I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
p
ute
r
E
ng
in
ee
ring
(
I
J
E
C
E)
Vo
l.
7
,
No
.
1
,
Feb
r
u
ar
y
201
7
,
p
p
.
4
0
2
~
4
0
7
I
SS
N:
2088
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v
7
i
1
.
p
p
4
0
2
-
4
0
7
402
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
jo
u
r
n
a
l.c
o
m/o
n
lin
e/in
d
ex
.
p
h
p
/I
JE
C
E
Strea
m
ing
Big
Da
ta
Ana
ly
sis
f
o
r R
e
a
l
-
T
i
m
e
Sen
ti
m
en
t
b
a
sed
Ta
rg
eted
Adv
er
ti
sing
L
e
k
ha
R.
Na
ir
,
Su
j
a
la
D.
Sh
et
t
y
,
Sid
dh
a
nt
Dee
pa
k
Sh
et
t
y
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter S
c
ien
c
e
,
Birl
a
In
sti
tu
te
o
f
T
e
c
h
n
o
lo
g
y
a
n
d
S
c
ien
c
e
(BIT
S
)
P
il
a
n
i,
Du
b
a
i
C
a
m
p
u
s,
Un
it
e
d
A
ra
b
Em
irate
s
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
l 1
5
,
2
0
1
6
R
ev
i
s
ed
Dec
2
5
,
2
0
1
6
A
cc
ep
ted
J
an
8
,
2
0
1
7
Big
Da
ta
c
o
n
stit
u
ti
n
g
f
ro
m
th
e
in
f
o
r
m
a
ti
o
n
sh
a
re
d
in
th
e
v
a
ri
o
u
s
so
c
ial
n
e
tw
o
rk
site
s
h
a
v
e
g
re
a
t
r
e
lev
a
n
c
e
f
o
r
re
se
a
rc
h
to
b
e
a
p
p
li
e
d
in
d
iv
e
rse
f
ield
s
li
k
e
m
a
rk
e
ti
n
g
,
p
o
l
it
ics
,
h
e
a
lt
h
o
r
d
isa
ste
r
m
a
n
a
g
e
m
e
n
t.
S
o
c
ial
n
e
t
w
o
rk
sites
li
k
e
F
a
c
e
b
o
o
k
a
n
d
T
w
it
ter
a
re
n
o
w
e
x
t
e
n
siv
e
l
y
u
se
d
f
o
r
c
o
n
d
u
c
ti
n
g
b
u
sin
e
ss
,
m
a
rk
e
ti
n
g
p
ro
d
u
c
ts
a
n
d
se
rv
ice
s
a
n
d
c
o
ll
e
c
ti
n
g
o
p
i
n
io
n
s
a
n
d
f
e
e
d
b
a
c
k
s
re
g
a
rd
in
g
th
e
sa
m
e
.
S
in
c
e
d
a
ta
g
a
th
e
re
d
f
ro
m
th
e
se
sites
re
g
a
rd
in
g
a
p
ro
d
u
c
t/
b
ra
n
d
a
re
u
p
-
to
-
d
a
te
a
n
d
a
re
m
o
stl
y
su
p
p
li
e
d
v
o
l
u
n
tarily
,
it
ten
d
s
to
b
e
m
o
re
re
a
li
stic,
m
a
ss
iv
e
a
n
d
re
f
l
e
c
ts
th
e
g
e
n
e
ra
l
p
u
b
li
c
o
p
i
n
i
o
n
.
I
ts
a
n
a
ly
si
s
o
n
re
a
l
ti
m
e
c
a
n
lea
d
to
a
c
c
u
ra
te i
n
sig
h
t
s
a
n
d
re
sp
o
n
d
in
g
to
th
e
re
su
lt
s
so
o
n
e
r
is
u
n
d
o
u
b
ted
ly
a
d
v
a
n
tag
e
o
u
s
th
a
n
re
sp
o
n
d
i
n
g
late
r.
In
th
is
p
a
p
e
r,
a
c
lo
u
d
b
a
se
d
s
y
ste
m
f
o
r
re
a
l
ti
m
e
targ
e
ted
a
d
v
e
rti
sin
g
b
a
se
d
o
n
t
w
e
e
t
se
n
ti
m
e
n
t
a
n
a
ly
sis
is
d
e
sig
n
e
d
a
n
d
im
p
lem
e
n
ted
u
si
n
g
th
e
b
ig
d
a
ta
p
r
o
c
e
ss
in
g
e
n
g
in
e
A
p
a
c
h
e
S
p
a
rk
,
u
ti
li
z
i
n
g
it
s
str
e
a
m
in
g
li
b
ra
ry
.
A
p
p
li
c
a
ti
o
n
is
m
e
a
n
t
to
p
ro
m
o
te cro
ss
se
ll
in
g
a
n
d
p
r
o
v
id
e
b
e
tt
e
r
c
u
sto
m
e
r
su
p
p
o
rt.
K
ey
w
o
r
d
:
B
ig
d
ata
Sp
ar
k
Stre
a
m
i
n
g
b
ig
d
ata
p
r
o
ce
s
s
in
g
T
ar
g
eted
ad
v
er
tis
in
g
T
w
ee
t
s
e
n
ti
m
e
n
t a
n
al
y
s
is
Co
p
y
rig
h
t
©
2
0
1
7
In
stit
u
te o
f
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
rin
g
a
n
d
S
c
ien
c
e
.
Al
l
rig
h
ts
re
se
rv
e
d
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
L
e
k
h
a
R
.
Nair
,
Dep
ar
t
m
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
,
B
I
T
S P
ilan
i,
Du
b
ai
C
a
m
p
u
s
P
.
O.
B
o
x
: 3
4
5
0
5
5
,
Du
b
ai
I
n
ter
n
atio
n
al
A
ca
d
e
m
ic
C
i
t
y
,
D
u
b
ai,
Un
ited
A
r
ab
E
m
ir
ates
E
m
ail: le
k
h
ar
n
air
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
So
cial
n
et
w
o
r
k
s
ites
h
a
v
e
b
e
co
m
e
a
p
r
o
m
i
n
en
t
p
lat
f
o
r
m
t
o
ex
p
r
ess
o
p
in
io
n
s
an
d
f
ee
d
b
ac
k
s
.
W
it
h
w
id
esp
r
ea
d
u
s
e
o
f
s
m
ar
tp
h
o
n
es
an
d
ev
er
g
r
o
w
i
n
g
p
o
p
u
lar
it
y
o
f
s
o
cial
n
et
w
o
r
k
s
i
tes,
m
o
s
t
p
eo
p
le
n
o
w
s
h
ar
e
th
eir
s
en
ti
m
en
t
s
an
d
ex
p
er
ien
ce
ab
o
u
t
an
y
n
e
w
m
ar
k
et
p
r
o
d
u
ct
al
m
o
s
t
i
n
s
ta
n
tl
y
i
n
th
e
s
o
cial
n
et
w
o
r
k
s
an
d
th
ese
p
o
s
t
s
h
a
v
e
g
r
ea
t
in
f
l
u
e
n
ce
in
th
e
b
u
y
i
n
g
p
atter
n
s
o
f
p
r
o
s
p
ec
tiv
e
cu
s
to
m
er
s
.
A
m
o
d
el
f
o
r
k
n
o
w
led
g
e
tr
an
s
f
er
f
r
o
m
s
o
cial
n
et
w
o
r
k
s
to
p
r
ed
ict
h
u
m
a
n
b
eh
a
v
io
r
is
g
i
v
en
i
n
[
1
]
w
h
ich
ca
n
b
e
ap
p
lied
in
s
o
cial
m
ar
k
et
in
g
.
B
u
s
in
e
s
s
m
ar
k
et
l
ea
d
er
s
h
av
e
id
en
ti
f
ied
t
h
e
p
o
ten
tial
o
f
t
h
e
se
s
ite
s
to
g
at
h
er
o
p
in
io
n
s
ab
o
u
t
a
p
r
o
d
u
ct
r
ath
er
th
an
co
n
d
u
cti
n
g
a
m
ar
k
et
s
u
r
v
e
y
,
as
t
h
e
d
ata
f
r
o
m
f
o
r
m
er
r
ef
lects
r
ec
en
t
o
p
in
io
n
s
,
m
o
s
tl
y
u
n
b
ia
s
ed
f
ee
li
n
g
s
w
h
ic
h
w
i
ll
b
e
m
o
r
e
r
ea
lis
tic
a
n
d
co
m
es
i
n
h
u
g
e
v
o
lu
m
es
r
ep
r
ese
n
ti
n
g
a
f
air
p
er
ce
n
ta
g
e
o
f
g
en
er
al
p
u
b
lic,
t
h
o
u
g
h
th
e
d
a
ta
b
ein
g
lar
g
el
y
u
n
s
tr
u
ct
u
r
ed
.
I
n
b
r
a
n
d
co
m
p
eti
tio
n
,
i
m
m
e
d
iate
ac
tio
n
s
ta
k
en
b
ased
o
n
cu
s
to
m
er
f
ee
d
b
ac
k
s
r
esu
lt
i
n
s
tr
ateg
ic
ad
v
an
ta
g
e
o
f
o
n
e
b
r
an
d
o
v
er
an
o
th
er
.
Satis
f
ied
cu
s
to
m
er
s
o
f
a
p
r
o
d
u
ct
ar
e
m
o
r
e
li
k
el
y
to
b
u
y
a
n
a
s
s
o
ciate
d
p
r
o
d
u
ct
f
r
o
m
th
e
s
a
m
e
b
r
an
d
i
f
a
n
e
f
f
ec
ti
v
e
m
ar
k
eti
n
g
s
tr
ate
g
y
tar
g
eti
n
g
th
o
s
e
c
u
s
to
m
er
s
is
s
u
cc
e
s
s
f
u
ll
y
i
m
p
le
m
e
n
ted
.
W
h
ile
a
co
n
te
n
ted
c
u
s
to
m
er
ca
n
b
r
in
g
i
n
m
o
r
e
r
ev
en
u
e,
ex
ce
s
s
i
v
e
n
e
g
ati
v
e
s
e
n
ti
m
en
ts
r
eg
ar
d
i
n
g
a
p
r
o
d
u
ct,
s
p
r
ea
d
in
g
o
v
er
s
o
cial
m
ed
ia,
c
an
ad
v
er
s
el
y
a
f
f
ec
t
th
e
s
ale
s
an
d
r
esu
lt in
lo
s
in
g
l
o
y
al
a
n
d
p
r
o
s
p
ec
tiv
e
cu
s
to
m
er
s
.
1
.
1
.
T
he
P
ro
ble
m
:
Str
i
k
i
ng
Whil
e
t
he
I
ro
n
i
s
Hot
C
r
o
s
s
-
s
ell
in
g
,
w
h
er
e
an
ad
d
it
io
n
al
p
r
o
d
u
ct
o
r
s
er
v
ice
i
s
s
o
ld
to
an
ex
i
s
ti
n
g
c
u
s
to
m
er
,
is
d
etailed
in
[
2
]
w
h
ich
r
eq
u
ir
es
ad
v
er
tis
i
n
g
t
h
e
p
r
ec
is
e
p
r
o
d
u
ct
to
th
e
ex
ac
t
cu
s
to
m
er
at
t
h
e
co
r
r
ec
t
ti
m
e.
I
n
t
h
e
cu
r
r
e
n
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2088
-
8708
S
tr
ea
min
g
B
ig
Da
ta
A
n
a
lysi
s
fo
r
R
ea
l
-
Time
S
en
timen
t B
a
s
ed
Ta
r
g
eted
A
d
ve
r
tis
in
g
(
Lekh
a
R
.
N
a
ir
)
403
h
ig
h
l
y
co
m
p
etiti
v
e
m
ar
k
eti
n
g
s
ce
n
ar
io
,
cr
o
s
s
-
s
ell
in
g
ca
n
b
r
in
g
in
h
u
g
e
r
ev
e
n
u
e
an
d
t
h
e
s
tr
ateg
y
is
v
er
y
ef
f
ec
tiv
e
w
h
en
th
e
ex
i
s
ti
n
g
cu
s
to
m
er
h
a
s
p
o
s
iti
v
e
s
en
ti
m
en
t
to
w
ar
d
s
t
h
e
o
w
n
ed
p
r
o
d
u
ct
w
h
i
le
tar
g
e
ted
ad
v
er
tis
i
n
g
to
t
h
ese
c
u
s
t
o
m
er
s
ca
n
i
n
cr
ea
s
e
r
et
u
r
n
o
n
in
v
e
s
t
m
e
n
t.
A
t
th
e
s
a
m
e
ti
m
e,
f
o
r
ev
er
y
b
r
an
d
it
is
n
ec
es
s
ar
y
to
tac
k
le
th
e
is
s
u
e
s
r
aised
b
y
t
h
e
u
n
s
ati
s
f
ied
c
u
s
to
m
er
an
d
to
p
ac
if
y
h
i
m
at
t
h
e
ea
r
lies
t
s
o
as
to
r
eg
ain
h
is
b
r
an
d
co
n
f
id
en
ce
.
A
tta
in
i
n
g
s
o
cial
n
et
w
o
r
k
s
o
u
r
ce
d
r
ea
l
tim
e
b
ig
d
ata
f
o
r
an
aly
s
i
s
is
n
o
t
ea
s
y
as
m
o
s
t
s
ite
s
lack
p
u
b
lic
ap
p
licatio
n
p
r
o
g
r
am
m
i
n
g
in
t
er
f
ac
e(
A
P
I
)
f
o
r
a
th
ir
d
p
ar
t
y
to
ac
ce
s
s
,
w
i
th
T
w
itter
b
ein
g
a
n
ex
ce
p
tio
n
.
A
cc
o
r
d
in
g
to
in
ter
n
et
s
tatis
t
ic
s
m
o
r
e
th
an
6
0
0
0
t
w
ee
ts
ar
e
p
o
s
ted
p
er
s
ec
o
n
d
w
h
ic
h
is
h
u
g
e
e
n
o
u
g
h
i
n
ter
m
s
o
f
v
o
l
u
m
e
an
d
v
elo
cit
y
to
b
e
h
an
d
led
b
y
tr
ad
itio
n
al
d
ata
a
n
al
y
tic
s
y
s
te
m
an
d
h
e
n
ce
n
ec
e
s
s
it
ates t
h
e
u
s
ag
e
o
f
a
b
ig
d
ata
p
r
o
ce
s
s
in
g
s
y
s
te
m
.
I
n
th
is
w
o
r
k
,
a
n
ap
ac
h
e
s
p
ar
k
b
ased
b
ig
d
ata
ap
p
licatio
n
is
m
o
d
elled
a
n
d
i
m
p
le
m
e
n
ted
o
n
clo
u
d
th
a
t
p
r
o
ce
s
s
es
r
ea
l
ti
m
e
t
w
ee
ts
r
e
g
ar
d
in
g
a
p
r
o
d
u
ct
x
an
d
id
e
n
ti
f
y
it
s
s
e
n
ti
m
en
t
.
I
f
th
e
s
en
t
i
m
e
n
t
is
n
e
g
ati
v
e,
cu
s
to
m
er
s
u
p
p
o
r
t
is
o
f
f
er
ed
i
n
s
ta
n
tl
y
a
n
d
f
ee
d
b
ac
k
is
r
eq
u
ested
t
h
r
o
u
g
h
d
ir
ec
t
m
e
s
s
a
g
e,
else,
ad
v
er
tis
e
m
e
n
t
o
f
an
ass
o
ciate
d
p
r
o
d
u
ct
y
is
tar
g
eted
to
th
e
u
s
er
.
L
o
ca
tio
n
o
f
th
e
u
s
er
is
al
s
o
co
llected
to
p
r
o
v
id
e
lo
ca
tio
n
s
p
ec
if
ic
s
er
v
ices
an
d
to
id
en
t
if
y
g
eo
g
r
ap
h
ic
ar
ea
s
w
h
er
e
m
ar
k
et
in
g
o
r
cu
s
to
m
er
s
er
v
ic
e
s
ec
tio
n
n
ee
d
to
b
e
co
n
ce
n
tr
ated
.
Si
n
ce
th
e
s
e
p
r
o
s
p
ec
tiv
e
c
u
s
to
m
er
s
ar
e
tar
g
eted
at
t
h
e
r
i
g
h
t
ti
m
e
w
h
en
t
h
e
y
h
av
e
e
x
p
r
ess
ed
t
h
eir
s
en
ti
m
e
n
ts
,
it i
s
o
b
v
io
u
s
th
at
t
h
is
co
u
ld
b
e
a
b
etter
m
ar
k
eti
n
g
s
tr
ate
g
y
.
1
.
2
.
Select
ing
Ass
o
cia
t
ed/Rec
o
mm
ende
d P
ro
du
ct
I
n
m
ar
k
et
b
as
k
et
a
n
al
y
s
i
s
,
c
u
s
to
m
er
tr
an
s
ac
tio
n
s
ar
e
a
n
al
y
s
e
d
to
r
ec
o
g
n
ize
t
h
eir
p
u
r
c
h
asi
n
g
p
atter
n
.
Ass
o
c
iatio
n
r
u
le
lear
n
in
g
[
3
]
i
s
a
m
et
h
o
d
to
id
en
tify
r
ela
tio
n
s
a
m
o
n
g
v
ar
iab
les
i
n
a
d
ataset
w
h
ic
h
ca
n
b
e
u
s
ed
to
f
in
d
r
elate
d
p
r
o
d
u
cts
i
n
c
u
s
to
m
er
tr
an
s
ac
tio
n
s
lead
in
g
to
ef
f
ec
tiv
e
m
ar
k
eti
n
g
d
ec
i
s
i
o
n
s
.
B
y
as
s
o
ciatio
n
an
al
y
s
is
,
f
o
r
a
p
r
o
d
u
ct
x
,
a
n
ass
o
ciate
d
p
r
o
d
u
ct
y
ca
n
b
e
i
d
en
tifie
d
w
h
ich
is
b
o
u
g
h
t
to
g
eth
er
w
it
h
o
r
af
ter
b
u
y
i
n
g
p
r
o
d
u
ct
x
.
R
ec
o
m
m
e
n
d
atio
n
s
y
s
te
m
s
id
en
ti
f
y
p
r
o
d
u
cts
to
b
e
r
ec
o
m
m
e
n
d
ed
b
ased
o
n
c
u
s
to
m
er
’
s
p
a
s
t
p
u
r
ch
a
s
es
an
d
o
th
er
u
s
er
s
b
eh
av
io
r
.
A
p
leth
o
r
a
o
f
w
o
r
k
h
a
v
e
b
ee
n
ca
r
r
ied
o
u
t
in
as
s
o
ciatio
n
an
al
y
s
is
[
4
-
5
]
an
d
r
ec
o
m
m
e
n
d
er
s
y
s
te
m
s
[
6
-
7
]
a
n
d
it
is
n
o
t
in
cl
u
d
ed
in
t
h
e
s
co
p
e
o
f
th
is
p
ap
er
w
h
er
e
it
i
s
ass
u
m
ed
th
at
t
h
e
ass
o
ciate
d
p
r
o
d
u
ct
y
an
d
th
e
p
r
o
d
u
ct
to
b
e
r
ec
o
m
m
e
n
d
ed
z
,
h
ad
alr
ea
d
y
b
ee
n
id
en
ti
f
ied
.
1.
3
.
Rela
t
ed
Wo
rk
s
Ma
n
y
r
e
s
ea
r
ch
w
o
r
k
s
h
a
v
e
b
ee
n
ca
r
r
ied
o
u
t
in
s
e
n
ti
m
e
n
t
a
n
al
y
s
i
s
[
8
]
.
Fin
d
in
g
c
u
s
to
m
er
s
en
ti
m
en
ts
to
w
ar
d
s
a
b
r
an
d
b
y
m
i
n
i
n
g
s
o
cial
m
ed
ia
tex
t
w
a
s
th
e
to
p
ic
o
f
[
9
]
w
h
ile
u
s
a
g
e
o
f
t
w
itte
r
d
ata
f
o
r
s
en
ti
m
en
t
an
al
y
s
is
w
as
d
is
cu
s
s
ed
i
n
[
1
0
]
.
Sev
er
al
w
o
r
k
s
w
er
e
d
o
n
e
f
o
r
r
ev
ea
li
n
g
s
e
n
ti
m
e
n
ts
r
eg
ar
d
in
g
p
er
s
o
n
s
o
r
p
r
o
d
u
cts
th
at
m
ad
e
u
s
e
o
f
t
w
i
tter
d
ata
[
1
1
-
1
3
]
.
I
n
m
o
s
t
o
f
t
h
e
w
o
r
k
s
,
an
a
l
y
s
is
w
a
s
p
er
f
o
r
m
ed
o
n
s
tatic
d
ata.
Usef
u
l
n
es
s
o
f
s
o
cial
m
ed
ia
i
n
b
u
s
in
e
s
s
i
s
an
ac
tiv
e
r
e
s
ea
r
ch
ar
ea
an
d
m
ar
k
eti
n
g
s
co
p
e
o
f
s
o
cial
m
ed
ia
i
s
d
eta
iled
in
[
1
4
]
.
R
elatio
n
s
h
i
p
m
ar
k
eti
n
g
v
ia
t
w
it
ter
is
t
h
e
to
p
ic
o
f
d
is
cu
s
s
io
n
o
f
[
1
5
]
,
w
h
ile
m
ar
k
et
in
g
h
elp
f
u
l
n
es
s
o
f
t
w
itter
i
n
h
o
tel
in
d
u
s
tr
y
i
s
ex
p
lai
n
ed
in
[
1
6
]
.
T
h
is
w
o
r
k
i
m
p
le
m
e
n
ts
a
u
to
m
a
ted
r
ea
l
tim
e
tar
g
eted
ad
v
er
tis
i
n
g
s
y
s
te
m
b
ased
o
n
r
ea
l
ti
m
e
s
en
t
i
m
e
n
t
an
al
y
s
is
o
f
t
w
itter
d
ata.
Do
n
e
f
r
o
m
a
B
ig
Data
p
er
s
p
ec
tiv
e,
th
e
s
y
s
te
m
is
h
i
g
h
l
y
s
ca
lab
le
as
it
m
a
k
e
s
u
s
e
o
f
b
ig
d
ata
p
r
o
ce
s
s
in
g
e
n
g
in
e
Sp
ar
k
,
w
h
ic
h
tak
e
s
in
to
ac
co
u
n
t
o
f
ch
alle
n
g
es a
n
d
o
p
p
o
r
tu
n
itie
s
o
f
b
ig
d
ata
[
1
7
]
2.
RE
S
E
ARCH
M
E
T
H
O
D
2
.
1
.
Da
t
a
s
et
:
T
w
it
t
er
Str
ea
m
ing
Da
t
a
T
w
it
ter
,
th
e
p
r
ev
ale
n
t
m
icr
o
b
l
o
g
g
i
n
g
s
ite
w
it
h
3
2
0
m
illi
o
n
m
o
n
t
h
l
y
ac
ti
v
e
ac
co
u
n
t
s
as
p
e
r
co
m
p
a
n
y
s
tatis
t
ics,
allo
w
s
u
s
er
to
s
en
d
1
4
0
ch
ar
ac
ter
li
m
ited
m
es
s
a
g
e
s
ter
m
ed
t
w
ee
t
s
,
v
i
s
ib
le
to
all.
On
e
ca
n
also
s
en
d
a
d
ir
ec
t
m
es
s
ag
e
w
h
ic
h
is
v
i
s
i
b
le
o
n
l
y
to
th
e
in
te
n
d
ed
u
s
er
.
T
w
it
ter
’
s
g
lo
b
al
s
tr
ea
m
o
f
d
at
a
ca
n
b
e
ac
ce
s
s
ed
w
it
h
t
h
e
aid
o
f
T
w
itter
s
tr
ea
m
i
n
g
A
P
I
.
Fo
r
th
i
s
r
ea
l
ti
m
e
ac
ce
s
s
to
t
w
ee
t
s
,
a
p
er
s
is
te
n
t
HT
T
P
co
n
n
ec
tio
n
i
s
r
eq
u
ir
ed
to
b
e
o
p
en
.
A
n
ap
p
lic
atio
n
in
ten
d
ed
to
u
s
e
T
w
itter
A
P
I
n
ee
d
to
o
b
tain
O
Au
t
h
ac
c
ess
to
k
e
n
o
n
b
e
h
al
f
o
f
a
t
w
it
ter
ac
co
u
n
t.
Au
th
o
r
i
ze
d
r
eq
u
ests
to
t
h
e
T
w
i
tter
S
tr
ea
m
i
n
g
A
P
I
ca
n
b
e
is
s
u
ed
b
y
th
e
ap
p
licatio
n
m
ak
in
g
u
s
e
o
f
ac
ce
s
s
to
k
e
n
a
n
d
s
ec
r
et
k
e
y
s
.
O
n
ce
t
h
e
co
n
n
e
ctio
n
i
s
es
tab
lis
h
ed
,
Sp
ar
k
Str
ea
m
i
n
g
b
u
ilt
o
n
t
h
e
to
p
o
f
s
p
ar
k
co
r
e
tak
es c
ar
e
o
f
th
e
r
ec
ep
tio
n
o
f
r
ea
l ti
m
e
t
w
e
ets
w
h
ic
h
t
h
en
p
r
o
ce
s
s
ed
b
y
s
p
ar
k
co
r
e
en
g
in
e.
2
.
2
.
T
o
o
ls
:
Apa
che
Sp
a
rk
a
n
d Spa
r
k
Str
ea
m
ing
L
ibra
ry
Sin
ce
tr
ad
itio
n
al
d
ata
p
r
o
ce
s
s
in
g
s
y
s
te
m
s
h
a
v
e
s
ca
lab
ili
t
y
is
s
u
es
a
n
d
ar
e
n
o
t
eq
u
ip
p
ed
to
h
an
d
l
e
s
tr
ea
m
i
n
g
d
ata
o
f
i
m
m
en
s
e
v
o
lu
m
e,
a
s
ca
lab
le
b
ig
d
ata
p
r
o
ce
s
s
in
g
s
y
s
te
m
is
p
r
ef
er
r
ed
f
o
r
th
is
ap
p
licatio
n
.
Sp
ar
k
[
1
8
]
is
an
o
p
en
s
o
u
r
ce
co
m
p
u
ti
n
g
en
g
i
n
e
m
ea
n
t
f
o
r
d
is
tr
ib
u
ted
d
ata
p
r
o
ce
s
s
i
n
g
.
Ha
d
o
o
p
[
1
9
]
,
th
e
f
ir
s
t
g
en
er
atio
n
b
ig
d
ata
p
r
o
ce
s
s
in
g
en
g
in
e
is
s
lo
w
l
y
b
ein
g
r
ep
lace
d
b
y
Sp
ar
k
w
h
ic
h
is
co
n
s
i
d
er
ed
as
th
e
s
ec
o
n
d
g
en
er
atio
n
B
ig
Data
p
r
o
ce
s
s
i
n
g
en
g
i
n
e
b
y
[
2
0
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
1
,
Feb
r
u
ar
y
201
7
:
4
0
2
–
407
404
Dr
iv
er
p
r
o
g
r
a
m
o
f
s
p
ar
k
ap
p
licatio
n
r
u
n
s
t
h
e
m
ai
n
f
u
n
cti
o
n
an
d
p
er
f
o
r
m
s
p
ar
allel
o
p
e
r
atio
n
s
o
n
v
ar
io
u
s
w
o
r
k
er
n
o
d
es
in
a
s
p
a
r
k
clu
s
ter
.
S
p
ar
k
u
s
e
s
th
e
co
n
ce
p
t
o
f
R
esil
ie
n
t
Di
s
tr
ib
u
ted
Data
s
et
(
R
DD)
[
2
1
]
,
w
h
ic
h
is
a
co
llectio
n
o
f
i
m
m
u
tab
le
o
b
j
ec
ts
s
eg
r
eg
ated
ac
r
o
s
s
th
e
clu
s
ter
n
o
d
es
f
o
r
p
er
f
o
r
m
in
g
p
ar
allel
o
p
er
atio
n
s
.
R
DD
s
ca
n
b
e
p
er
s
is
ted
in
m
e
m
o
r
y
f
o
r
r
ep
etiti
v
e
u
s
e
an
d
d
u
e
to
t
h
is
in
-
m
e
m
o
r
y
an
al
y
tic
s
,
s
p
ar
k
p
er
f
o
r
m
s
f
as
ter
th
a
n
th
e
Ha
d
o
o
p
,
esp
ec
ially
in
iter
ati
v
e
ap
p
licatio
n
s
.
T
h
o
u
g
h
Sp
ar
k
is
m
ai
n
l
y
a
b
atch
p
r
o
ce
s
s
in
g
en
g
i
n
e,
Sp
ar
k
ec
o
s
y
s
te
m
is
eq
u
ip
p
ed
w
it
h
Sp
ar
k
Stre
a
m
i
n
g
t
h
at
i
s
d
esti
n
ed
f
o
r
s
tr
ea
m
in
g
d
ata
p
r
o
ce
s
s
in
g
a
s
g
i
v
en
i
n
Fi
g
u
r
e
1
.
I
n
s
p
ar
k
s
tr
ea
m
i
n
g
,
co
n
ti
n
u
o
u
s
s
tr
ea
m
o
f
d
ata
is
r
ep
r
esen
ted
b
y
d
is
cr
etize
d
s
tr
ea
m
(
Dst
r
ea
m
)
w
h
ic
h
is
a
s
eq
u
en
ce
o
f
R
D
Ds.
I
n
th
is
w
o
r
k
,
s
p
ar
k
s
tr
ea
m
in
g
r
ec
eiv
e
s
an
d
h
an
d
le
s
th
e
r
ea
l
ti
m
e
t
w
ee
t
s
f
r
o
m
th
e
T
w
i
tter
S
tr
ea
m
i
n
g
A
P
I
af
ter
estab
li
s
h
i
n
g
th
e
co
n
n
e
ctio
n
.
Fig
u
r
e
1
.
Sp
ar
k
w
i
th
S
tr
ea
m
i
n
g
:
A
r
ch
itec
tu
r
al
d
iag
r
a
m
2
.
3
.
T
he
Sy
s
t
e
m
M
o
del
T
h
e
w
o
r
k
f
lo
w
m
o
d
el
o
f
t
h
e
s
y
s
te
m
is
g
iv
e
n
i
n
Fi
g
u
r
e
2
,
w
h
ic
h
is
b
u
ilt
ar
o
u
n
d
Sp
ar
k
.
On
ce
t
h
e
co
n
n
ec
tio
n
w
i
th
T
w
itter
s
tr
ea
m
i
n
g
A
P
I
is
estab
li
s
h
ed
,
f
r
o
m
a
m
o
n
g
th
o
u
s
a
n
d
s
o
f
t
w
ee
t
s
p
o
s
ted
p
er
s
ec
o
n
d
,
th
e
ap
p
licatio
n
f
ilter
t
w
ee
ts
r
e
g
ar
d
in
g
a
p
ar
ticu
lar
p
r
o
d
u
ct
x
.
Sp
ar
k
Stre
a
m
i
n
g
h
a
n
d
les
t
h
is
s
tr
ea
m
i
n
g
d
ata
an
d
p
ac
k
th
ese
t
w
ee
t
s
in
to
b
atch
e
s
an
d
h
an
d
o
v
er
to
u
n
d
er
l
y
i
n
g
s
p
ar
k
co
r
e
en
g
in
e
f
o
r
p
r
o
ce
s
s
i
n
g
.
Se
n
ti
m
e
n
t
o
f
ea
ch
t
w
ee
t i
s
an
a
l
y
ze
d
i
n
r
ea
l
ti
m
e
a
n
d
if
f
o
u
n
d
p
o
s
iti
v
e/
n
eu
tr
al,
ad
v
er
tis
e
m
en
t o
f
an
a
s
s
o
c
iate
d
p
r
o
d
u
ct
y
o
r
a
r
ec
o
m
m
e
n
d
ed
p
r
o
d
u
ct
z
is
tar
g
eted
to
th
e
t
w
ee
ter
,
w
h
ile
s
te
p
s
ar
e
tak
en
to
o
f
f
er
cu
s
to
m
er
s
u
p
p
o
r
t
an
d
g
ath
er
r
elev
an
t
i
n
f
o
r
m
atio
n
r
eg
ar
d
in
g
d
is
s
ati
s
f
ac
tio
n
i
n
ca
s
e
o
f
n
e
g
ati
v
e
t
w
ee
ts
,
s
o
th
at
r
e
m
ed
ial
m
ea
s
u
r
es
ca
n
b
e
tak
en
i
m
m
ed
iatel
y
to
p
r
ev
en
t l
o
s
in
g
p
r
o
s
p
ec
tiv
e
cu
s
to
m
er
s
Fig
u
r
e
2
.
W
o
r
k
Flo
w
M
o
d
el
o
f
th
e
A
p
p
licatio
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2088
-
8708
S
tr
ea
min
g
B
ig
Da
ta
A
n
a
lysi
s
fo
r
R
ea
l
-
Time
S
en
timen
t B
a
s
ed
Ta
r
g
eted
A
d
ve
r
tis
in
g
(
Lekh
a
R
.
N
a
ir
)
405
2
.
4
.
P
r
o
du
ct
Senti
m
ent
Ana
l
y
s
is
Ma
n
y
r
esear
c
h
w
o
r
k
s
h
av
e
b
ee
n
ca
r
r
ied
o
u
t
in
s
en
t
i
m
e
n
t
an
al
y
s
is
.
S
tan
f
o
r
d
o
f
f
er
s
a
n
o
p
en
s
o
u
r
ce
s
en
ti
m
e
n
t
an
al
y
ze
r
lib
r
ar
y
th
a
t
ca
n
b
e
u
s
ed
ef
f
ec
ti
v
el
y
to
ca
r
r
y
o
u
t
s
e
n
ti
m
e
n
t
an
al
y
s
is
.
T
h
er
e
is
n
o
f
o
o
l
p
r
o
o
f
alg
o
r
ith
m
f
o
r
s
en
ti
m
e
n
t
an
al
y
s
i
s
as
m
an
y
N
L
P
alg
o
r
ith
m
s
s
tu
m
b
le
o
n
ac
c
u
r
atel
y
i
d
en
tify
i
n
g
s
ar
ca
s
ti
c
co
m
m
e
n
t
s
.
Fo
r
th
i
s
p
r
o
to
ty
p
e,
w
e
h
a
v
e
u
s
ed
th
e
m
o
s
t
p
r
i
m
it
iv
e
t
y
p
e
o
f
s
e
n
ti
m
e
n
t
a
n
al
y
ti
c
m
et
h
o
d
o
f
f
i
n
d
i
n
g
th
e
r
elati
v
e
co
u
n
t
o
f
p
o
s
iti
v
e
an
d
n
e
g
ati
v
e
s
e
n
ti
m
en
t
h
o
ld
in
g
w
o
r
d
s
i
n
th
e
t
w
ee
t.
A
lar
g
e
co
llectio
n
o
f
o
v
er
5
0
0
0
w
o
r
d
s
li
k
e
g
o
o
d
,
am
az
i
n
g
etc.
co
m
m
o
n
l
y
u
s
ed
to
ex
p
r
ess
p
o
s
itiv
e
s
e
n
ti
m
e
n
ts
ar
e
co
m
p
iled
in
a
tex
t
f
ile
to
b
e
u
s
ed
as
a
lo
o
k
u
p
tab
le.
Sa
m
e
is
d
o
n
e
f
o
r
n
eg
a
tiv
e
w
o
r
d
s
as
w
ell.
A
co
u
n
ter
is
in
itial
ized
to
ze
r
o
,
ass
u
m
in
g
n
e
u
tr
al
s
en
t
i
m
e
n
t,
a
n
d
f
o
r
ea
ch
w
o
r
d
i
n
th
e
t
w
ee
t,
co
m
p
ar
is
o
n
is
d
o
n
e
w
i
th
a
s
e
t
o
f
p
o
s
i
tiv
e
w
o
r
d
s
an
d
n
eg
at
iv
e
w
o
r
d
s
.
I
f
th
e
w
o
r
d
is
ass
o
ciate
d
w
it
h
p
o
s
itiv
e
s
e
n
ti
m
e
n
t,
co
u
n
ter
is
in
cr
e
m
e
n
ted
o
r
if
n
eg
ati
v
e,
th
e
co
u
n
ter
is
d
ec
r
e
m
en
ted
an
d
th
e
s
i
g
n
o
f
th
e
f
in
a
l
co
u
n
ter
v
alu
e
d
eter
m
in
e
s
w
h
e
t
h
er
th
e
p
r
o
d
u
ct
is
ass
o
ciate
d
w
it
h
p
o
s
i
tiv
e,
n
e
g
ativ
e
o
r
n
e
u
tr
al
s
e
n
ti
m
e
n
t.
T
h
o
u
g
h
t
h
e
m
et
h
o
d
is
s
i
m
p
le,
i
t
is
h
a
v
i
n
g
o
b
v
io
u
s
d
r
a
w
b
ac
k
s
a
n
d
ca
n
b
e
r
ep
lace
d
b
y
an
y
co
n
v
e
n
ien
t
s
en
ti
m
en
t
an
al
y
t
ic
alg
o
r
it
h
m
.
2
.
5
.
L
o
ca
t
i
o
n Specif
ic
Serv
ices
Fo
r
lo
ca
tio
n
en
ab
led
t
w
ee
ts
,
u
s
er
lo
ca
tio
n
is
id
en
ti
f
ied
f
r
o
m
t
h
e
t
w
ee
t,
an
d
lo
ca
tio
n
s
p
e
cif
ic
o
f
f
er
s
an
d
s
er
v
ices
ar
e
tar
g
eted
to
t
h
ese
u
s
er
s
.
B
y
m
ap
p
in
g
t
w
ee
t
lo
ca
tio
n
s
b
ased
o
n
s
en
ti
m
e
n
t
s
,
g
eo
g
r
ap
h
ic
ar
ea
s
w
h
er
e
atten
tio
n
is
r
eq
u
ir
ed
ca
n
b
e
id
en
ti
f
ied
an
d
ap
p
r
o
p
r
iate
ac
tio
n
s
ca
n
b
e
ta
k
en
.
2.
6
.
Alg
o
ri
t
h
m
Select
a
p
r
o
d
u
ct
x
a.
Fin
d
as
s
o
ciate
d
p
r
o
d
u
ct
y
u
s
in
g
ass
o
ciatio
n
an
al
y
s
is
b.
Fin
d
a
p
r
o
d
u
ct
z
th
at
ca
n
b
e
r
ec
o
m
m
e
n
d
ed
to
u
s
er
u
s
in
g
r
ec
o
m
m
en
d
atio
n
s
y
s
te
m
.
c.
W
h
ile
(
t
w
itter
A
P
I
co
n
n
ec
tio
n
is
tr
u
e)
a.
Fil
ter
t
w
ee
t stre
a
m
r
eg
ar
d
in
g
t
h
e
p
r
o
d
u
ct
b.
Fo
r
ea
ch
t
w
ee
t (
t
w
ee
t(
i)
)
1.
Get
u
s
er
n
a
m
e(
u
s
er
(
i)
an
d
lo
ca
tio
n
(
lo
c(
i)
)
2.
Fin
d
s
e
n
ti
m
e
n
t o
f
t
h
e
t
w
ee
t se
n
ti(i)
I
f
(
s
en
ti(i)
==
p
o
s
itiv
e
O
R
n
e
u
tr
al)
A
d
v
er
tis
e
as
s
o
ciate
d
p
r
o
d
u
ct
y
an
d
z
to
th
e
u
s
er
(
i)
I
f
(
lo
c(
i)
is
n
o
t n
u
ll)
ad
v
er
tis
e
lo
ca
tio
n
lo
c(
i)
s
p
ec
if
ic
o
f
f
er
s
to
t
h
e
u
s
er
(
i)
else
Of
f
er
cu
s
to
m
er
s
u
p
p
o
r
t to
u
s
er
(
i)
an
d
r
eq
u
est f
o
r
u
s
er
(
i)
f
ee
d
b
ac
k
3
.
Sav
e
t
w
ee
t(
i)
,
lo
c(
i)
an
d
s
e
n
ti(i)
f
o
r
f
u
r
th
er
a
n
al
y
s
i
s
.
3.
RE
SU
L
T
S
A
ND
AN
AL
Y
SI
S
T
h
o
u
g
h
m
an
y
w
o
r
k
s
r
eg
ar
d
i
n
g
s
e
n
ti
m
e
n
t
an
al
y
s
is
o
f
t
w
i
tter
d
ata
w
er
e
d
o
n
e
b
ef
o
r
e,
th
is
w
o
r
k
u
tili
ze
s
r
ea
l
ti
m
e
t
w
ee
t
s
e
n
ti
m
en
t
an
al
y
s
i
s
f
o
r
r
ea
l
ti
m
e
t
ar
g
eted
ad
v
er
tis
i
n
g
m
ak
in
g
u
s
e
o
f
s
ca
lab
le
o
p
en
s
o
u
r
ce
s
p
ar
k
s
tr
ea
m
in
g
,
w
h
ic
h
w
as
n
o
t
atte
m
p
ted
b
ef
o
r
e.
T
h
e
ap
p
licatio
n
w
as
b
u
ilt
u
s
i
n
g
Si
m
p
le
B
u
ild
T
o
o
l
(
SB
T
)
an
d
r
u
n
o
n
a
Sp
ar
k
clu
s
ter
w
i
th
a
m
aster
an
d
t
w
o
s
la
v
e
n
o
d
es
co
n
f
i
g
u
r
ed
o
n
i5
p
r
o
ce
s
s
o
r
,
4
GB
R
A
M
an
d
Ub
u
n
tu
1
4
.
0
4
o
p
e
r
atin
g
s
y
s
te
m
.
I
t
w
a
s
also
s
u
cc
e
s
s
f
u
ll
y
d
ep
lo
y
ed
o
n
Am
az
o
n
E
la
s
ti
c
C
o
m
p
u
te
C
lo
u
d
(
E
C
2
)
.
Sp
ar
k
C
lu
s
ter
w
ith
t2
.
m
icr
o
co
n
f
ig
u
r
atio
n
w
a
s
cr
ea
ted
an
d
af
ter
test
i
n
g
th
e
ap
p
lica
tio
n
,
th
e
cl
u
s
ter
w
a
s
d
estro
y
ed
.
Sp
ar
k
ec
2
s
cr
ip
t
w
a
s
u
til
ized
in
la
u
n
c
h
i
n
g
a
n
d
m
a
n
ag
i
n
g
s
p
ar
k
cl
u
s
ter
i
n
E
C
2
cl
o
u
d
.
T
ab
le
1
.
R
ec
eiv
ed
T
w
ee
ts
a
n
d
R
ea
l T
im
e
R
esp
o
n
s
e
b
ased
o
n
S
e
n
ti
m
en
t a
n
d
L
o
ca
tio
n
R
e
a
l
T
i
me
Tw
e
e
t
s R
e
c
e
i
v
e
d
S
e
n
t
i
me
n
t
i
d
e
n
t
i
f
i
e
d
,
L
o
c
a
t
i
o
n
D
i
r
e
c
t
M
e
ssag
e
S
e
n
t
(
T
a
r
g
e
t
e
d
A
d
v
e
r
t
i
se
me
n
t
)
my
x
p
a
d
1
0
w
o
r
k
s fi
n
e
P
o
si
t
i
v
e
,
n
u
l
l
L
i
mi
t
e
d
p
e
r
i
o
d
o
f
f
e
r
,
1
0
%
d
i
s
c
o
u
n
t
o
n
a
l
l
O
r
a
n
g
e
mo
b
i
l
e
a
c
c
e
sso
r
i
e
s
P
r
i
c
e
o
f
X
P
a
d
-
1
0
i
s g
o
o
d
b
u
t
p
i
c
t
u
r
e
q
u
a
l
i
t
y
i
s
p
o
o
r
N
e
u
t
r
a
l
,
D
u
b
a
i
L
i
mi
t
e
d
p
e
r
i
o
d
o
f
f
e
r
,
1
0
%
d
i
s
c
o
u
n
t
o
n
a
l
l
O
r
a
n
g
e
mo
b
i
l
e
a
c
c
e
sso
r
i
e
s
A
maz
i
n
g
o
f
f
e
r
:
C
l
e
a
r
a
n
c
e
S
a
l
e
a
t
O
r
a
n
g
e
i
-
st
o
r
e
s a
t
D
e
i
r
a
C
i
t
y
C
e
n
t
e
r
,
D
u
b
a
i
N
e
w
x
p
a
d
1
0
s
u
c
k
s,
d
o
n
t
buy
N
e
g
a
t
i
v
e
,
I
n
d
i
a
P
l
e
a
se
c
a
l
l
t
o
l
l
f
r
e
e
n
o
8
0
0
-
1
2
3
4
f
o
r
a
l
l
y
o
u
r
c
o
mp
l
a
i
n
t
s
o
r
v
i
si
t
w
w
.
o
r
a
n
g
e
.
c
o
m/
c
u
st
c
a
r
e
t
o
se
r
v
e
y
o
u
b
e
t
t
e
r
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
1
,
Feb
r
u
ar
y
201
7
:
4
0
2
–
407
406
T
h
e
ap
p
licatio
n
w
a
s
in
i
tiall
y
test
ed
b
y
f
ilter
i
n
g
t
w
ee
ts
r
eg
a
r
d
in
g
p
o
p
u
lar
p
r
o
d
u
cts
av
ailab
le
in
th
e
m
ar
k
et
an
d
its
s
e
n
ti
m
e
n
t
w
er
e
an
al
y
ze
d
,
lo
ca
tio
n
id
en
ti
f
ie
d
an
d
s
a
v
ed
i
n
a
f
i
le.
T
ar
g
eted
ad
v
er
tis
i
n
g
w
a
s
d
is
ab
led
in
th
is
ca
s
e.
T
h
ir
t
y
t
o
s
ix
t
y
t
w
ee
t
s
p
er
m
i
n
u
te
w
er
e
o
b
s
er
v
ed
r
eg
ar
d
in
g
alr
ea
d
y
estab
lis
h
ed
m
ar
k
et
p
r
o
d
u
cts,
b
u
t th
e
n
u
m
b
er
i
s
ex
p
ec
ted
to
s
h
o
o
t u
p
in
th
e
i
n
itia
l p
er
io
d
s
w
h
e
n
a
n
e
w
p
r
o
d
u
ct
is
lau
n
c
h
ed
.
.
T
h
e
ap
p
licatio
n
w
a
s
test
ed
b
y
s
en
d
i
n
g
p
o
s
itiv
e
an
d
n
e
g
at
iv
e
s
en
ti
m
en
t
t
w
ee
t
s
f
r
o
m
5
d
if
f
er
en
t
t
w
i
tter
ac
co
u
n
t
s
ab
o
u
t
a
h
y
p
o
th
etica
l
p
r
o
d
u
ct
x
P
ad
-
1
0
f
r
o
m
co
m
p
a
n
y
Or
a
n
g
e.
A
ll
th
e
t
w
ee
ts
w
er
e
r
ec
eiv
ed
i
n
r
ea
l
ti
m
e
a
n
d
its
s
e
n
ti
m
e
n
ts
w
er
e
id
en
t
if
ied
a
n
d
ac
co
r
d
in
g
l
y
p
r
o
m
o
tio
n
a
l
o
f
f
er
m
e
s
s
ag
e
s
o
r
f
ee
d
b
ac
k
r
eq
u
est/
cu
s
to
m
er
s
u
p
p
o
r
t
d
eta
ils
w
er
e
s
e
n
t
b
y
t
h
e
ap
p
licati
o
n
as
d
ir
ec
t
m
es
s
ag
e
to
ea
c
h
t
w
ee
ter
a
s
g
iv
e
n
i
n
T
ab
le
1
.
A
ls
o
th
e
t
w
ee
t d
etail
s
w
er
e
r
ec
o
r
d
ed
in
a
f
ile
f
o
r
d
etailed
an
al
y
s
is
later
.
4.
DIS
CU
SS
I
O
N
S
I
n
th
is
p
ap
er
a
s
ca
lab
le
s
p
ar
k
ap
p
licatio
n
to
p
er
f
o
r
m
r
ea
l
t
i
m
e
tar
g
eted
ad
v
er
ti
s
in
g
to
p
r
o
s
p
ec
tiv
e
cu
s
to
m
er
s
b
ased
o
n
th
e
s
en
t
i
m
e
n
ts
e
x
p
r
ess
ed
o
n
r
elate
d
p
r
o
d
u
cts
o
n
t
w
i
tter
is
i
m
p
l
e
m
en
ted
.
Sin
ce
n
o
s
en
ti
m
e
n
t a
n
al
y
s
i
s
al
g
o
r
ith
m
g
iv
es a
f
o
o
l p
r
o
o
f
r
esu
lt,
t
h
e
o
b
s
er
v
ed
s
e
n
ti
m
e
n
t
m
a
y
b
e
d
if
f
e
r
en
t i
n
s
o
m
e
ca
s
e
s
,
b
u
t
s
i
n
ce
t
h
e
ap
p
licatio
n
is
a
b
o
u
t
r
ea
l
ti
m
e
tar
g
eted
ad
v
er
t
is
in
g
,
it
w
i
ll
n
o
t
h
a
v
e
an
y
n
e
g
ati
v
e
ef
f
ec
t
o
n
th
e
p
er
f
o
r
m
a
n
ce
.
T
w
it
ter
u
s
er
s
w
h
o
ar
e
v
er
y
m
u
c
h
co
n
ce
r
n
ed
ab
o
u
t
t
h
eir
p
r
iv
ac
y
m
i
g
h
t
d
is
ab
le
lo
ca
tio
n
tr
ac
k
in
g
,
w
h
er
e
lo
ca
tio
n
s
p
ec
i
f
ic
s
er
v
i
ce
s
b
ec
o
m
e
s
in
s
i
g
n
i
f
ica
n
t.
Als
o
,
if
th
e
u
s
er
d
is
ab
les
t
h
e
o
p
tio
n
o
f
r
ec
eiv
in
g
d
ir
ec
t m
e
s
s
a
g
es
f
r
o
m
e
v
er
y
o
n
e,
it
w
ill b
e
h
ar
d
to
tar
g
et
t
h
at
u
s
er
f
o
r
ad
v
er
ti
s
in
g
.
5.
CO
NCLU
SI
O
N
T
h
e
B
ig
Data
a
n
al
y
tic
s
y
s
te
m
m
ea
n
t
f
o
r
r
ea
l
ti
m
e
tar
g
eted
ad
v
er
tis
i
n
g
w
h
er
e
tar
g
et
id
en
t
if
icatio
n
i
s
d
o
n
e
o
n
th
e
b
as
is
o
f
cu
s
to
m
er
s
en
ti
m
e
n
ts
s
h
ar
ed
o
n
t
w
it
ter
,
w
as
s
u
cc
ess
f
u
ll
y
b
u
i
lt
ar
o
u
n
d
t
h
e
b
ig
d
ata
p
r
o
ce
s
s
in
g
s
y
s
te
m
A
p
ac
h
e
Sp
ar
k
an
d
test
ed
o
n
Am
az
o
n
E
C
2
clo
u
d
.
T
h
e
s
am
e
ap
p
licatio
n
w
it
h
s
lig
h
t
m
o
d
i
f
icatio
n
ca
n
b
e
u
s
ed
in
in
ter
n
a
tio
n
al
p
o
liti
cs
f
o
r
d
ir
ec
t
ca
m
p
aig
n
i
n
g
a
n
d
to
tak
e
co
r
r
ec
tiv
e
m
ea
s
u
r
e
s
b
ased
o
n
p
u
b
lic
o
p
in
io
n
s
as
w
ell
as
to
f
o
r
m
u
late
w
i
n
n
in
g
s
tr
ateg
y
b
ased
o
n
p
r
ed
ictio
n
s
in
elec
tio
n
s
.
I
n
th
is
w
o
r
k
in
a
d
d
itio
n
to
r
ea
l
ti
m
e
a
n
al
y
s
i
s
,
t
h
e
i
n
d
iv
id
u
al
t
w
ee
t
w
it
h
it
s
lo
ca
tio
n
an
d
p
r
ed
icted
s
en
ti
m
e
n
t
is
s
to
r
ed
in
a
csv
f
i
le
w
h
ic
h
ca
n
b
e
m
i
n
ed
to
g
ain
in
s
i
g
h
ts
to
w
ar
d
s
a
lo
n
g
ter
m
p
o
lic
y
f
o
r
m
u
l
atio
n
.
RE
F
E
R
E
NC
E
S
[1
]
E.
Zh
o
n
g
,
W
.
F
a
n
,
J.W
.
L
.
X
ia
o
a
n
d
Y.
L
i,
"
Co
mS
o
c
:
Ad
a
p
ti
v
e
T
r
a
n
sfe
r
o
f
Us
e
r
Beh
a
v
i
o
rs
o
v
e
r
Co
mp
o
site
S
o
c
i
a
l
Ne
two
rk
"
,
in
1
8
th
A
CM
S
IG
KD
D i
n
tern
a
ti
o
n
a
l
c
o
n
f
e
re
n
c
e
o
n
Kn
o
w
led
g
e
d
isc
o
v
e
r
y
a
n
d
d
a
ta m
in
in
g
,
2
0
1
2
.
[2
]
S
.
L
i,
B.
S
u
n
a
n
d
L
.
M
.
A
la
n
,
"
Cro
ss
-
se
ll
in
g
th
e
rig
h
t
p
ro
d
u
c
t
to
th
e
rig
h
t
c
u
sto
m
e
r
a
t
th
e
rig
h
t
ti
m
e
"
,
J
o
u
rn
a
l
o
f
M
a
rk
e
ti
n
g
Res
e
a
rc
h
,
v
o
l
.
4
8
,
n
o
.
4
,
p
p
.
6
8
3
-
7
0
0
,
2
0
1
1
.
[3
]
R.
A
g
r
a
w
a
l,
T
.
I
m
ieliń
sk
i
a
n
d
A.
S
w
a
m
i,
"
M
in
in
g
a
ss
o
c
iatio
n
ru
l
e
s
b
e
tw
e
e
n
se
ts
o
f
it
e
m
s
in
larg
e
d
a
tab
a
se
s"
,
in
ACM
S
IGM
OD
in
ter
n
a
ti
o
n
a
l
c
o
n
f
e
re
n
c
e
o
n
M
a
n
a
g
e
me
n
t
o
f
d
a
t
a
,
1
9
9
3
.
[4
]
C.
C.
A
g
g
a
r
w
a
l,
C.
P
ro
c
o
p
i
u
c
a
n
d
P
.
S
.
Yu
,
"
F
in
d
in
g
l
o
c
a
li
z
e
d
a
ss
o
c
iatio
n
s
in
m
a
rk
e
t
b
a
sk
e
t
d
a
ta"
,
IEE
E
T
ra
n
sa
c
ti
o
n
s
o
n
Kn
o
wled
g
e
a
n
d
Da
ta
E
n
g
in
e
e
rin
g
,
v
o
l.
1
4
,
n
o
.
1
,
p
p
.
5
1
-
6
2
,
2
0
0
2
.
[5
]
M.
Ku
b
a
t,
A
.
Ha
f
e
z
,
V
.
V
.
Ra
g
h
a
v
a
n
,
J.R.
L
e
k
k
a
la
a
n
d
W
.
K
.
Ch
e
n
,
"
Item
se
t
tre
e
s
f
o
r
targ
e
ted
a
ss
o
c
iatio
n
q
u
e
ry
in
g
"
,
IEE
E
T
ra
n
sa
c
t
io
n
s o
n
Kn
o
wled
g
e
a
n
d
D
a
ta
En
g
i
n
e
e
rin
g
,
v
o
l.
1
5
,
n
o
.
6
,
p
p
.
1
5
2
2
-
1
5
3
4
,
2
0
0
3
.
[6
]
H.K.
Kim
,
J.K.
Ki
m
a
n
d
Y.U.
R
y
u
,
"
P
e
rso
n
a
li
z
e
d
Re
c
o
m
m
e
n
d
a
ti
o
n
o
v
e
r
a
Cu
sto
m
e
r
Ne
t
w
o
rk
f
o
r
Ub
iq
u
it
o
u
s
S
h
o
p
p
i
n
g
"
,
IEE
E
T
r
a
n
s
a
c
ti
o
n
s o
n
S
e
rv
ice
s Co
mp
u
ti
n
g
,
v
o
l.
2
,
n
o
.
2
,
p
p
.
1
4
0
-
1
5
1
,
2
0
0
9
.
[7
]
K.A
.
A
l
m
o
h
se
n
a
n
d
A
.
J.
Hu
d
a
,
"
Re
c
o
m
m
e
n
d
e
r
S
y
ste
m
s
in
L
ig
h
t
o
f
Big
Da
ta
"
,
In
ter
n
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
El
e
c
trica
l
a
n
d
Co
m
p
u
ter
E
n
g
i
n
e
e
rin
g
(
IJ
ECE
)
,
v
o
l.
5
,
n
o
.
6
,
2
0
1
5
.
[8
]
B.
L
iu
,
"
S
e
n
ti
m
e
n
t
a
n
a
l
y
sis
a
n
d
o
p
in
i
o
n
m
in
in
g
"
,
S
y
n
th
e
sis
lec
tu
re
s
o
n
h
u
m
a
n
l
a
n
g
u
a
g
e
tec
h
n
o
lo
g
ie
s
,
v
o
l.
5
,
n
o
.
1
,
p
p
.
1
-
1
6
7
,
2
0
1
2
.
[9
]
M.
M
.
M
o
sta
f
a
,
"
M
o
re
th
a
n
w
o
rd
s:
S
o
c
ial
n
e
tw
o
rk
s’
te
x
t
m
in
in
g
fo
r
c
o
n
su
m
e
r
b
ra
n
d
se
n
ti
m
e
n
ts"
,
Exp
e
rt
S
y
ste
ms
wit
h
A
p
p
l
ica
ti
o
n
s
,
v
o
l.
4
0
,
n
o
.
1
0
,
p
p
.
4
2
4
1
-
4
2
5
1
,
2
0
1
3
.
[1
0
]
A
.
P
a
k
a
n
d
P
.
P
a
tri
c
k
,
"
Tw
it
ter
a
s
a
Co
rp
u
s
f
o
r
S
e
n
ti
m
e
n
t
A
n
a
l
y
sis
a
n
d
Op
i
n
io
n
M
in
i
n
g
"
,
L
RE
C
,
v
o
l.
1
0
,
p
p
.
1
3
2
0
-
1
3
2
6
,
2
0
1
0
.
[1
1
]
S.
L
iu
e
t
a
l.
,
"
TA
S
C:T
o
p
ic
-
A
d
a
p
ti
v
e
S
e
n
ti
m
e
n
t
C
las
si
f
ic
a
ti
o
n
o
n
Dy
n
a
m
ic
Twe
e
ts
"
,
IEE
E
T
ra
n
sa
c
ti
o
n
s
o
n
Kn
o
wled
g
e
a
n
d
D
a
ta
En
g
i
n
e
e
rin
g
,
v
o
l.
2
7
,
n
o
.
6
,
p
p
.
1
6
9
6
-
1
7
0
9
,
2
0
1
5
.
[1
2
]
P
.
R.
Ca
v
a
li
n
e
t
a
l.
,
"
A
sc
a
lab
le
a
r
c
h
it
e
c
tu
re
f
o
r
re
a
l
-
ti
m
e
a
n
a
l
y
sis
o
f
m
icro
b
lo
g
g
in
g
d
a
ta"
,
IBM
J
o
u
r
n
a
l
o
f
Res
e
a
rc
h
a
n
d
De
v
e
lo
p
me
n
t
,
v
o
l.
5
9
,
n
o
.
2
/3
,
p
p
.
1
6
-
1
,
2
0
1
5
.
[1
3
]
X
.
Ch
e
n
y
a
n
,
Y.
Ya
n
g
a
n
d
H.
Ch
u
n
-
Ke
u
n
g
,
"
Hid
d
e
n
i
n
-
g
a
m
e
in
telli
g
e
n
c
e
in
NBA
p
la
y
e
rs
'
t
w
e
e
ts
"
,
Co
mm
u
n
ica
ti
o
n
s o
f
th
e
ACM
,
v
o
l
.
5
8
,
n
o
.
1
1
,
p
p
.
8
0
-
8
9
,
2
0
1
5
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2088
-
8708
S
tr
ea
min
g
B
ig
Da
ta
A
n
a
lysi
s
fo
r
R
ea
l
-
Time
S
en
timen
t B
a
s
ed
Ta
r
g
eted
A
d
ve
r
tis
in
g
(
Lekh
a
R
.
N
a
ir
)
407
[1
4
]
M
.
S
.
Ya
d
a
v
e
t
a
l.
,
"
S
o
c
ial
c
o
m
m
e
rc
e
:
a
c
o
n
ti
n
g
e
n
c
y
f
ra
m
e
w
o
rk
f
o
r
a
ss
e
ss
in
g
m
a
rk
e
ti
n
g
p
o
ten
ti
a
l"
,
El
se
v
ier
J
o
u
rn
a
l
o
f
In
ter
a
c
ti
v
e
M
a
rk
e
ti
n
g
,
v
o
l.
2
7
,
n
o
.
4
,
p
p
.
3
1
1
-
3
2
3
,
2
0
1
3
.
[1
5
]
B.
A
.
W
a
t
k
in
s
a
n
d
R
.
L
e
w
is
,
"
T
w
it
ter
a
s
G
a
tewa
y
t
o
Rel
a
ti
o
n
sh
ip
M
a
rk
e
ti
n
g
:
A
C
o
n
ten
t
A
n
a
lys
is
o
f
Re
l
a
ti
o
n
sh
i
p
Bu
il
d
in
g
v
ia
T
wit
ter
"
,
in
S
o
c
ial
M
e
d
ia an
d
S
trate
g
ie Co
m
m
u
n
ica
ti
o
n
s,
UK
,
P
a
lg
ra
v
e
M
a
c
m
il
lan
,
2
0
1
3
,
p
p
.
2
5
-
4
4
.
[1
6
]
X
.
Y.
L
e
u
n
g
,
B.
Bil
ly
a
n
d
A
.
S
.
Ku
rt,
"
T
h
e
m
a
r
k
e
ti
n
g
e
ff
e
c
ti
v
e
n
e
ss
o
f
so
c
ial
m
e
d
ia
in
th
e
h
o
tel
in
d
u
stry
a
c
o
m
p
a
riso
n
o
f
f
a
c
e
b
o
o
k
a
n
d
tw
it
ter
"
,
J
o
u
rn
a
l
o
f
H
o
sp
it
a
li
ty
&
T
o
u
rism
Res
e
a
rc
h
,
v
o
l.
3
9
,
n
o
.
2
,
p
p
.
1
4
7
-
1
6
9
,
2
0
1
5
.
[1
7
]
H.
Ba
g
h
e
ri
a
n
d
A
.
A
b
d
u
sa
la
m
,
"
Big
Da
ta:
c
h
a
ll
e
n
g
e
s,
o
p
p
o
rt
u
n
it
ies
a
n
d
Clo
u
d
b
a
se
d
so
lu
ti
o
n
s"
,
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
El
e
c
trica
l
a
n
d
C
o
mp
u
t
e
r E
n
g
i
n
e
e
rin
g
(
IJ
ECE
)
,
v
o
l.
5
,
n
o
.
2
,
p
.
3
4
0
,
2
0
1
5
.
[1
8
]
[
On
li
n
e
]
.
A
v
a
il
a
b
le:
h
tt
p
s://
sp
a
rk
.
a
p
a
c
h
e
.
o
rg
/d
o
c
s/late
st/.
[
A
c
c
e
ss
e
d
1
5
F
e
b
ru
a
ry
2
0
1
6
].
[1
9
]
T
.
W
h
it
e
,
"
Ha
d
o
o
p
:
T
h
e
De
f
in
it
iv
e
G
u
id
e
,
3
rd
Ed
it
io
n
"
,
O'
Re
il
l
y
M
e
d
ia,
Ca
li
f
o
rn
ia,
2
0
1
2
.
[2
0
]
F
.
G
e
b
a
ra
,
H.
Ho
f
st
e
e
a
n
d
K.
No
wk
a
,
"
S
e
c
o
n
d
-
G
e
n
e
ra
ti
o
n
Big
Da
ta
S
y
ste
m
s
"
,
IEE
E
Co
mp
u
ter
,
v
o
l.
4
8
,
n
o
.
1
,
p
p
.
36
-
4
1
,
2
0
1
5
.
[2
1
]
M
.
Zah
a
ria,
M
.
Ch
o
w
d
h
u
ry
,
M
.
J.
F
ra
n
k
li
n
,
S
.
S
h
e
n
k
e
r
a
n
d
I.
S
t
o
ica
,
"
S
p
a
rk
:
Clu
ste
r
Co
m
p
u
ti
n
g
w
it
h
W
o
rk
in
g
S
e
ts"
,
in
US
ENIX
c
o
n
fer
e
n
c
e
o
n
Ho
t
to
p
ics
in
c
lo
u
d
c
o
mp
u
ti
n
g
,
2
0
1
0
.
Evaluation Warning : The document was created with Spire.PDF for Python.