achi
n
e L
earnin
g
Po
licy
Evalu
atio
n
2.2.
Sma
r
t
ci
ties,
p
olicy e
va
lu
ati
on a
nd opini
on mi
ning
Ci
ti
es
in
country
li
ke
In
dia
a
re
gove
rn
e
d
by
m
ult
iple
or
ga
nizat
ion
s
a
nd
authorit
ie
s.
The
dif
fer
e
nt
sp
at
ia
l
entit
ie
s
with
m
ulti
ple
bounda
ries
deter
ef
fecti
ve
pla
nn
i
ng
an
d
gove
rn
a
nce.
Th
us,
to
r
eal
iz
e
the
sm
art
ci
ty
-
s
m
art
natio
n
m
issi
on
,
te
chnolo
gy
w
hic
h
entai
ls
&
ca
ptures
ef
fecti
ve
ci
ti
zen
par
ti
ci
pation
is
des
ired.
“Peo
ple
-
ce
ntri
c”
strat
egic
te
chnolo
gy
com
pone
nts
are
im
per
at
ive
to
create
sm
art
o
utcom
es
fo
r
c
it
iz
ens.
Po
li
cy
evaluati
on
is
an
ef
fectual
an
d
em
piri
cal
inv
est
igati
on
process
t
o
a
ppraise
t
he
perf
or
m
ance
of
a
poli
cy
by
m
easur
ing
it
s
wo
r
k
pr
act
i
ces
in
a
real
ti
m
e
env
ir
on
m
ent.
T
his
proce
ss
com
par
es
the
pr
ogress
c
urve
of
a
po
li
cy
over
ti
m
e
by
answ
e
ring
var
i
ous
que
sti
on
s
s
uch
a
s
exam
ining
the
accom
plish
m
e
nt
of
goal
s,
c
he
ckin
g
the
qual
it
y,
de
ci
din
g
f
uture
pros
pects
a
nd
m
easur
es
of
a
genda,
im
ple
m
enting
strat
e
gy
with
t
heir
e
xpect
ed
ou
t
pu
t
an
d
so
on.
It
is
a
com
plex
proce
ss
i
n
vie
w
of
it
s
ass
ociat
ion
with
num
ero
us
real
ti
m
e
con
st
raints
su
c
h
as
cost,
ti
m
e,
effor
t
,
s
o
c
i
a
l
a
n
d
l
e
g
a
l
c
o
n
s
i
d
e
r
a
t
i
o
n
,
r
e
s
o
u
r
c
e
o
p
t
i
m
i
s
a
t
i
o
n
a
n
d
m
u
c
h
m
o
r
e
.
A
p
o
l
i
c
y
m
a
y
b
e
e
v
a
l
u
a
t
e
d
b
o
t
h
i
n
inter
nal
as
w
el
l
as
extern
al
env
i
ronm
ent.
In
te
r
nal
evaluat
ion
incl
ud
es d
iffe
ren
t
act
ivit
ie
s
su
c
h
as
exec
utio
n
of
pre
-
set
pro
ce
dures
a
nd
m
eth
od
ologies
f
or
est
i
m
at
ing
the
pro
gr
es
s,
desi
gn
i
ng
a
te
am
of
well
trai
ned
a
nd
prof
essi
onal
ly
qu
al
ifie
d
pu
blic
of
fici
al
s,
pro
visio
n
of
a
ppr
opriat
e
log
ist
ic
s
(polic
y
ob
je
ct
ives
,
com
pleti
on
tim
e
li
ne,
re
qu
isi
t
e
resou
rces
et
c.)
to
e
valuat
or
s
wh
e
reas
e
xter
nal
evalu
at
ion
req
uires
c
oope
rati
on
and
pa
rtic
ipati
on
of
the
ta
r
ge
t
aud
ie
nc
e
in
t
his
crit
ic
al
exe
rcise
f
or
rati
ng
the
poli
cy
pa
ce.
Th
us
,
c
ons
ulti
ng
the
pu
blic
opin
ion
c
om
es
ou
t
to
be
a
n
es
sent
ia
l
and
fruit
fu
l
pr
act
ic
e
in
t
he
process
of
poli
cy
evaluati
on
wh
ic
h
fo
ste
rs
the
ne
e
d
of
i
n
d
u
c
t
i
n
g
o
p
i
n
i
o
n
m
i
n
i
n
g
i
n
p
o
l
i
c
y
e
v
a
l
u
a
t
i
o
n
f
o
r
a
c
o
m
p
r
e
h
e
n
s
i
v
e
d
e
v
e
l
o
p
m
e
n
t
o
f
a
c
o
u
n
t
r
y
a
t
t
r
a
c
t
i
n
g
p
e
o
p
l
e
&
i
n
v
e
s
t
m
e
n
t
s
.
3.
SMART EL
E
CTRIGO
V:
S
AU
BH
A
GYA
YOJN
A
E
VAL
UA
TIO
N U
SIN
G
OP
INION
MI
NI
NG
Saubha
gya,
or
Pr
a
dh
a
n
Ma
ntr
i
Saha
j
Bi
j
li
H
ar
Gh
a
r
Y
oj
a
na
,
is
a
soc
io
ce
ntric
se
rv
ic
e
la
un
c
he
d
on
Septem
ber
25,
2017
(
"Sa
ubha
gya
schem
e:
All
you
need
t
o
kn
ow
"
,
20
17
[
3,
4,
12)
to
m
ake
the
dre
a
m
of
"el
ect
rici
ty
fo
r
al
l"
a
reali
ty
.
T
his
go
vernm
ent
schem
e
aims
at
nationwide h
ouse
hold
el
ec
trific
at
ion
f
or
pe
op
l
e
welfar
e
by
m
a
king
t
heir
li
fe
qu
al
it
y
bette
r
.
The
gove
r
nm
e
nt
visions
the
schem
e
as
extr
e
m
el
y
ben
efici
al
in
the
interest
of
ci
ti
zen
con
si
der
i
ng
it
s
pos
it
ive
i
m
plicatio
ns
over
var
i
ou
s
sect
or
s
li
ke
ed
ucati
on,
healt
h,
agr
ic
ultur
e
, e
ntrepren
e
urshi
p et
c.
Und
oubtedly,
Saubha
gya
is
a
sta
irs
of
s
uc
cess
to
bri
ng
prom
inent
f
uture
opport
un
it
ie
s
f
or
I
ndia
.
Nev
e
rtheless
,
the
eval
uation
of
t
his
poli
cy
will
ref
le
ct
the
exact
sta
ti
sti
cs
of
public
reacti
on
a
nd
their
incli
nation
a
bo
ut the s
ch
em
e.
Hen
ce
, th
is
pa
per
i
nduces th
e
techn
iq
ues of
op
i
nion m
ining
in po
li
cy
ev
a
luati
on
with
a
view
t
o
optim
ise
the
process
e
xem
plifie
d
over
Sa
ubha
gya.
T
he
e
ntire
syst
em
sc
hem
a
is
sp
li
t
up
int
o
four
m
odules
as
represe
nted
i
n
Table
3.
T
he
execu
ti
on
of
m
od
ules
is
carried
out
in
the
fo
ll
owin
g
or
de
r:
D
at
a
Acquisi
ti
on,
P
re
-
processi
ng
of
data,
Cl
assifi
cat
ion
by
m
achine
le
ar
ni
ng
te
c
hn
i
ques
an
d
E
valuati
on
by
sta
nd
a
rd ef
fica
cy
m
easur
es
w
hich
a
re
discu
s
sed
i
n
s
ubseq
ue
nt secti
ons.
Table
3.
T
weet
c
ollec
ti
on
r
ec
ord o
n
d
ai
ly
b
a
sis
Ph
ase I
Ph
ase II
Ph
ase III
Ph
a
se IV
Date
Nu
m
b
e
r
o
f
Tweets
Date
Nu
m
b
e
r
o
f
Tweets
Date
Nu
m
b
e
r
o
f
Tweets
Date
Nu
m
b
e
r
o
f
Tweets
9
/1
1
/2
0
1
7
12
1
6
/1
1
/
2
0
1
7
59
2
3
/1
1
/
2
0
1
7
56
3
0
/1
1
/
2
0
1
7
65
1
0
/1
1
/
2
0
1
7
19
1
7
/1
1
/
2
0
1
7
177
2
4
/1
1
/
2
0
1
7
282
1
/1
2
/2
0
1
7
112
1
1
/1
1
/
2
0
1
7
4
1
8
/1
1
/
2
0
1
7
33
2
5
/1
1
/
2
0
1
7
75
2
/1
2
/2
0
1
7
11
1
2
/1
1
/
2
0
1
7
4
1
9
/1
1
/
2
0
1
7
17
2
6
/1
1
/
2
0
1
7
36
3
/1
2
/2
0
1
7
9
1
3
/1
1
/
2
0
1
7
37
2
0
/1
1
/
2
0
1
7
17
2
7
/1
1
/
2
0
1
7
3
4
/1
2
/2
0
1
7
2
1
4
/1
1
/
2
0
1
7
135
2
1
/1
1
/
2
0
1
7
31
2
8
/1
1
/
2
0
1
7
9
5
/1
2
/2
0
1
7
0
1
5
/1
1
/
2
0
1
7
76
2
2
/1
1
/
2
0
1
7
59
2
9
/1
1
/
2
0
1
7
16
6
/1
2
/2
0
1
7
27
7
/1
2
/2
0
1
7
38
8
/1
2
/2
0
1
7
50
9
/1
2
/2
0
1
7
4
3.1.
Acqu
isi
ti
on
of
data and i
ts
p
re
-
pr
ocessin
g
The
data
has
been
colle
ct
ed
by
m
eans
of
a
so
ci
al
netw
orki
ng
sit
e,
T
witt
er.
It
offer
s
a
platf
or
m
to
sh
are
vie
ws
ov
er
a
par
ti
cula
r
su
bject
m
at
te
r
by
sev
eral
us
e
rs
world
wide
wit
h
diff
e
re
nt
cultural,
ed
uca
ti
on
al
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
Su
st
aina
ble go
vern
an
ce
in
s
m
ar
t ci
ti
es and u
se o
f s
upervi
se
d
le
arni
ng bas
ed op
i
nion
mini
ng
(
Hena Iq
ba
l
)
493
and
s
ocial
bac
kgr
ounds.
To
that
en
d,
it
pr
ovides
a
global
ou
tl
oo
k
of
pu
bl
ic
senti
m
ents
and
their
incli
nation
towa
rd
s
any
m
at
te
r
of
co
nce
r
n,
t
op
ic
,
eve
nt
or
phen
om
ena
[
21
-
24
]
.
T
witt
er
sear
ch
API
(appli
cat
ion
pa
ckag
e
interfa
ce
)
has
been
us
e
d
f
or
the
ext
racti
on
of
Sa
ubha
gy
a
schem
e
tw
eet
s.
Fo
ll
owin
g
are
the
ste
ps
for
data acq
uisit
io
n:
-
An
a
ppli
cat
ion
util
isi
ng
the
gem
"t
witt
er"
is
dev
el
op
e
d
i
n
Ru
by
on
Ra
il
s.
Tweets
are
extracte
d
f
rom
twit
te
r
searc
h API by
us
i
ng a
rub
y i
nterf
ace
cal
le
d
twit
te
r
r
ub
y
gem
.
-
Re
gistrati
on
of the d
evel
oped a
pp
li
cat
io
n
is d
on
e
by r
ecei
vi
ng
t
he
acce
ss tok
e
n
i.e.
o
aut
h
cre
den
ti
al
s fo
r
cod
e
integ
rati
on.
-
So
m
e
of
the
ha
sh
ta
gs
us
e
d
in
this
pr
ocess
are:
#S
a
ubha
gyaSchem
e,
#Sau
bh
a
gyaY
oj
na,
#S
a
ubha
gy
a,
#S
a
ubha
g
ya
Plan.
-
Finall
y, the m
essages
co
ns
ist
of p
a
rtic
ular ke
ywords
r
et
urne
d by T
witt
er APIs
a
re c
ollec
te
d.
-
Total
co
unt
of
tweet
s
colle
ct
ed
durin
g
t
he
ti
m
el
ine
of
f
our
weeks
afte
r
the
la
un
c
h
of
sc
he
m
e
is
1262
a
nd
their
daily
stat
us
c
ount.
Pr
e
-
proces
sin
g
of
the
g
at
her
e
d
tweet
s
is
perform
ed
for
cl
e
an,
c
onsist
ent
and
cl
assifi
ed
data.
F
urt
he
r,
diff
e
re
nt
for
th
e
cl
assifi
cat
ion
of
norm
al
iz
ed
tweet
s
we
us
e
d
dif
fer
e
nt
m
a
chine
la
ngua
ge
.
Steps
in
vo
l
ve
d
in
the pr
ocedur
e
of pre
-
processi
ng are:
-
Ther
e
is
hu
ge
possibil
it
y
of
d
upli
cat
e
tw
eet
s
in
the
c
ourse
of
da
ta
colle
ct
ion
,
so
firstly
in
dat
a
pre
-
pr
ocessin
g we
need to
rem
ov
e
the
duplica
te
tweets.
-
Ther
e
are
m
any
sp
eci
al
cha
r
act
ers
$,#
,*
et
c,
an
d
num
ber
so
we
nee
d
t
o
discar
d
nu
m
ber
s
an
d
s
peci
al
char
act
e
rs
s
-
Rem
ov
e all
UR
L li
nk
s
an
d w
ords
li
ke
is, am
are, t
he
et
c.
-
Mi
ni
m
iz
ation
of cr
um
pled
w
ords
i
n
thei
r
r
oot f
or
m
by ste
m
m
ing
.
-
Feat
ur
e
selec
ti
on in res
ultant
data an
d
it
s cla
ssific
at
ion
of s
entim
ent polarit
y.
-
Lives,
P
oor
,
P
eop
le
,
Fam
iliarisa
ti
on
,
Sc
hem
e,
Co
nn
ect
ivit
y,
Ligh
ti
ng,
B
ulb
,
Sh
e
ddin
g,
Crack
dow
n
a
nd
so
on are
s
om
e
of the
ex
am
ples of select
ed
a
tt
ribu
te
s.
3.2.
Opinion
classi
ficat
i
on
based
o
n
mac
hine le
arnin
g techni
ques
In
t
his
pa
pe
r,
we
em
pirical
ly
analy
se
va
rio
us
sta
ndar
d
m
ac
hin
e
le
ar
ning
a
lgorit
hm
s
li
s
te
d
in
Table
4
.
And
preci
sion,
recall
and acc
uracy
[
13
]
li
ste
d i
n
T
a
ble
5.
Table
4.
Ma
c
hi
ne
l
ear
ning
t
ec
hn
i
qu
e
s
[
13
]
Na
m
e
of
T
echn
iq
u
e(s)
Descripti
o
n
Na
ive B
a
yesia
n
Part
o
f
a
class
of
sim
p
l
e pro
b
ab
ilistic
class
if
iers us
ed
f
o
r
th
e esti
m
a
tio
n
of
class
if
icatio
n
para
m
eters.
S
u
p
p
o
rt Vector
Ma
ch
in
e
Belo
n
g
s
to
th
e
cla
ss
o
f
d
iscri
m
in
ativ
e
su
p
ervis
ed
learnin
g
b
ased
class
if
ier
f
o
r
id
en
tif
icatio
n
o
f
class
if
icatio
n
pattern
.
It
classif
ies th
e
d
ata by
a
hyperp
la
n
e.
Mu
ltila
yer Pe
rcep
t
ro
n
Rep
resents
a
n
etw
o
rk o
f
neu
ron
s n
a
m
e
d
percept
ro
n
r
el
ated
to artif
icial ne
u
ral
n
etwo
rk.
k
-
Nea
res
t Neig
h
b
o
u
rs
Fu
n
d
a
m
en
tal
m
ac
h
in
e
learnin
g
alg
o
rith
m
th
at
d
o
es
n
o
t
m
ak
e
an
y
u
n
d
erly
in
g
ass
u
m
p
tio
n
s
ab
o
u
t
d
ata
d
istrib
u
tio
n
.
Here
,
class
if
icatio
n
o
f
o
b
jects
is
b
ased
o
n
th
e
v
o
tin
g
o
f
n
eig
h
b
o
u
rs
an
d
th
e
class
ass
ig
n
ed
to
th
e ob
ject is us
u
ally
a
m
o
n
g
its k n
ear
est n
eig
h
b
o
u
rs
[21]
Decisio
n
Tree
Tr
ee
m
o
d
el
o
f
d
eci
sio
n
s
an
d
th
eir
res
u
lts
u
sed
as
a
d
eci
sio
n
su
p
p
o
rt
to
o
l.
Dev
elo
p
ed
f
ro
m
to
p
to
b
o
tto
m
with
a
sin
g
le r
o
o
t
n
o
d
e at
th
e top
and
b
ranch
in
g
of
sev
e
ral
leaf
n
o
d
es with
p
rob
ab
le ou
tco
m
es
.
Ran
d
o
m F
o
res
t
Ad
v
an
ce
m
en
t of
decis
io
n
tr
ees to
get
m
o
re
sp
ecif
ic r
esu
lts.
Lin
ea
r Regr
ess
io
n
Linear
ap
p
roach
u
sed
for
p
redictiv
e
an
aly
sis
to
id
en
tify
wh
eth
er
p
redicto
r
v
ariables
are
ab
le
to
p
redict
d
ep
en
d
en
t variab
les an
d
whi
ch
variabl
e ar
e
sig
n
if
ican
t p
redicto
rs of
d
ep
en
d
en
t variab
les.
K
-
sta
r
Ins
tan
ce based
clas
sif
ier
d
eter
m
in
es
sim
i
larit
y
f
u
n
ctio
n
s b
y
us
in
g
entro
p
y
bas
ed
dis
tan
ce fu
n
ct
io
n
Bag
g
in
g
It
i
m
p
rov
es class
if
icatio
n
by
co
m
b
in
i
n
g
r
an
d
o
m
gen
er
at
ed
tr
ain
in
g
sets.
Ada
b
o
o
st
Ad
ap
tiv
e
b
o
o
sting
,
m
a
ch
in
e
learnin
g
m
eta
alg
o
rith
m
.
A
stro
n
g
class
i
f
ier
h
as
b
een
d
ev
elo
p
ed
b
y
co
m
b
in
in
g
vario
u
s weak class
if
iers to
i
m
p
rov
e perf
o
r
m
a
n
ce.
Table
5.
Stan
da
rd
p
e
rfor
m
ance
i
nd
ic
at
or
s
[
13
]
Na
m
e
Descripti
o
n
Precisio
n
Co
rr
ectn
ess
in
m
e
a
su
r
in
g
the ratio o
f
tr
u
e
to
pred
icted
po
sit
iv
es
Recall
Co
m
p
leten
ess
in
m
easu
ring
the ratio
o
f
tr
u
e to actual p
o
sitiv
es
Accura
cy
Prop
o
rtion
of
tr
u
e r
esu
lts v
s th
e total
n
u
m
b
er
of
cases e
x
a
m
in
ed
3.3.
Polic
y
e
valua
t
ion based
on
opi
nion mi
ning
Sentim
ent
real
i
zat
ion
an
d
their
cl
assifi
cat
ion
in
colle
ct
ed
t
weet
is
do
ne
by
app
ly
ing
opi
nion
m
ining
te
chn
iq
ues
[
14
]
.
The
con
ce
pt
of
inco
r
porati
ng
th
oughts
or
viewpoint
of
s
ta
keholde
rs
an
d
en
d
us
e
rs
in
po
li
cy
evaluati
on
give
rise
to
s
how
case
the
pr
eci
se
portrayal
of
po
l
ic
y
pe
rfo
r
m
ance.
As
a
s
te
p
f
orward
t
o
this,
op
i
nion
ass
oci
at
ed
with
t
he
c
ollec
te
d
tweet
s
of
Sa
ubha
gya
schem
e
has
be
en
m
ined
an
d
cat
egorized
i
n
three
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol. 11,
No.
1,
Febr
uar
y
2021 :
48
9
-
49
7
494
cl
asses:
po
sit
ive
(P),
neg
at
i
ve
(
N)
a
nd
ne
utral
(
Nu).
T
he
daily
sta
tus
repor
t
of
twee
ts
cl
assifi
cat
ion
in
a
ll
the fo
ur
phases
is rec
orde
d
in
Table
6.
Table
6
.
Distri
bu
ti
on
of
o
pi
nio
n
p
olarit
y o
ve
r
f
our
p
hases
of
d
at
a
a
cq
uisit
ion
Date
N
Nu
P
Total
Date
N
Nu
P
Total
Ph
ase I
Ph
ase II
9
/1
1
/2
0
1
7
0
4
7
11
1
6
/1
1
/
2
0
1
7
1
5
37
43
1
0
/1
1
/
2
0
1
7
0
4
16
20
1
7
/1
1
/
2
0
1
7
2
24
19
45
1
1
/1
1
/
2
0
1
7
0
3
2
5
1
8
/1
1
/
2
0
1
7
11
6
10
27
1
2
/1
1
/
2
0
1
7
0
2
1
3
1
9
/1
1
/
2
0
1
7
3
3
6
12
1
3
/1
1
/
2
0
1
7
0
3
29
32
2
0
/1
1
/
2
0
1
7
2
1
14
17
1
4
/1
1
/
2
0
1
7
0
2
133
135
2
1
/1
1
/
2
0
1
7
1
5
21
27
1
5
/1
1
/
2
0
1
7
0
9
66
75
2
2
/1
1
/
2
0
1
7
1
14
32
47
Total
0
27
254
281
Total
21
58
1
39
218
Ph
ase III
Ph
ase IV
2
3
/1
1
/
2
0
1
7
7
24
21
52
3
0
/1
1
/
2
0
1
7
3
45
4
52
2
4
/1
1
/
2
0
1
7
5
30
242
277
1
/1
2
/2
0
1
7
26
80
5
111
2
5
/1
1
/
2
0
1
7
7
17
49
73
2
/1
2
/2
0
1
7
0
5
0
8
2
6
/1
1
/
2
0
1
7
4
2
30
36
3
/1
2
/2
0
1
7
2
4
2
8
2
7
/1
1
/
2
0
1
7
0
0
3
3
4
/1
2
/2
0
1
7
0
1
1
2
2
8
/1
1
/
2
0
1
7
1
0
6
7
5
/1
2
/2
0
1
7
0
0
0
0
2
9
/1
1
/
2
0
1
7
2
1
12
15
6
/1
2
/2
0
1
7
3
0
24
27
Total
27
74
362
463
7
/1
2
/2
0
1
7
1
31
5
38
8
/1
2
/2
0
1
7
0
45
5
50
9
/1
2
/2
0
1
7
1
2
1
³
Total
37
213
50
300
The
sta
tus
of
posit
ive,
neg
at
i
ve
a
nd
ne
utral
tweet
s
of
Sa
ubhag
ya
Y
ojn
a
a
cro
ss
phases
is
dep
ic
te
d
i
n
F
igure
2
.
T
he
canv
a
s
re
pr
e
se
nts
an
ap
par
e
nt
po
sit
ive
de
viati
on
of
pe
op
l
e
towa
rds
the
schem
e
with
90%
of
po
sit
ive
a
nd
10%
of
ne
utral
tweet
s
duri
ng
ph
a
se
one.
N
o
in
div
i
du
al
figure
of
ne
gative
tweet
has
be
e
n
confro
nted
i
n
t
his
ph
ase
o
ut
of
281
c
ollec
te
d t
weets.
Figure
2
.
Stat
us o
f
p
os
it
ive
,
n
egati
ve
a
nd
n
e
utral tweet
s
of
s
au
bh
a
gya
y
oj
na
ac
ro
s
s
ph
as
es
The
facts
an
d
record
s
ref
le
ct
s
a
con
trast
in
g
reacti
on
of
ci
ti
zen
in
second
ph
ase,
al
th
ough
the
sta
ts
represe
nt
63%
of
posit
ive
tw
eet
s,
9.6%
of
ne
gative
twe
et
s
and
26.
6%
of
ne
utral
tw
eet
s
sti
ll
resu
lt
ing
i
n
m
axi
m
al
cou
nt
of
po
sit
ive
tweet
s.
Ph
a
se
three
rec
orde
d
an
identic
al
ph
e
nom
eno
n
of
earli
er
ph
a
se
with
a
highest
per
c
entage
of
posit
ive
tweet
s
that
is
78%
f
ollo
w
ed
by
16%
of
neu
t
ral
an
d
6%
of
ne
gative
tweet
s.
An
une
xp
e
ct
ed
rise
i
n
per
ce
nt
age
of
ne
utral
tweet
s
with
71%
is
obse
rv
e
d
in
ph
ase
f
ou
r
c
on
t
rar
y
t
o
posit
ive
tweet
s
with
17%
an
d
ne
gativ
e
tweet
s
with
12%.
T
he
a
bru
pt
fluct
uation
in
the
fig
ur
es
of
ne
utral
vs
.
posit
iv
e
tweet
s
is
on
ac
count
of
m
axi
m
u
m
par
ti
ci
pat
ion
of
pu
blic
in
e
xpressi
ng
their
t
houghts
i
n
the
init
ia
l
ph
ases
of
la
un
c
h
of
sc
he
m
e.
Howe
ver
,
a
la
rg
e
num
ber
of
inf
orm
ational
tweet
s
has
be
en
po
ste
d
by
m
edia
or
ci
vic
in
la
st
ph
a
se c
on
ce
rn
i
ng the m
os
t rec
ent m
easur
es f
i
naliz
e
d for t
he po
li
cy
ex
e
cutio
n.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
Su
st
aina
ble go
vern
an
ce
in
s
m
ar
t ci
ti
es and u
se o
f s
upervi
se
d
le
arni
ng bas
ed op
i
nion
mini
ng
(
Hena Iq
ba
l
)
495
4.
RESU
LT
S
AND FI
ND
I
NGS
Ov
e
r
a
co
un
t
of
1262
tweet
s
colle
ct
ed
fro
m
Twitt
er,
a
set
of
te
n
sup
erv
ise
d
m
achi
ne
le
arn
i
ng
al
gorithm
s
hav
e
been
a
pp
li
ed
to
fin
d
the
pol
arit
y
of
opinio
n
f
or
this
sc
he
m
e
[
25
]
.
SV
M
,
kNN,
MLP,
R
F,
LR,
DT,
K
-
sta
r,
B
agg
i
ng,
NB
a
nd
Ad
a
boos
t
ar
e
the
a
ppr
oaches
us
e
d
f
or
t
he
cal
culat
ion
of
sta
ndar
d
e
valuati
on
m
easur
es
a
nd
t
he
resp
e
ct
ive
e
xp
e
rim
ental
resu
lt
s
f
or
acc
uracy
,
preci
sio
n
and
recall
a
re
ind
e
xed
in
Ta
ble
7.
Com
par
at
ive
analy
sis
of
s
upe
rv
ise
d
m
achine
le
arn
in
g
al
go
rithm
s
has
bee
n
car
ried
ou
t
t
o
dete
rm
ine
the
best
cl
assifi
er
f
or
c
at
egorizi
ng
se
nt
i
m
ents
a
s
dire
ct
qu
a
ntifie
rs
of
po
li
cy
.
F
ollo
wing
obser
vations
hav
e
bee
n
li
ste
d
in v
ie
w of ab
ove
resu
lt
s:
-
The
hi
ghest
accuracy
ha
s
bee
n
achie
ved
by
Suppor
t
Vecto
r
Ma
chi
ne
with
91.
77%
an
d
Ad
a
boos
t
bei
ng
the lea
st wit
h 7
2.41%.
-
kNN
a
nd
MLP
r
eflect
s en
c
ourag
i
ng
resu
lt
s, bo
t
h
with
91.
47% accu
racy b
ut kNN
is
bein
g
bette
r
in te
r
m
s
of preci
sio
n wit
h
91.
6%
and
MLP wit
h 9
1%
.
-
Nex
t
le
vel
ha
s
bee
n
ac
hieve
d
by
Ra
ndom
Forest
with
a
n
acc
ur
acy
of
91.
07
%
f
ollo
wed
by
Li
nea
r
Re
gr
essi
on w
it
h 90.87%
.
-
Accuracy
of
D
eci
sion
T
ree is
90.67%
ch
ase
d by
K
-
sta
ir
w
it
h 89.97%
.
-
Ba
gg
i
ng r
ec
ords
acc
ur
acy
of
88.36%
foll
owed by Nai
ve
B
ay
es w
it
h
a
coun
t
of
77.73%.
Table
7
.
Stat
ist
ic
s o
f
s
ta
ndar
d
e
ff
ic
acy
m
easur
es
for
m
achine
l
earn
i
ng
t
ech
niques
ML
Techn
iq
u
e(s)
Precisio
n
Recall
Accurac
y
Su
p
p
o
rt
Vector M
achi
n
e
9
1
.5
9
1
.8
9
1
.77
K
-
n
eare
st
neig
h
b
o
u
r
9
1
.6
9
1
.5
9
1
.47
Multilaye
r
Pe
rcept
ron
91
9
1
.5
9
1
.47
Ran
d
o
m
For
est
9
1
.2
9
1
.1
9
1
.07
Linear Regr
ess
io
n
9
0
.5
9
0
.9
9
0
.87
Decisio
n
T
ree
9
0
.5
9
0
.7
9
0
.67
K
-
star
90
90
8
9
.97
Bag
g
in
g
8
7
.7
8
8
.4
8
8
.36
Naiv
e Bay
es
8
4
.1
7
7
.7
7
7
.73
Ad
ab
o
o
st
6
8
.7
7
2
.4
7
2
.41
Am
on
g
al
l,
SV
M
excee
ds
in
al
l
the
par
a
m
et
ers
of
accuracy,
pr
eci
sio
n
an
d
recall
.
The
picto
ria
l
represe
ntati
on
of
rel
at
ive
perform
ances
of
eff
ic
acy
m
eas
ur
es
for
the
a
bove
m
achine
le
arn
i
ng
al
gorithm
s
is
dep
ic
te
d i
n
F
ig
ur
e
3
.
Figure
3
.
Stan
da
rd
e
f
ficacy
m
easur
e
for
m
ac
hin
e
l
ea
rn
i
ng
t
echn
i
qu
e
s
5.
CONCL
US
I
O
N
As
a
com
m
on
m
an
or
the
ta
r
ge
t
aud
ie
nce
is
the b
eh
olde
r
of a
po
li
cy
,
he
nce
co
ns
ide
rin
g
t
he
ir
ideas
or
views
in
the
e
valuati
on
proc
ess
is
essenti
al
ly
req
uire
d.
T
he
pe
op
le
pe
rs
pecti
ve
hel
ps
the
poli
cy
m
ak
ers
in
decidin
g
t
he
f
ut
ur
e
pros
pects,
correct
ive a
nd
pr
e
ve
ntive m
ea
su
res
r
ec
omm
e
nd
e
d for
po
li
cy
in
or
der
t
o
m
a
ke
it
a
su
ccess.
T
he
refor
e
,
the
aim
of
this
pa
per
was
to
induct
the
co
ncep
t
of
op
i
nion
m
ining
for
poli
cy
evaluati
on
as
a
sm
art
ci
ty
te
chn
ol
ogy
so
luti
on
f
or
s
us
ta
ina
ble
gove
rn
a
nce.
T
o
e
xem
plify
the
m
od
el
,
a
la
te
st
gove
rn
m
ental
s
c
h
e
m
e
,
P
r
a
d
h
a
n
M
a
n
t
r
i
S
a
h
a
j
B
i
j
l
i
H
a
r
G
h
a
r
Y
o
j
a
n
a
o
r
S
a
u
b
h
a
g
y
a
(
H
o
u
s
e
h
o
l
d
E
l
e
c
t
r
i
c
i
t
y
s
e
r
v
i
c
e
)
i
s
c
o
n
s
i
d
e
r
e
d
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol. 11,
No.
1,
Febr
uar
y
2021 :
48
9
-
49
7
496
Gove
rn
m
ent
Int
el
li
gen
ce
le
ve
rag
es opinio
n
m
ining
to
i
nclu
de
ci
ti
zen
in
put
to
enhance
t
he
po
li
cy
an
d
decisi
on
-
m
aking
cy
cl
e.
It h
el
ps
go
vernm
ents
and
org
a
nizat
ion
s
in r
es
ourc
e
m
anag
e
m
ent,
stren
gth
e
ns
po
li
ti
ca
l
m
e
a
s
u
r
e
s
/
m
e
c
h
a
n
i
s
m
s
,
r
e
v
o
l
u
t
i
o
n
i
z
e
s
t
h
e
s
t
y
l
e
o
f
d
e
l
i
v
e
r
y
o
f
b
a
s
i
c
s
e
r
v
i
c
e
s
t
o
e
s
t
a
b
l
i
s
h
a
f
o
c
u
s
e
d
m
o
d
e
l
o
f
g
o
v
e
r
n
a
n
c
e
f
o
r
p
e
o
p
l
e
.
T
h
e
s
c
o
p
e
o
f
g
o
v
e
r
n
m
e
n
t
i
n
f
o
r
m
a
t
i
o
n
i
n
c
l
u
d
e
s
s
t
r
a
t
e
g
y
,
c
o
n
s
u
l
t
a
t
i
v
e
p
a
r
t
i
c
i
p
a
t
i
o
n
,
l
o
b
b
y
i
n
g
e
t
c
.
The
sc
hem
e
is
evaluate
d
by
a
naly
zi
ng
publi
c
sentim
ents
extracte
d
f
ro
m
t
weets
by
a
pply
ing
opinio
n
m
ining
te
c
hn
i
qu
e
s.
A
s
et
of
super
vise
d
m
achine
le
arn
i
ng
te
c
hn
i
qu
es
ha
ve
bee
n
a
pp
li
ed
a
nd
c
om
par
ed
on
the
ba
sis
of
ac
cur
acy
,
preci
sion
an
d
recall
nam
ely,
suppo
rt
vect
or
m
achine
,
k
-
near
e
st
neig
hbour
,
m
ulti
layer
per
ce
ptr
on,
ra
ndom
fo
rest,
li
near
regressio
n,
decisi
on
tr
ee,
K
-
sta
r,
B
ag
gi
ng,
naive
bayes
an
d
a
da
boost
.
Fo
ll
owin
g
co
nc
lusio
ns
ha
ve
been
dr
a
wn
ba
sed
on
this
res
earch
w
ork:
(a
)
63%
of
the
tweet
s
com
es
ou
t
to
be
po
sit
ive
sig
nifi
es a f
a
vourable
r
es
pons
e
of p
e
op
le
t
ow
a
r
ds
t
his polic
y, (
b)
30% of t
he
twe
et
s ar
e
neu
tr
al
w
hic
h
include
s
lot
of
inf
or
m
at
ion
al
tweet
s
poste
d
by
ci
vic
or
m
edia
reg
a
rd
i
ng
la
te
st
upda
te
s
of
sc
hem
e
or
it
s
i
m
ple
m
entat
io
n
pr
ocess,
(c)
7%
of
t
h
e
tweet
s
res
ults
in
ne
gative
,
(
d)
SV
M
ou
t
pe
rfor
m
s
a
m
ong
al
l
the
im
ple
m
ented
s
up
e
r
vised
m
achine
le
ar
nin
g
te
c
hniq
ues
with
a
n
acc
uracy
of
91.
77
,
(e)
kNN
a
nd
MLP
sh
are
s
the
nex
t
le
vel
with
an
accuracy
of
91.47
f
ollo
wed
by
RF,
LR,
DT
,
K
-
sta
r,
Ba
ggin
g,
NB
an
d
A
da
boos
t
,
(f)
A
da
boos
t
be
ing
t
he
lo
west
w
it
h
a
n
acc
ura
cy
o
f
72.4
1
The
res
ults
are
pr
om
isi
ng
an
d
validat
e
the
use
of
opini
on
m
ining
as
a
sm
art
te
chn
olog
y
so
luti
on
f
or
gove
rn
m
ent
poli
cy
evaluati
on.
The
re
is
a
hu
ge
sco
pe
of
e
va
luati
ng
m
or
e
po
li
ci
es
to
buil
d
a
sm
art
ci
t
y
too
l.
More
ov
e
r,
the
process
of
eval
uation
ca
n
be
e
nh
a
nce
d
by
im
pro
ving
the
ta
sk
of
sentim
ent
cl
assifi
cat
ion
usi
ng
so
ft c
om
pu
ti
ng &
e
vo
l
utio
nary
techn
iq
ues
.
ACKN
OWLE
DGE
MENTS
We
w
ou
l
d
li
ke
to
than
k
our
c
ollea
gu
e
s
f
or
pro
vid
i
ng
the
i
nsi
gh
t
an
d
e
xp
e
rtise
that
gr
eat
ly
helped
in
wr
it
in
g
this
co
m
ple
te
research
arti
cl
e.
Also
we
wan
t
to
s
upport
our
re
s
pecti
ve
insti
tuti
on
s
w
he
re
we
are
work
i
ng wh
o
a
re c
on
sta
ntly
m
ot
ivati
ng
t
o
s
ub
m
it
r
esearch
artic
le
s of this
h
ig
h st
an
dards
.
REFERE
NCE
S
[1]
M.
P.
S.
Bhat
i
a
and
A.
K.
Khalid,
“A
primer
on
the
web
informa
ti
on
ret
r
ie
v
al
par
adi
gm
,”
Journal
of
Theoreti
cal
&
Appl
ie
d
Informa
ti
on
Te
chnol
ogy
,
vol.
4
,
no.
7
,
pp.
657
-
662,
2008.
[2]
M.
P.
S.
Bhat
i
a
and
A.
Kum
ar,
“
Para
digm
shifts:
from
pre
-
web
informati
on
s
y
st
e
m
s
to
rec
ent
we
b
-
base
d
contextu
al
informati
on
r
et
r
i
eva
l
,”
We
bolog
y
,
vol
.
7
,
no
.
1
,
pp
.
1
-
7
,
2010
.
[3]
Utpal
Bhaskar
“
Saubhag
y
a
sche
m
e:
All
y
ou
ne
ed
to
know
,”
live
mint
,
2017.
[
Online
]
.
Available:
htt
p://ww
w.l
ivem
int
.
com/Indust
r
y
/6
agWD
5oCYBi6jokfKc3QjP/
Saubhag
y
a
-
sche
m
e
-
All
-
y
ou
-
n
eed
-
to
-
know.ht
m
l
.
[4]
“
Policy
,
”
[Onlin
e]
.
Available:
h
ttps
:/
/e
n
.
wikipe
d
i
a.
org/wik
i/
Poli
c
y
.
[5]
A.
Kum
ar
and
T
.
M.
Sebasti
an,
“
Senti
m
ent
ana
l
y
sis
on
twit
te
r
,”
Inte
rnational
Jo
urnal
of
Comput
er
Sci
ence
Iss
ues
(
IJCSI
)
,
vol.
9
,
n
o.
4,
pp
.
372
-
37
8,
2012
.
[6]
“
The
Po
li
c
y
m
ak
ing
Proce
ss
,
”
Cl
if
fsnote
s
.
[Onlin
e]
.
Available:
ht
tps:/
/www
.
cliffsnote
s.c
om
/stud
y
-
guide
s/ameri
can
-
gover
nm
ent
/pub
li
c
-
pol
icy
/
the
-
po
li
c
y
m
aki
ng
-
pro
c
ess
.
[7]
“
The
publi
c
poli
c
y
dev
e
lopment
c
y
cle
,”
U.S.
F
ire
Admini
stratio
n
,
2017.
[Online
]
.
Availa
ble
:
htt
ps://
ww
w.usfa.
fema
.
gov/t
r
ai
n
ing/
cof
f
ee
_br
ea
k
/053117.
html
.
[8]
“
Understa
nding
the
politic
s
o
f
publi
c
pol
icy
,”
Stage
s
of
Policy
Dev
el
opm
ent
,
2013
.
[Online
]
.
Availabl
e
:
htt
p://us
ers.
han
c
ock.
ne
t/
l
enn
y
esq
/c
hap
te
r%2003
.
p
df
.
[9]
“
Policy
,
”
[Onlin
e]
.
Available:
h
ttps
:/
/e
n
.
oxfor
ddictiona
ri
es.
com/d
e
fini
ti
on
/pol
i
c
y
.
[10]
“
Mining,
”
[Onl
i
ne]
.
Available:
h
tt
ps://
en.
wik
ipe
d
ia
.
org
/wiki
/Min
i
ng
.
[11]
“
Opinion,
”
[Onl
i
ne]
.
Available:
h
tt
ps://
en.
oxfordd
ic
ti
on
aries.c
om
/
def
ini
t
ion/
opin
io
n
.
[12]
Paul
C.
,
“
Policy
concept
s
in
1000
words
:
T
he
polic
y
c
y
cle
and
i
ts
stag
es,
”
2013
.
[On
li
ne]
.
Avai
la
bl
e
:
htt
ps://
p
aul
c
ai
rn
e
y
.
wordpress
.
co
m
/2013/
11/11/
p
oli
c
y
-
conc
ep
ts
-
in
-
1000
-
words
-
the
-
polic
y
-
c
y
c
le
-
a
nd
-
its
-
stage
s
.
[13]
A.
Kum
ar
and
A.
Jaiswal,
“
Em
piri
c
al
Stud
y
of
Twit
ter
and
Tu
m
blr
for
Senti
m
ent
Anal
y
sis
usi
ng
Soft
Com
puti
ng
Te
chn
ique
s
,”
i
n
Proce
ed
ings o
f
t
he
World
Congr
ess on
Engi
n
ee
ri
ng
and
Compute
r Sc
ie
n
ce
,
v
ol
.
1
,
pp.
1
-
5,
2017.
[14]
A.
Kum
ar
and
A.
Sharm
a,
“
Para
digm
Shifts
from
E
-
Governa
nc
e
to
S
-
Governa
n
ce
,”
The
Hum
an
El
eme
nt
o
f
Bi
g
Data:
Iss
ues,
An
aly
tics,
and
P
erf
orm
ance
,
p
p
.
21
3
-
234,
2016
.
[15]
A.
Kum
ar
and
A.
Sharm
a,
“
Sy
stemati
c
Li
t
era
t
ure
Review
on
Opinion
Minin
g
of
Big
Data
for
Governm
ent
Inte
lligen
ce
,”
W
ebol
ogy
,
vo
l.
14
,
no.
2
,
pp.
6
-
47,
2017.
[16]
K.
Ravi
and
V.
Ravi
,
“
A
surve
y
on
op
i
nion
m
ini
ng
a
nd
senti
m
ent
ana
l
y
sis:
t
asks,
appr
oac
h
es
a
nd
appl
i
ca
t
ions
,”
K
nowle
dge
-
Based
Syste
ms
,
vol.
89
,
pp
.
14
-
46
,
201
5
.
[17]
H.
L
.
Har
tman,
“
SM
E
Mining
En
gine
er
ing
Handb
ook
,
”
So
ciety f
or
Mini
ng
,
M
et
al
lu
rgy,
and
Ex
p
lorati
on
In
c
,
1992
.
[18]
“
Data
base
R
enew
abl
e Ene
rg
y
&
Mining:
W
ind &
Solar,
”
TH
ENERG
Y
sus
t
ainable
consulting
.
[19]
K.
Dave
,
e
t
al.,
“
Mining
the
peanut
gal
l
er
y
:
Opi
nion
ext
ra
ction
and
sem
ant
ic
class
ifi
cation
of
p
roduc
t
rev
ie
ws
,
”
i
n
Proc
ee
dings
o
f
th
e
12th
Int
ernati
onal
Con
fe
ren
ce
on
World
W
id
e
We
b
(
WWW)
,
pp.
519
-
528
,
20
03
.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
Su
st
aina
ble go
vern
an
ce
in
s
m
ar
t ci
ti
es and u
se o
f s
upervi
se
d
le
arni
ng bas
ed op
i
nion
mini
ng
(
Hena Iq
ba
l
)
497
[20]
J.
Rea
d
,
“
Us
ing
emotic
ons
to
r
educ
e
depe
nd
en
c
y
in
m
a
chi
ne
l
ea
rning
techniq
ues
for
senti
m
e
nt
class
ifi
c
at
ion
,
”
i
n
Proc
ee
dings
of
ACL
-
05
,
43
nd
Me
e
ti
ng
o
f
the
Associat
ion
for
Computational
Linguisti
cs.
Associa
ti
on
fo
r
Computati
onal
L
ingui
stic
s
,
pp
.
4
3
-
48,
2005
.
[21]
“
W
ordNet
,”
[On
li
ne]
.
A
va
il
ab
le
:
htt
p://wordnet
.
pr
inc
e
ton.
edu
/
.
[22]
A.
Agarwal
,
e
t
a
l.
,
“
Senti
m
ent
A
naly
s
is
of
Twit
t
e
r
Data
,
”
i
n
Proc
ee
dings
of
th
e
A
CL
2011
Worksh
op
on
Languages
in
Soc
ial
M
edi
a
,
pp.
30
-
38,
2011
.
[23]
A.
Pak
and
P.
Paroube
k
,
“
Twitte
r
as
a
Corpus
for
Senti
m
e
nt
Ana
l
y
sis
an
d
Opinio
n
Mining
,
”
i
n
Proce
ed
ings
of
the
Sev
en
th
Conf
ere
nce
on
Int
ernational
Language
Re
sour
ce
s and
E
val
uati
on
,
pp
.
13
20
-
1326,
2010
.
[24]
A.
Kum
ar
and
T.
M.
Sebast
ia
n
,
“
Mac
hine
l
ea
r
ning
assisted
Se
nti
m
ent
Anal
y
si
s
,
”
Proceedi
ngs
of
Inte
rnationa
l
Confe
renc
e
on
C
o
mputer
Scienc
e
&
Engi
nee
ring (
ICCSE’2012)
,
pp.
123
-
130
,
201
2
.
[25]
V.
Hatziva
ss
il
og
lou
and
K.
McK
eown,
“
Predictin
g
the
sem
a
nti
c
orie
nt
at
ion
of
ad
je
c
ti
ves
,
”
i
n
Pro
ce
ed
ings
of
the
Joi
nt
ACL/E
ACL Conf
ere
n
ce
,
pp
.
174
-
181,
2004
.
BIOGR
AP
H
I
ES
OF
A
UTH
ORS
He
na
Iq
b
al
hav
e
comple
te
d
the
doct
ora
te
d
egr
e
e
(Ph.D.)
from
In
dia
in
th
e
y
e
ar
2
015
and
h
ave
bee
n
te
a
chi
ng
fo
r
the
past
m
an
y
y
e
ars
in
UA
E.
S
he
has
aro
und
1
0
y
ea
rs
of
Aca
d
emic
t
eachi
ng
expe
ri
ence.
As
a
te
a
che
r
,
her
m
a
in
goal
is
to
m
ot
iva
t
e
student
s
to
do
the
ir
best
an
d
ext
end
their
own
per
sonal
lim
it
s.
She
devi
ses
progra
m
s,
ac
cor
ding
to
the
s
y
l
la
bus
req
uire
m
e
nts,
tha
t
exp
and
on
pre
vious
kn
owledge
and
en
cour
age
studen
t
s
to
expl
ore
new
and
int
ere
sti
ng
poss
ibi
li
ti
es.
She
has
orga
ni
z
ed
var
ious
works
hops
and
sem
ina
rs
and
h
as
given
m
an
y
gu
est
l
e
ct
ure
s.
On
e
ac
h
occ
asion
she
ha
d
worked
and
m
ana
ged
to
demo
nstrat
e
excel
l
ent
pla
nning
,
com
m
unic
at
ion
and
te
am
work
skill
s.
She
is
ver
y
ac
t
i
ve
in
her
rese
arch
work.
She
has
m
ore
tha
n
10
rese
arc
h
pape
rs
publi
c
at
ions
in
ref
err
ed
journ
a
l
s
and
int
ern
at
io
nal
conf
er
enc
es
.
Her
rese
arc
h
are
as
include
software
eng
ineeri
ng,
m
obil
e
a
ppli
c
at
ion
,
c
y
b
e
r
sec
uri
t
y
.
At
pre
sent
she
is
working
as
an
As
sistant
Profess
or
in
IT
dep
artm
ent
at Al
Da
r U
nive
rsit
y
Co
ll
e
ge,
Dub
ai
,
UA
E.
Su
jni
Paul
is
a
fac
u
lty
in
the
Depa
rtment
of
Com
pute
r
Inform
at
ion
Scie
nc
e
in
the
Higher
Coll
ege
s
of
T
ec
hnolog
y
.
She
has
aro
und
1
6
y
ea
rs
of
Ac
ade
m
ic
teac
hin
g
expe
r
ie
n
ce.
She
complet
ed
h
er
Ph.D i
n
the
y
e
ar
2009.
Her
res
ea
rch
areas i
ncl
u
de
par
a
ll
e
l
and
d
istri
bute
d
d
at
a
m
ini
ng,
w
eb
ser
vic
es
and
techn
ologi
es,
b
ig
data
and
c
loud
co
m
puti
ng.
She
ha
s
m
ore
tha
n
45
rese
arc
h
pap
ers
publi
c
at
ions
in
r
efe
rre
d
journ
al
s
and
in
te
rna
ti
on
al
conf
er
ences.
At
pre
sent
she
is
supervising
3
Ph.D.
rese
arc
h
sc
hola
rs
and
one
PhD
student
has
gra
duated.
She
has
orga
nized
var
ious
works
hops
and
conf
ere
n
ce
s
and
has
giv
en
m
an
y
guest
l
ec
tur
es.
She
has
cont
ributed
in
cur
ric
u
lum
desi
gn
and
h
as
be
en
a
m
ember
of
the
boar
d
o
f
studie
s
in
va
rious
rep
ut
ed
orga
nizati
ons.
S
he
is
an
au
thor
o
f
a
Chapter
in
a
book
ti
tled
Dist
r
ibut
ed
Da
ta
Min
ing.
She
is
ve
r
y
kee
n
in
comm
u
nity
service
and
has
supervise
d
m
an
y
student
s
i
n
doing
service
rel
a
te
d
proj
ec
t
s
.
She
is a
n
ed
it
ori
al
bo
ard
m
ember
and
r
eviewer
of
diffe
ren
t
jou
rna
l
s.
Dr.
Khali
qu
z
z
a
man
Khan
rec
ei
ved
his
Ph.D
in
Stat
isti
c
s
from
Agra
Univer
sit
y
,
Agra,
India
i
n
2004.
He
re
ce
iv
ed
his
MA
in
Stat
ist
ic
s
and
BA
(Honors
)
in
Econom
ic
s
from
A
li
gar
h
Mus
li
m
Univer
sit
y
,
Al
ig
arh
,
Ind
ia
in
19
75
and
1978
r
espe
ctively
.
His
r
ese
arc
h
intere
sts
inc
lud
e
ar
ea
s
cove
ring
m
an
ag
ement
espe
ci
a
lly
ind
ustria
l
m
anagem
ent
,
and
qua
nti
tative
fina
n
ce
with
spec
i
fic
int
er
est
in
s
ec
uri
t
y
and
por
tfol
io
ana
l
y
ses.
At
pre
sent
he
is
worki
ng
as
an
As
sista
nt
Profess
or
in
Al
Dar
Univ
ersity
Co
llege, Duba
i
.
Evaluation Warning : The document was created with Spire.PDF for Python.