I
nte
rna
t
io
na
l J
o
urna
l o
f
Rec
o
nfig
ura
ble a
nd
E
m
be
dd
e
d Sy
s
t
e
m
s
(
I
J
R
E
S)
Vo
l.
14
,
No
.
2
,
J
u
ly
20
25
,
p
p
.
472
~
4
8
0
I
SS
N:
2089
-
4864
,
DOI
:
1
0
.
1
1
5
9
1
/i
j
r
es
.
v
14
.
i
2
.
pp
4
7
2
-
480
472
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
r
es.ia
esco
r
e.
co
m
O
pti
m
i
zing
so
cia
l
m
edia
ana
ly
tics
w
ith
the
da
ta qua
lity
enha
nce
m
en
t
a
nd
ana
ly
tics
fra
m
ew
o
rk
for supe
rio
r
da
ta qua
lity
B
.
K
a
r
t
hick
1
,2
,
T
.
M
ey
y
a
pp
an
1
1
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
A
l
a
g
a
p
p
a
U
n
i
v
e
r
si
t
y
,
K
a
r
a
i
k
u
d
i
,
I
n
d
i
a
2
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
S
y
e
d
H
a
me
e
d
h
a
A
r
t
s a
n
d
S
c
i
e
n
c
e
C
o
l
l
e
g
e
,
K
i
l
a
k
a
r
a
i
,
I
n
d
i
a
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Au
g
1
8
,
2
0
2
4
R
ev
i
s
ed
A
p
r
8
,
2
0
2
5
A
cc
ep
ted
J
u
n
1
0
,
2
0
2
5
T
h
is
p
a
p
e
r
in
tr
o
d
u
c
e
s
th
e
d
a
ta
q
u
a
li
ty
e
n
h
a
n
c
e
m
e
n
t
a
n
d
a
n
a
ly
ti
c
s
(DQ
EA
)
f
ra
m
e
w
o
rk
to
e
n
h
a
n
c
e
d
a
ta
q
u
a
li
ty
in
so
c
ial
m
e
d
ia
a
n
a
l
y
ti
c
s
th
ro
u
g
h
m
a
c
h
in
e
lea
rn
in
g
(M
L
)
a
lg
o
rit
h
m
s.
T
h
e
e
ff
ica
c
y
o
f
th
e
f
r
a
m
e
w
o
rk
is
v
a
li
d
a
ted
th
ro
u
g
h
f
e
a
tu
re
s
tes
ted
a
g
a
in
st
h
u
m
a
n
c
o
d
e
rs
o
n
Am
a
z
o
n
M
e
c
h
a
n
ica
l
T
u
rk
,
a
c
h
iev
in
g
a
n
in
ter
-
c
o
d
e
r
re
li
a
b
il
it
y
sc
o
re
o
f
0
.
8
5
,
in
d
ica
ti
n
g
h
ig
h
a
g
re
e
m
e
n
t.
F
u
rt
h
e
r
m
o
re
,
tw
o
c
a
se
stu
d
ies
w
it
h
a
larg
e
so
c
ial
m
e
d
ia
d
a
tas
e
t
f
ro
m
T
u
m
b
lr
w
e
r
e
c
o
n
d
u
c
ted
to
d
e
m
o
n
stra
te
th
e
e
ffe
c
ti
v
e
n
e
ss
o
f
th
e
p
ro
p
o
se
d
c
o
n
t
e
n
t
f
e
a
tu
re
s.
In
th
e
f
irst
c
a
se
stu
d
y
,
th
e
DQ
EA
f
ra
m
e
w
o
rk
re
d
u
c
e
d
d
a
ta
n
o
ise
b
y
3
0
%
a
n
d
b
ias
b
y
2
5
%
,
w
h
il
e
in
c
re
a
sin
g
c
o
m
p
lete
n
e
ss
b
y
2
0
%
.
In
th
e
se
c
o
n
d
c
a
se
stu
d
y
,
th
e
f
ra
m
e
w
o
rk
im
p
ro
v
e
d
d
a
ta
c
o
n
siste
n
c
y
b
y
3
5
%
a
n
d
o
v
e
ra
ll
d
a
ta
q
u
a
li
ty
sc
o
re
b
y
2
8
%
.
Co
m
p
a
ra
ti
v
e
a
n
a
l
y
sis
w
it
h
sta
te
-
of
-
th
e
-
a
rt
m
o
d
e
ls,
in
c
lu
d
in
g
ra
n
d
o
m
f
o
re
st
a
n
d
su
p
p
o
r
t
v
e
c
to
r
m
a
c
h
in
e
s
(S
V
M
),
s
h
o
w
e
d
sig
n
if
ica
n
t
i
m
p
ro
v
e
m
e
n
ts
in
d
a
ta
re
li
a
b
il
it
y
a
n
d
d
e
c
isio
n
-
m
a
k
in
g
a
c
c
u
ra
c
y
.
S
p
e
c
i
f
ica
ll
y
,
t
h
e
DQ
EA
f
ra
m
e
w
o
rk
o
u
tp
e
rf
o
rm
e
d
th
e
ra
n
d
o
m
f
o
re
st
m
o
d
e
l
b
y
1
5
%
in
a
c
c
u
ra
c
y
a
n
d
2
0
%
i
n
tr
u
e
p
o
siti
v
e
ra
te,
a
n
d
th
e
S
V
M
m
o
d
e
l
b
y
1
0
%
i
n
e
rr
o
r
ra
te red
u
c
ti
o
n
a
n
d
1
8
%
i
n
re
li
a
b
il
it
y
.
T
h
e
re
su
lt
s
u
n
d
e
rsc
o
re
th
e
p
o
ten
ti
a
l
o
f
a
d
v
a
n
c
e
d
d
a
ta
a
n
a
ly
ti
c
s
to
o
ls
i
n
tran
sf
o
rm
in
g
so
c
ial
m
e
d
ia
d
a
ta
in
to
a
v
a
lu
a
b
le
a
ss
e
t
f
o
r
or
g
a
n
iza
ti
o
n
s,
h
ig
h
l
ig
h
ti
n
g
th
e
p
ra
c
ti
c
a
l
im
p
li
c
a
ti
o
n
s
a
n
d
f
u
tu
r
e
re
se
a
rc
h
d
irec
ti
o
n
s
in
th
is
d
o
m
a
in
.
K
ey
w
o
r
d
s
:
B
ig
d
ata
D
ata
q
u
alit
y
R
an
d
o
m
f
o
r
est
S
o
cial
m
ed
ia
an
al
y
s
is
Su
p
p
o
r
t v
ec
to
r
m
ac
h
i
n
e
T
u
m
b
lr
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
B
.
Kar
h
tick
Dep
ar
t
m
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
,
A
lag
ap
p
a
Un
i
v
er
s
i
t
y
Kar
aik
u
d
i
-
6
3
0
0
0
3
,
T
am
il Na
d
u
,
I
n
d
ia
E
m
ail: b
k
ar
t
h
ic
k
1
9
8
0
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
p
r
o
lif
er
atio
n
o
f
s
o
cial
m
e
d
ia
p
latf
o
r
m
s
in
r
ec
en
t
y
ea
r
s
h
as
tr
an
s
f
o
r
m
ed
th
e
w
a
y
i
n
d
iv
id
u
al
s
an
d
o
r
g
an
izatio
n
s
co
m
m
u
n
icate
,
s
h
ar
e
i
n
f
o
r
m
atio
n
,
a
n
d
e
n
g
a
g
e
w
it
h
th
eir
a
u
d
ie
n
ce
s
.
P
lat
f
o
r
m
s
s
u
ch
as
Face
b
o
o
k
,
T
w
it
ter
,
I
n
s
ta
g
r
a
m
,
a
n
d
T
u
m
b
lr
h
av
e
b
ec
o
m
e
i
n
te
g
r
al
p
ar
ts
o
f
d
ail
y
li
f
e,
g
e
n
er
ati
n
g
v
a
s
t
a
m
o
u
n
t
s
o
f
u
s
er
-
g
en
er
ated
co
n
te
n
t.
T
h
is
co
n
ten
t p
r
o
v
id
es a
r
ich
s
o
u
r
ce
o
f
d
at
a
th
at
ca
n
b
e
an
al
y
ze
d
to
g
ai
n
in
s
i
g
h
ts
in
to
p
u
b
lic
o
p
in
io
n
,
co
n
s
u
m
er
b
eh
av
io
r
,
m
ar
k
et
tr
en
d
s
,
an
d
m
o
r
e.
Ho
w
e
v
er
,
d
esp
ite
th
e
i
m
m
en
s
e
p
o
ten
tial
o
f
s
o
cial
m
ed
ia
d
ata,
th
e
q
u
alit
y
o
f
t
h
is
d
ata
is
o
f
te
n
co
m
p
r
o
m
is
ed
b
y
v
ar
io
u
s
f
ac
to
r
s
s
u
c
h
as
n
o
is
e,
b
ias,
an
d
in
co
m
p
lete
n
es
s
,
p
o
s
in
g
s
ig
n
i
f
i
ca
n
t
ch
alle
n
g
es
to
r
esear
ch
er
s
an
d
an
al
y
s
ts
[
1
]
–
[
6
]
.
No
is
e
i
n
s
o
cial
m
ed
ia
d
ata
r
ef
er
s
to
ir
r
elev
an
t
o
r
ex
tr
an
eo
u
s
in
f
o
r
m
a
tio
n
t
h
at
d
o
es
n
o
t
co
n
tr
ib
u
te
to
m
ea
n
i
n
g
f
u
l
an
al
y
s
is
.
T
h
is
ca
n
in
cl
u
d
e
s
p
a
m
,
o
f
f
-
to
p
ic
p
o
s
ts
,
an
d
d
u
p
licate
co
n
ten
t,
w
h
i
ch
ca
n
d
is
to
r
t
an
al
y
tical
o
u
t
co
m
e
s
an
d
lead
to
er
r
o
n
eo
u
s
co
n
cl
u
s
io
n
s
.
B
ia
s
in
s
o
cial
m
ed
ia
d
ata
ar
is
e
s
f
r
o
m
t
h
e
i
n
h
er
e
n
t
s
u
b
j
ec
tiv
it
y
a
n
d
v
ar
y
in
g
p
er
s
p
ec
tiv
es
o
f
u
s
er
s
,
a
s
w
el
l
as
t
h
e
al
g
o
r
ith
m
s
t
h
at
c
u
r
ate
co
n
te
n
t
[
7
]
–
[
1
0
]
.
T
h
is
ca
n
r
esu
lt
in
s
k
e
w
ed
d
atasets
th
a
t
d
o
n
o
t
ac
cu
r
atel
y
r
e
p
r
esen
t
th
e
b
r
o
ad
er
p
o
p
u
latio
n
o
r
p
h
en
o
m
en
a
b
ei
n
g
s
tu
d
ie
d
.
I
n
co
m
p
lete
n
es
s
,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4864
Op
timiz
in
g
s
o
cia
l m
ed
ia
a
n
a
ly
tics
w
ith
th
e
d
a
ta
q
u
a
lity e
n
h
a
n
ce
men
t a
n
d
a
n
a
lytics
…
(
B.
K
a
r
th
ick
)
473
an
o
th
er
cr
itical
is
s
u
e,
o
cc
u
r
s
w
h
e
n
d
atasets
lack
s
u
f
f
icie
n
t
d
ata
p
o
in
ts
o
r
h
av
e
m
is
s
i
n
g
i
n
f
o
r
m
atio
n
,
lead
in
g
to
g
ap
s
in
a
n
al
y
s
is
a
n
d
u
n
r
elia
b
le
r
esu
lts
.
A
d
d
r
ess
i
n
g
t
h
ese
d
ata
q
u
ali
t
y
is
s
u
es
i
s
cr
u
ci
al
f
o
r
en
s
u
r
i
n
g
th
e
r
eliab
ilit
y
a
n
d
v
alid
it
y
o
f
i
n
s
i
g
h
t
s
d
er
iv
ed
f
r
o
m
s
o
cial
m
ed
ia
an
al
y
tics
[
1
1
]
–
[
1
4
]
.
T
r
ad
itio
n
al
ap
p
r
o
ac
h
es
to
en
h
a
n
ci
n
g
d
ata
q
u
alit
y
,
s
u
c
h
as
b
u
s
i
n
ess
d
ec
is
io
n
m
a
n
ag
e
m
en
t
s
y
s
te
m
s
(
B
DM
S),
h
av
e
b
ee
n
em
p
lo
y
ed
to
m
iti
g
ate
th
e
s
e
ch
alle
n
g
es.
Ho
w
e
v
er
,
th
ese
m
eth
o
d
s
o
f
te
n
f
all
s
h
o
r
t
d
u
e
to
th
eir
r
elian
ce
o
n
p
r
ed
ef
in
ed
r
u
les
an
d
m
a
n
u
al
i
n
ter
v
e
n
tio
n
s
,
w
h
i
ch
m
a
y
n
o
t
s
ca
le
e
f
f
ec
ti
v
el
y
w
it
h
t
h
e
d
y
n
a
m
ic
a
n
d
v
o
lu
m
i
n
o
u
s
n
at
u
r
e
o
f
s
o
cial
m
ed
ia
d
ata
[1
5
]
–
[
1
8
]
.
T
h
er
e
i
s
a
p
r
ess
in
g
n
ee
d
f
o
r
in
n
o
v
ati
v
e
f
r
a
m
e
w
o
r
k
s
t
h
at
ca
n
s
y
s
te
m
atica
ll
y
i
m
p
r
o
v
e
d
ata
q
u
alit
y
w
h
ile
le
v
er
ag
i
n
g
th
e
ca
p
ab
ilit
ies
o
f
m
o
d
er
n
d
ata
an
al
y
tic
s
to
o
ls
.
I
n
r
esp
o
n
s
e
to
th
is
n
ee
d
,
th
is
p
ap
er
in
tr
o
d
u
ce
s
th
e
d
ata
q
u
alit
y
en
h
an
ce
m
e
n
t
an
d
an
al
y
tics
(
DQE
A
)
f
r
a
m
e
w
o
r
k
,
a
n
o
v
el
ap
p
r
o
ac
h
d
esig
n
ed
to
en
h
an
ce
t
h
e
q
u
alit
y
o
f
s
o
c
ial
m
ed
ia
d
ata
th
r
o
u
g
h
ad
v
a
n
ce
d
d
ata
an
aly
t
ics
tech
n
iq
u
es.
Un
lik
e
tr
ad
itio
n
a
l
m
et
h
o
d
s
,
th
e
DQE
A
f
r
a
m
e
w
o
r
k
u
ti
lizes
a
co
m
b
i
n
atio
n
o
f
a
u
to
m
ated
d
ata
p
r
o
ce
s
s
i
n
g
,
in
te
g
r
atio
n
,
a
n
d
tr
an
s
f
o
r
m
atio
n
tech
n
iq
u
es
to
ad
d
r
ess
n
o
is
e,
b
ias,
an
d
i
n
co
m
p
lete
n
e
s
s
m
o
r
e
e
f
f
ec
ti
v
e
l
y
[
1
9
]
–
[
2
4
]
.
T
h
e
f
r
a
m
e
w
o
r
k
is
i
m
p
le
m
e
n
ted
u
s
in
g
s
tate
-
of
-
t
h
e
-
ar
t
d
ata
an
al
y
tic
s
to
o
ls
s
u
c
h
as
s
tr
u
ct
u
r
e
d
q
u
er
y
lan
g
u
ag
e
(
SQL
)
,
T
ab
leau
,
an
d
A
p
ac
h
e
Sp
ar
k
,
w
h
ich
o
f
f
er
r
o
b
u
s
t
ca
p
ab
ilit
ies
f
o
r
d
ata
m
a
n
ip
u
latio
n
,
v
is
u
al
izatio
n
,
an
d
lar
g
e
-
s
ca
le
p
r
o
ce
s
s
i
n
g
.
T
h
e
D
QE
A
f
r
a
m
e
w
o
r
k
i
n
co
r
p
o
r
ates
s
ev
er
al
k
e
y
co
m
p
o
n
e
n
ts
ai
m
e
d
at
i
m
p
r
o
v
in
g
d
ata
q
u
alit
y
.
First,
it
e
m
p
lo
y
s
s
o
p
h
is
t
icate
d
d
ata
clea
n
in
g
tec
h
n
iq
u
e
s
to
f
ilter
o
u
t
n
o
is
e
a
n
d
ir
r
elev
an
t
co
n
te
n
t,
en
s
u
r
in
g
t
h
at
th
e
r
e
m
ai
n
i
n
g
d
ata
is
p
er
ti
n
en
t
a
n
d
m
ea
n
i
n
g
f
u
l.
T
h
ese
tec
h
n
iq
u
e
s
i
n
cl
u
d
e
th
e
u
s
e
o
f
p
atter
n
r
ec
o
g
n
itio
n
,
k
e
y
w
o
r
d
f
ilter
i
n
g
,
an
d
s
tati
s
tical
m
et
h
o
d
s
to
id
en
ti
f
y
a
n
d
r
e
m
o
v
e
u
n
w
a
n
ted
i
n
f
o
r
m
atio
n
.
Seco
n
d
,
th
e
f
r
a
m
e
w
o
r
k
ad
d
r
ess
es
b
ias
b
y
in
teg
r
ati
n
g
d
ata
f
r
o
m
m
u
ltip
le
s
o
u
r
ce
s
an
d
ap
p
l
y
in
g
n
o
r
m
aliza
tio
n
tech
n
iq
u
es
to
m
iti
g
ate
th
e
e
f
f
ec
ts
o
f
s
u
b
j
ec
tiv
e
p
er
s
p
ec
tiv
e
s
an
d
alg
o
r
ith
m
ic
c
u
r
atio
n
.
T
h
is
h
elp
s
to
cr
ea
te
a
m
o
r
e
b
ala
n
ce
d
an
d
r
ep
r
esen
tativ
e
d
ataset.
T
h
ir
d
,
th
e
f
r
a
m
e
w
o
r
k
tac
k
les
i
n
co
m
p
le
ten
e
s
s
b
y
e
m
p
lo
y
i
n
g
d
ata
in
te
g
r
atio
n
a
n
d
tr
an
s
f
o
r
m
atio
n
m
eth
o
d
s
t
h
at
f
il
l
g
ap
s
in
t
h
e
d
ata
an
d
e
n
s
u
r
e
co
n
s
is
te
n
c
y
ac
r
o
s
s
d
i
f
f
er
e
n
t
d
atasets
.
T
h
e
co
n
tr
ib
u
tio
n
s
o
f
th
e
p
r
o
p
o
s
ed
w
o
r
k
ar
e
g
iv
e
n
a
s
f
o
llo
w
s
:
−
T
h
e
in
tr
o
d
u
ctio
n
o
f
t
h
e
DQE
A
f
r
a
m
e
w
o
r
k
r
ep
r
esen
ts
a
s
i
g
n
i
f
ica
n
t
ad
v
a
n
ce
m
e
n
t
in
th
e
f
ield
o
f
s
o
cia
l
m
ed
ia
d
ata
q
u
alit
y
en
h
a
n
ce
m
en
t.
I
t
o
f
f
er
s
a
n
o
v
el
ap
p
r
o
ac
h
th
at
lev
er
ag
es
m
o
d
er
n
d
ata
an
al
y
tics
to
o
ls
to
ad
d
r
ess
cr
itical
d
ata
q
u
alit
y
is
s
u
es.
−
B
y
i
n
co
r
p
o
r
atin
g
au
to
m
ated
d
ata
clea
n
in
g
,
in
te
g
r
atio
n
,
an
d
tr
an
s
f
o
r
m
at
io
n
tech
n
i
q
u
es,
th
e
DQE
A
f
r
a
m
e
w
o
r
k
ef
f
ec
ti
v
el
y
r
ed
u
ce
s
n
o
is
e,
m
i
tig
a
tes b
ias,
an
d
f
ill
s
d
ata
g
ap
s
,
en
s
u
r
in
g
h
ig
h
er
d
ata
q
u
alit
y
.
−
T
h
e
f
r
a
m
e
w
o
r
k
’
s
f
ea
tu
r
e
s
ar
e
r
ig
o
r
o
u
s
l
y
v
alid
ated
ag
ai
n
s
t
h
u
m
a
n
co
d
er
s
o
n
Am
az
o
n
M
ec
h
an
ica
l
T
u
r
k
,
ac
h
iev
in
g
a
h
i
g
h
i
n
ter
-
co
d
er
r
eliab
ilit
y
s
co
r
e
o
f
0
.
8
5
,
w
h
ic
h
u
n
d
er
s
co
r
es t
h
e
ac
c
u
r
ac
y
a
n
d
r
eliab
ilit
y
o
f
t
h
e
f
r
a
m
e
w
o
r
k
.
−
T
h
r
o
u
g
h
t
w
o
ca
s
e
s
t
u
d
ies
w
it
h
T
u
m
b
lr
d
ata,
th
e
DQE
A
f
r
a
m
e
w
o
r
k
d
e
m
o
n
s
tr
ate
s
p
r
ac
tical
i
m
p
r
o
v
e
m
e
n
t
s
in
d
ata
q
u
a
lit
y
m
etr
ics,
i
n
cl
u
d
in
g
a
3
0
%
r
ed
u
ctio
n
i
n
n
o
is
e,
a
2
5
%
r
ed
u
ctio
n
in
b
ia
s
,
an
d
a
2
0
%
in
cr
ea
s
e
in
co
m
p
lete
n
es
s
.
2.
L
I
T
E
R
AT
U
RE
R
E
VI
E
W
T
h
e
liter
atu
r
e
o
n
d
ata
q
u
ality
e
n
h
an
ce
m
e
n
t
i
n
s
o
cial
m
e
d
ia
an
al
y
tic
s
u
n
d
er
s
co
r
es
th
e
p
er
v
asiv
e
ch
alle
n
g
e
s
o
f
n
o
is
e,
b
ias,
an
d
in
co
m
p
lete
n
es
s
in
h
er
en
t
in
s
o
cial
m
ed
ia
d
ata,
alo
n
g
w
i
th
th
e
ev
o
lv
in
g
m
et
h
o
d
s
an
d
li
m
ita
tio
n
s
i
n
ad
d
r
ess
in
g
th
ese
is
s
u
e
s
.
T
r
a
d
itio
n
al
ap
p
r
o
ac
h
es
lik
e
B
DM
S
h
a
v
e
b
ee
n
f
o
u
n
d
atio
n
a
l
b
u
t
o
f
ten
s
tr
u
g
g
le
w
it
h
th
e
d
y
n
a
m
ic
an
d
u
n
s
tr
u
ctu
r
ed
n
at
u
r
e
o
f
s
o
cial
m
ed
ia
co
n
ten
t.
B
er
ar
d
i
et
a
l
.
[
2
]
ex
p
lo
r
e
d
h
as
h
ta
g
s
e
g
m
en
ta
tio
n
a
n
d
te
x
t
q
u
alit
y
r
an
k
i
n
g
to
i
m
p
r
o
v
e
d
ata
r
elev
an
ce
a
n
d
ac
cu
r
ac
y
,
h
ig
h
li
g
h
t
i
n
g
in
itial
ef
f
o
r
t
s
to
s
tr
u
ct
u
r
e
an
d
f
ilter
s
o
cial
m
ed
ia
d
ata
ef
f
ec
ti
v
el
y
.
Sin
g
h
an
d
Ver
m
a
[
1
1
]
p
r
o
p
o
s
ed
an
ef
f
ec
ti
v
e
p
ar
allel
p
r
o
ce
s
s
in
g
f
r
a
m
e
w
o
r
k
f
o
r
s
o
cial
m
ed
ia
an
al
y
tic
s
,
ai
m
i
n
g
to
en
h
an
ce
s
ca
lab
ilit
y
an
d
p
r
o
ce
s
s
in
g
s
p
ee
d
b
u
t
f
ac
ed
ch
alle
n
g
es
i
n
m
ai
n
t
ai
n
in
g
d
ata
in
te
g
r
it
y
ac
r
o
s
s
d
is
tr
ib
u
ted
en
v
ir
o
n
m
en
ts
.
Mu
s
ta
f
a
et
a
l
.
[
1
3
]
e
m
p
lo
y
ed
m
ac
h
i
n
e
lear
n
i
n
g
(
ML
)
to
p
r
ed
ict
cr
ick
et
m
atc
h
o
u
tco
m
e
s
b
ased
o
n
s
o
cial
n
et
w
o
r
k
o
p
in
io
n
s
,
d
em
o
n
s
tr
ati
n
g
t
h
e
p
o
ten
tial
o
f
p
r
ed
ictiv
e
a
n
al
y
tics
b
u
t
n
o
ti
n
g
th
e
v
ar
iab
ilit
y
i
n
d
ata
q
u
a
lit
y
a
n
d
s
e
n
ti
m
e
n
t
an
al
y
s
is
ac
cu
r
ac
y
.
Si
n
g
h
e
t
a
l
.
[
1
0
]
in
v
esti
g
ated
T
w
itt
er
an
al
y
tics
f
o
r
p
r
ed
ictin
g
elec
tio
n
o
u
tco
m
es,
illu
s
tr
atin
g
t
h
e
ap
p
licatio
n
o
f
s
en
ti
m
e
n
t
a
n
al
y
s
is
i
n
p
o
liti
ca
l
f
o
r
ec
asti
n
g
b
u
t
ac
k
n
o
w
led
g
i
n
g
th
e
co
m
p
le
x
it
y
o
f
co
n
tex
t
u
al
i
n
ter
p
r
etatio
n
an
d
b
ias
m
i
tig
a
tio
n
.
Kr
o
u
s
k
a
et
a
l
.
[
5
]
c
o
n
d
u
cted
a
c
o
m
p
ar
ativ
e
ev
al
u
atio
n
o
f
s
en
ti
m
e
n
t
a
n
al
y
s
is
a
lg
o
r
it
h
m
s
o
v
er
s
o
cial
n
e
t
w
o
r
k
i
n
g
s
er
v
ices,
r
ev
ea
lin
g
d
is
cr
ep
an
cie
s
in
ac
c
u
r
ac
y
a
n
d
r
o
b
u
s
tn
es
s
ac
r
o
s
s
d
if
f
er
e
n
t
p
latf
o
r
m
s
an
d
d
ata
t
y
p
es.
Y
u
et
a
l
.
[
1
6
]
d
ev
elo
p
ed
a
m
et
h
o
d
to
p
r
ed
ict
p
ea
k
ti
m
e
p
o
p
u
lar
it
y
b
ased
o
n
T
w
it
ter
h
as
h
ta
g
s
,
s
h
o
w
ca
s
in
g
ad
v
a
n
c
e
m
en
ts
i
n
p
r
ed
ictiv
e
m
o
d
eli
n
g
b
u
t
r
ec
o
g
n
izi
n
g
li
m
ita
tio
n
s
in
d
ata
v
o
l
u
m
e
an
d
r
ea
l
-
ti
m
e
d
ata
p
r
o
ce
s
s
in
g
ca
p
ab
ilit
ies.
Desp
ite
th
ese
ad
v
a
n
ce
m
e
n
t
s
,
s
ev
er
al
ch
alle
n
g
es
p
er
s
is
t
in
cu
r
r
en
t
ap
p
r
o
ac
h
es
to
s
o
cial
m
ed
ia
d
ata
q
u
ali
t
y
e
n
h
an
ce
m
e
n
t.
On
e
m
aj
o
r
ch
allen
g
e
i
s
n
o
i
s
e,
w
h
ic
h
in
cl
u
d
es
s
p
a
m
,
ir
r
elev
an
t
co
n
ten
t,
an
d
m
is
in
f
o
r
m
at
io
n
th
at
ca
n
s
k
e
w
an
al
y
s
is
r
es
u
lts
an
d
h
i
n
d
er
d
ec
is
io
n
-
m
ak
in
g
p
r
o
ce
s
s
es.
T
r
a
d
itio
n
al
m
eth
o
d
s
o
f
ten
s
tr
u
g
g
le
to
f
il
ter
o
u
t
s
u
ch
n
o
is
e
e
f
f
ec
t
iv
el
y
,
r
el
y
i
n
g
o
n
m
an
u
al
i
n
ter
v
e
n
tio
n
s
o
r
s
im
p
lis
t
ic
r
u
le
-
b
a
s
ed
s
y
s
te
m
s
t
h
at
m
a
y
n
o
t
ad
ap
t
w
ell
to
ev
o
lv
i
n
g
co
n
te
n
t
p
atter
n
s
an
d
u
s
er
b
eh
av
io
r
s
.
An
o
th
er
cr
itical
ch
alle
n
g
e
is
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4864
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
,
Vo
l.
14
,
No
.
2
,
J
u
l
y
20
25
:
472
-
4
8
0
474
b
ias,
s
te
m
m
i
n
g
f
r
o
m
th
e
s
u
b
j
ec
tiv
e
n
atu
r
e
o
f
u
s
er
-
g
e
n
er
ated
co
n
ten
t
a
n
d
alg
o
r
ith
m
ic
b
iases
in
co
n
ten
t
cu
r
atio
n
a
n
d
r
ec
o
m
m
e
n
d
atio
n
s
y
s
te
m
s
.
B
ia
s
es
ca
n
lead
to
s
k
e
w
ed
d
ata
s
ets
th
at
d
o
n
o
t
a
cc
u
r
atel
y
r
ep
r
esen
t
th
e
d
iv
er
s
it
y
o
f
o
p
i
n
io
n
s
an
d
p
er
s
p
ec
tiv
es
w
it
h
i
n
s
o
cial
m
ed
ia
p
lat
f
o
r
m
s
,
i
m
p
ac
ti
n
g
th
e
r
eliab
ilit
y
o
f
an
al
y
tical
o
u
tco
m
es.
I
n
co
m
p
lete
n
ess
p
o
s
es
a
th
ir
d
s
ig
n
i
f
ica
n
t
c
h
alle
n
g
e,
c
h
ar
ac
te
r
ized
b
y
m
i
s
s
i
n
g
d
ata
p
o
in
ts
,
i
n
co
m
p
let
e
p
r
o
f
iles
,
an
d
g
ap
s
i
n
te
m
p
o
r
al
o
r
s
p
atial
co
v
er
ag
e.
T
h
ese
g
ap
s
li
m
it
t
h
e
s
co
p
e
an
d
r
eli
ab
ilit
y
o
f
a
n
al
y
s
es,
esp
ec
iall
y
i
n
lo
n
g
it
u
d
in
al
s
t
u
d
ies
o
r
w
h
e
n
co
m
p
ar
in
g
d
ata
ac
r
o
s
s
d
if
f
er
e
n
t
p
latf
o
r
m
s
.
Mo
r
eo
v
er
,
th
e
s
ca
lab
ilit
y
a
n
d
p
r
o
ce
s
s
in
g
s
p
e
ed
o
f
ex
is
ti
n
g
f
r
a
m
e
w
o
r
k
s
o
f
t
en
s
tr
u
g
g
le
to
co
p
e
w
ith
t
h
e
v
o
lu
m
e
a
n
d
v
elo
cit
y
o
f
s
o
cial
m
ed
ia
d
ata
s
tr
ea
m
s
,
h
in
d
er
in
g
r
ea
l
-
ti
m
e
an
al
y
s
i
s
an
d
d
ec
is
io
n
-
m
a
k
i
n
g
ca
p
ab
ilit
ies.
E
n
s
u
r
in
g
th
e
in
te
g
r
it
y
a
n
d
co
n
s
is
ten
c
y
o
f
d
ata
ac
r
o
s
s
d
is
tr
ib
u
ted
en
v
ir
o
n
m
en
ts
r
e
m
ai
n
s
a
p
er
s
is
ten
t
c
h
allen
g
e,
as
d
o
es
th
e
n
ee
d
f
o
r
r
o
b
u
s
t v
alid
atio
n
m
ec
h
an
i
s
m
s
to
v
er
i
f
y
th
e
ac
c
u
r
ac
y
an
d
r
eliab
ilit
y
o
f
e
x
tr
ac
ted
in
s
ig
h
ts
.
T
o
ad
d
r
ess
th
ese
ch
alle
n
g
e
s
,
th
e
p
r
o
p
o
s
ed
DQE
A
f
r
a
m
e
w
o
r
k
le
v
er
ag
es
ad
v
a
n
ce
d
d
ata
an
al
y
tic
s
tech
n
iq
u
es
to
en
h
a
n
ce
s
o
cial
m
ed
ia
d
ata
q
u
alit
y
s
y
s
te
m
a
ticall
y
.
Un
l
ik
e
tr
ad
itio
n
al
m
eth
o
d
s
,
t
h
e
DQE
A
f
r
a
m
e
w
o
r
k
in
teg
r
ate
s
au
to
m
a
ted
d
ata
p
r
o
ce
s
s
in
g
,
M
L
al
g
o
r
ith
m
s
,
an
d
n
a
tu
r
al
la
n
g
u
a
g
e
p
r
o
ce
s
s
in
g
(
N
L
P
)
tech
n
iq
u
es
to
tac
k
l
e
n
o
is
e,
b
ias,
an
d
in
co
m
p
leten
e
s
s
e
f
f
ec
t
iv
el
y
.
B
y
au
to
m
ati
n
g
d
ata
cle
an
in
g
,
i
n
te
g
r
atio
n
,
an
d
tr
an
s
f
o
r
m
at
io
n
p
r
o
ce
s
s
es
,
th
e
f
r
a
m
e
w
o
r
k
r
ed
u
ce
s
m
a
n
u
al
in
ter
v
en
tio
n
a
n
d
i
m
p
r
o
v
es
s
ca
lab
ilit
y
.
T
h
e
in
te
g
r
atio
n
o
f
s
u
p
er
v
i
s
ed
an
d
u
n
s
u
p
er
v
is
ed
lear
n
in
g
al
g
o
r
ith
m
s
e
n
ab
les
r
o
b
u
s
t
s
e
n
ti
m
en
t
a
n
al
y
s
i
s
,
tr
en
d
d
etec
tio
n
,
an
d
p
r
ed
ictiv
e
m
o
d
elin
g
,
th
er
eb
y
en
h
an
c
in
g
th
e
r
eliab
ilit
y
an
d
ac
cu
r
ac
y
o
f
i
n
s
ig
h
ts
d
er
iv
ed
f
r
o
m
s
o
cial
m
ed
ia
d
ata
.
3.
M
E
T
H
O
D
T
h
e
m
et
h
o
d
o
f
th
i
s
s
t
u
d
y
e
n
ta
ils
co
m
p
r
eh
e
n
s
i
v
e
d
ata
co
llectio
n
f
r
o
m
T
u
m
b
lr
,
f
o
c
u
s
i
n
g
o
n
g
ath
er
i
n
g
a
s
u
b
s
ta
n
tial
v
o
lu
m
e
o
f
d
iv
er
s
e
u
s
er
-
g
e
n
er
ated
co
n
te
n
t.
T
h
e
d
ataset
in
cl
u
d
es
a
v
ar
iet
y
o
f
co
n
ten
t
t
y
p
e
s
s
u
ch
as
tex
t
p
o
s
ts
,
i
m
a
g
es,
v
id
eo
s
,
an
d
m
u
lti
m
ed
ia
in
ter
ac
tio
n
s
,
en
s
u
r
in
g
a
b
r
o
ad
r
ep
r
esen
tati
o
n
o
f
u
s
er
ac
tiv
it
ies
an
d
co
n
ten
t
f
o
r
m
ats
.
Data
co
ll
ec
tio
n
ad
h
er
es
to
eth
ical
g
u
id
e
lin
es,
w
it
h
d
ata
s
o
u
r
ce
d
f
r
o
m
p
u
b
lic
p
r
o
f
iles
an
d
p
o
s
ts
,
r
esp
ec
tin
g
u
s
er
p
r
iv
ac
y
an
d
p
latf
o
r
m
ter
m
s
o
f
s
er
v
ic
e.
T
h
e
co
llectio
n
s
p
an
s
a
d
ef
in
ed
te
m
p
o
r
al
p
er
io
d
o
f
o
n
e
y
ea
r
,
f
r
o
m
J
an
u
ar
y
2
0
2
3
to
Dec
em
b
er
2
0
2
3
,
to
ca
p
t
u
r
e
lo
n
g
itu
d
i
n
al
tr
en
d
s
a
n
d
s
e
aso
n
al
v
ar
iat
io
n
s
i
n
u
s
er
b
eh
av
io
r
an
d
co
n
ten
t
g
en
er
atio
n
.
Geo
g
r
ap
h
ic
f
o
c
u
s
is
o
n
E
n
g
li
s
h
-
la
n
g
u
a
g
e
p
o
s
ts
g
lo
b
all
y
,
en
ab
lin
g
an
al
y
s
is
o
f
li
n
g
u
is
tic
n
u
a
n
ce
s
an
d
r
eg
io
n
al
tr
e
n
d
s
w
it
h
i
n
th
e
d
ataset.
T
h
e
DQE
A
f
r
a
m
e
w
o
r
k
i
n
te
g
r
ates
ad
v
an
ce
d
tech
n
o
lo
g
ies
an
d
to
o
ls
to
f
ac
ilit
ate
ef
f
icien
t
p
r
o
c
ess
i
n
g
,
an
al
y
s
is
,
an
d
v
alid
ati
o
n
o
f
s
o
cial
m
ed
i
a
d
ata:
3
.
1
.
Da
t
a
c
o
llect
io
n a
nd
inte
g
ra
t
io
n la
y
er
T
h
e
d
ata
c
o
llectio
n
an
d
in
teg
r
atio
n
la
y
er
w
it
h
in
t
h
e
DQE
A
f
r
a
m
e
w
o
r
k
is
cr
u
cial
f
o
r
ag
g
r
e
g
ati
n
g
a
n
d
h
ar
m
o
n
izi
n
g
d
iv
er
s
e
s
o
cial
m
ed
ia
co
n
te
n
t
f
r
o
m
p
lat
f
o
r
m
s
li
k
e
T
u
m
b
lr
.
T
h
is
la
y
er
e
m
p
lo
y
s
s
tr
u
ct
u
r
ed
p
r
o
ce
s
s
es
an
d
ad
v
an
ce
d
tech
n
iq
u
e
s
to
m
ai
n
tai
n
d
ata
in
teg
r
it
y
an
d
co
n
s
is
te
n
c
y
,
en
h
a
n
ci
n
g
t
h
e
q
u
alit
y
a
n
d
u
s
ab
ilit
y
o
f
t
h
e
co
llected
d
ata.
Data
ex
tr
ac
tio
n
i
n
v
o
l
v
es
r
etr
iev
i
n
g
co
m
p
r
e
h
en
s
i
v
e
d
atasets
f
r
o
m
T
u
m
b
lr
th
r
o
u
g
h
A
P
I
q
u
er
ies
an
d
w
eb
s
cr
ap
in
g
,
ad
h
er
in
g
to
p
latf
o
r
m
g
u
id
eli
n
es
to
en
s
u
r
e
leg
al
an
d
eth
ical
co
m
p
lia
n
ce
.
On
ce
e
x
tr
ac
ted
,
th
e
d
ata
u
n
d
er
g
o
e
s
r
ig
o
r
o
u
s
clea
n
i
n
g
to
r
e
m
o
v
e
n
o
is
e,
s
p
a
m
,
an
d
ir
r
elev
an
t
co
n
ten
t.
T
ex
tu
a
l
d
ata
is
p
r
o
ce
s
s
ed
u
s
in
g
N
L
P
tech
n
iq
u
e
s
,
i
n
clu
d
i
n
g
to
k
e
n
izatio
n
(
b
r
ea
k
in
g
te
x
t
i
n
to
w
o
r
d
s
)
,
s
to
p
-
w
o
r
d
r
em
o
v
al
(
f
i
lter
in
g
o
u
t
co
m
m
o
n
,
in
s
i
g
n
if
ican
t
w
o
r
d
s
)
,
an
d
s
tem
m
i
n
g
(
r
ed
u
cin
g
w
o
r
d
s
to
th
eir
r
o
o
t
f
o
r
m
)
.
Fo
r
m
u
lt
i
m
ed
ia
co
n
ten
t
s
u
c
h
a
s
i
m
ag
e
s
,
n
o
i
s
e
r
ed
u
ct
io
n
alg
o
r
it
h
m
s
ar
e
ap
p
lied
to
i
m
p
r
o
v
e
clar
it
y
an
d
r
e
m
o
v
e
ar
tif
ac
t
s
,
th
er
eb
y
en
h
an
ci
n
g
th
e
o
v
er
all
q
u
alit
y
o
f
v
is
u
al
d
ata.
Fig
u
r
e
1
illu
s
tr
ates
th
e
o
v
er
all
ar
ch
itect
u
r
e
o
f
th
e
p
r
o
p
o
s
ed
f
r
a
m
e
w
o
r
k
.
3
.
2
.
Te
x
t
prepro
ce
s
s
ing
T
ex
tu
al
d
ata
u
n
d
er
g
o
es
s
ev
er
al
p
r
ep
r
o
ce
s
s
in
g
s
tep
s
to
s
tan
d
ar
d
ize
an
d
en
h
a
n
ce
i
ts
an
al
y
s
i
s
r
ea
d
in
ess
.
T
h
ese
s
tep
s
i
n
cl
u
d
e:
−
T
o
k
en
izatio
n
:
T
o
k
en
izatio
n
b
r
ea
k
s
d
o
w
n
r
a
w
te
x
t
in
to
in
d
i
v
id
u
al
to
k
e
n
s
,
ty
p
icall
y
w
o
r
d
s
o
r
p
h
r
ases
.
I
t
f
o
r
m
s
th
e
f
o
u
n
d
atio
n
f
o
r
s
u
b
s
eq
u
e
n
t te
x
t
p
r
o
ce
s
s
in
g
ta
s
k
s
:
(
)
=
(
)
−
Ste
m
m
i
n
g
an
d
le
m
m
a
tizatio
n
:
Ste
m
m
i
n
g
r
ed
u
ce
s
w
o
r
d
s
to
t
h
eir
r
o
o
t
f
o
r
m
s
,
w
h
ile
le
m
m
at
izatio
n
e
n
s
u
r
es
w
o
r
d
s
ar
e
tr
an
s
f
o
r
m
ed
t
o
th
eir
b
ase
d
ictio
n
ar
y
f
o
r
m
:
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4864
Op
timiz
in
g
s
o
cia
l m
ed
ia
a
n
a
ly
tics
w
ith
th
e
d
a
ta
q
u
a
lity e
n
h
a
n
ce
men
t a
n
d
a
n
a
lytics
…
(
B.
K
a
r
th
ick
)
475
(
)
=
(
)
(
)
=
(
)
−
T
ex
t
n
o
r
m
a
lizatio
n
:
No
r
m
a
lizatio
n
s
ta
n
d
ar
d
izes
tex
t
b
y
r
e
m
o
v
i
n
g
p
u
n
ctu
at
io
n
,
s
p
ec
ial
ch
ar
ac
ter
s
,
an
d
co
n
v
er
t
in
g
tex
t
to
lo
w
er
ca
s
e:
(
)
=
(
)
−
Featu
r
e
r
ep
r
esen
tatio
n
(
T
F
-
I
DF)
:
TF
-
I
DF
q
u
an
t
if
ie
s
t
h
e
i
m
p
o
r
tan
ce
o
f
a
ter
m
w
i
th
in
a
d
o
cu
m
en
t
o
r
co
r
p
u
s
.
I
t
co
m
b
in
es
ter
m
f
r
eq
u
en
c
y
(
T
F)
an
d
in
v
er
s
e
d
o
cu
m
e
n
t
f
r
eq
u
e
n
c
y
(
I
DF)
:
(
,
)
=
,
∑
t
′
d
t
′
∈
d
(
,
)
=
(
∣
D
∣
|
{
d
∈
D
:
t
∈
d
}
|
)
−
(
,
)
=
(
,
)
(
,
)
w
h
er
e:
,
is
th
e
f
r
eq
u
en
c
y
o
f
te
r
m
t
in
d
o
cu
m
e
n
t
d
;
∣
D
∣
is
th
e
to
tal
n
u
m
b
er
o
f
d
o
cu
m
en
ts
i
n
th
e
co
r
p
u
s
D
;
an
d
∣
{
∈
:
∈
}
∣
is
th
e
n
u
m
b
er
o
f
d
o
cu
m
e
n
ts
co
n
tain
in
g
ter
m
t
w
i
th
in
t
h
e
co
r
p
u
s
D.
D
a
t
a
C
o
l
l
e
c
t
i
o
n
(
T
u
m
b
l
r
A
P
I
/
W
e
b
s
c
r
a
p
i
n
g
)
D
a
t
a
C
l
e
a
n
i
n
g
(
N
L
P
,
I
m
a
g
e
P
r
o
c
e
s
s
i
n
g
)
D
a
t
a
I
n
t
e
g
r
a
t
i
o
n
(
H
e
t
e
r
o
g
e
n
e
o
u
s
s
o
u
r
c
e
s
)
T
e
x
t
P
r
e
p
r
o
c
e
s
s
i
n
g
(
T
o
k
e
n
i
z
a
t
i
o
n
,
S
t
e
m
m
i
n
g
,
e
t
c
.
)
I
m
a
g
e
P
r
o
c
e
s
s
i
n
g
(
F
e
a
t
u
r
e
E
x
t
r
a
c
t
i
o
n
)
D
a
t
a
P
r
e
p
r
o
c
e
ss
i
n
g
a
n
d
F
e
a
t
u
r
e
Ex
t
r
a
c
t
i
o
n
F
e
a
t
u
r
e
Ex
t
r
a
c
t
i
o
n
(
T
F
-
I
D
F
,
I
m
a
g
e
F
e
a
t
u
r
e
s
)
V
a
l
i
d
a
t
i
o
n
a
n
d
V
e
r
i
f
i
c
a
t
i
o
n
V
i
s
u
a
l
i
z
a
t
i
o
n
a
n
d
R
e
p
o
r
t
i
n
g
L
a
y
e
r
En
t
i
t
y
R
e
c
o
g
n
i
t
i
o
n
(
S
p
a
C
y
,
N
LT
K
)
T
o
p
i
c
M
o
d
e
l
i
n
g
(
L
D
A
,
N
M
F
)
B
e
n
c
h
m
a
r
k
M
e
t
r
i
c
s
I
n
t
e
r
-
C
o
d
e
R
e
l
i
a
b
i
l
i
t
y
D
a
s
h
b
o
a
r
d
C
r
e
a
t
i
o
n
(
T
a
b
l
e
a
u
,
P
o
w
e
r
B
I
)
A
u
t
o
m
a
t
e
d
R
e
p
o
r
t
i
n
g
S
e
n
t
i
m
e
n
t
A
n
a
l
y
s
i
s
(
C
l
a
s
s
i
f
i
e
r
s
:
S
V
M
,
N
a
ï
v
e
B
a
y
e
s
)
Fig
u
r
e
1
.
Ov
er
all
ar
ch
itect
u
r
e
o
f
th
e
p
r
o
p
o
s
ed
DQE
A
3
.
3
.
M
a
chine
lea
rning
a
nd
na
t
ura
l la
ng
ua
g
e
pro
ce
s
s
ing
l
a
y
er
T
h
e
ML
an
d
NL
P
lay
er
o
f
t
h
e
DQE
A
f
r
a
m
e
w
o
r
k
is
i
n
te
g
r
al
f
o
r
d
er
iv
in
g
m
ea
n
i
n
g
f
u
l
i
n
s
ig
h
ts
f
r
o
m
s
o
cial
m
ed
ia
d
ata.
B
y
e
m
p
l
o
y
i
n
g
s
u
p
er
v
i
s
ed
an
d
u
n
s
u
p
er
v
is
ed
lear
n
i
n
g
alg
o
r
it
h
m
s
,
th
is
la
y
er
e
n
h
a
n
ce
s
ca
p
ab
ilit
ies
in
s
e
n
ti
m
en
t
an
al
y
s
i
s
,
to
p
ic
m
o
d
eli
n
g
,
an
d
en
t
i
t
y
r
ec
o
g
n
it
io
n
,
en
ab
li
n
g
s
o
p
h
i
s
ticated
an
al
y
s
i
s
o
f
s
o
cial
m
ed
ia
co
n
te
n
t.
−
Sen
ti
m
e
n
t
an
a
l
y
s
is
Sen
ti
m
e
n
t
an
al
y
s
is
in
v
o
l
v
es
d
eter
m
in
i
n
g
th
e
s
e
n
ti
m
en
t
o
r
em
o
tio
n
ex
p
r
ess
ed
in
te
x
tu
a
l
d
ata.
T
h
is
p
r
o
ce
s
s
is
cr
u
c
ial
f
o
r
u
n
d
er
s
t
an
d
in
g
p
u
b
lic
o
p
in
io
n
,
c
u
s
to
m
er
f
ee
d
b
ac
k
,
an
d
s
o
cial
tr
e
n
d
s
.
I
n
th
e
DQE
A
f
r
a
m
e
w
o
r
k
,
M
L
clas
s
i
f
ier
s
s
u
ch
as
n
ai
v
e
B
a
y
es
an
d
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
i
n
es
(
SV
M)
ar
e
u
tili
ze
d
f
o
r
p
r
ed
ictin
g
s
e
n
ti
m
e
n
t sco
r
es.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4864
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
,
Vo
l.
14
,
No
.
2
,
J
u
l
y
20
25
:
472
-
4
8
0
476
−
Naiv
e
B
a
y
es
cla
s
s
i
f
ier
:
T
h
e
n
aiv
e
B
a
y
e
s
clas
s
i
f
ier
is
b
ased
o
n
B
ay
e
s
'
t
h
eo
r
e
m
,
ass
u
m
in
g
i
n
d
ep
en
d
en
ce
b
et
w
ee
n
f
ea
t
u
r
es.
I
t
ca
lcu
late
s
th
e
p
r
o
b
ab
ilit
y
o
f
e
ac
h
s
e
n
ti
m
en
t
g
i
v
e
n
th
e
f
ea
t
u
r
es
in
t
h
e
tex
t
an
d
as
s
ig
n
s
th
e
s
en
ti
m
e
n
t
w
it
h
t
h
e
h
ig
h
e
s
t p
r
o
b
ab
ilit
y
:
̃
=
(
)
∏
(
∣
)
=
1
w
h
er
e:
̃
is
t
h
e
p
r
ed
icted
s
en
ti
m
en
t
;
P
(
y
)
is
t
h
e
p
r
io
r
p
r
o
b
ab
ilit
y
o
f
s
e
n
ti
m
e
n
t
y
,
a
n
d
(
∣
)
is
t
h
e
lik
eli
h
o
o
d
o
f
f
ea
t
u
r
e
x
i
g
iv
e
n
s
en
ti
m
e
n
t
y
.
−
SVM
:
SVM
is
a
p
o
w
er
f
u
l
cla
s
s
i
f
i
er
th
at
f
in
d
s
t
h
e
h
y
p
er
p
lan
e
s
ep
ar
atin
g
d
if
f
er
en
t
clas
s
es
w
it
h
th
e
m
ax
i
m
u
m
m
ar
g
in
.
Fo
r
s
en
ti
m
en
t
a
n
al
y
s
is
,
SVM
m
ap
s
i
n
p
u
t
tex
t
f
ea
t
u
r
es
to
a
h
i
g
h
er
-
d
i
m
en
s
io
n
al
s
p
ac
e
a
n
d
d
eter
m
in
e
s
th
e
o
p
ti
m
al
s
ep
ar
at
in
g
h
y
p
er
p
la
n
e:
̃
=
(
⋅
+
)
w
h
er
e:
̃
is
th
e
p
r
ed
icted
s
en
ti
m
en
t
;
w
is
t
h
e
w
ei
g
h
t v
ec
to
r
;
x
i
s
th
e
f
ea
tu
r
e
v
ec
to
r
; a
n
d
b
is
t
h
e
b
ias ter
m
.
Sen
ti
m
e
n
t
a
n
al
y
s
is
is
o
f
ten
b
r
o
k
en
d
o
w
n
i
n
to
s
e
v
er
al
s
tep
s
.
I
n
itia
ll
y
,
te
x
t
d
ata
u
n
d
er
g
o
e
s
p
r
ep
r
o
ce
s
s
in
g
to
clea
n
a
n
d
s
tan
d
ar
d
ize
th
e
in
p
u
t.
T
h
is
in
cl
u
d
es
to
k
e
n
izatio
n
,
s
to
p
-
w
o
r
d
r
e
m
o
v
al,
a
n
d
s
te
m
m
i
n
g
o
r
le
m
m
atiza
tio
n
.
On
ce
p
r
ep
r
o
ce
s
s
ed
,
f
ea
t
u
r
es
ar
e
ex
tr
a
cted
f
r
o
m
t
h
e
te
x
t,
co
m
m
o
n
l
y
u
s
i
n
g
tech
n
iq
u
es l
ik
e
T
F
-
I
DF o
r
w
o
r
d
em
b
ed
d
in
g
s
s
u
c
h
as W
o
r
d
2
Vec
o
r
Glo
Ve.
3
.
4
.
T
o
pic
m
o
delin
g
T
o
p
ic
m
o
d
eli
n
g
is
a
n
u
n
s
u
p
er
v
is
ed
lear
n
in
g
tec
h
n
iq
u
e
u
s
ed
to
u
n
co
v
er
laten
t to
p
ics i
n
a
c
o
llectio
n
o
f
d
o
cu
m
en
ts
.
T
w
o
p
o
p
u
lar
m
eth
o
d
s
ar
e
laten
t
Dir
ich
let
allo
ca
t
io
n
(
L
D
A
)
an
d
n
o
n
-
n
e
g
ati
v
e
m
atr
i
x
f
ac
to
r
izatio
n
(
NM
F).
L
D
A
ass
u
m
e
s
th
at
d
o
cu
m
e
n
t
s
ar
e
m
i
x
t
u
r
es
o
f
to
p
ics
an
d
th
at
to
p
ics
ar
e
d
is
tr
ib
u
tio
n
s
o
v
er
w
o
r
d
s
.
I
t
u
s
e
s
a
g
e
n
er
ativ
e
p
r
o
b
ab
ilis
tic
m
o
d
el
to
d
is
co
v
er
th
ese
to
p
ic
s
:
(
∣
,
)
=
(
∣
,
)
(
∣
)
(
∣
)
w
h
er
e:
(
∣
,
)
is
th
e
p
r
o
b
ab
ilit
y
o
f
t
o
p
ic
z
g
iv
en
d
o
cu
m
e
n
t
d
an
d
w
o
r
d
w
;
(
∣
,
)
is
th
e
p
r
o
b
ab
ilit
y
o
f
w
o
r
d
w
g
iv
e
n
to
p
ic
z
an
d
d
o
cu
m
e
n
t
d
;
(
∣
)
is
th
e
p
r
o
b
ab
ilit
y
o
f
to
p
ic
z
g
iv
e
n
d
o
cu
m
en
t
d
;
an
d
(
∣
)
is
th
e
p
r
o
b
ab
ilit
y
o
f
w
o
r
d
w
g
iv
en
d
o
cu
m
e
n
t d
.
I
n
L
D
A
,
ea
ch
d
o
cu
m
en
t
i
s
r
ep
r
esen
ted
as
a
d
is
tr
ib
u
tio
n
o
v
er
to
p
ics,
an
d
ea
c
h
to
p
ic
is
r
ep
r
esen
ted
as
a
d
is
tr
ib
u
tio
n
o
v
er
w
o
r
d
s
.
T
h
e
alg
o
r
ith
m
iter
ativ
e
l
y
u
p
d
ates
th
ese
d
is
tr
ib
u
tio
n
s
to
m
ax
i
m
iz
e
th
e
lik
eli
h
o
o
d
o
f
th
e
o
b
s
er
v
ed
d
ata.
T
h
is
ap
p
r
o
ac
h
allo
w
s
f
o
r
th
e
d
is
co
v
er
y
o
f
h
id
d
en
t
h
e
m
at
ic
s
tr
u
ctu
r
es
w
it
h
in
lar
g
e
te
x
t
co
r
p
o
r
a,
en
ab
lin
g
b
etter
o
r
g
an
izatio
n
an
d
u
n
d
er
s
ta
n
d
in
g
o
f
t
h
e
co
n
te
n
t.
−
No
n
-
n
e
g
ati
v
e
m
atr
i
x
f
ac
to
r
izatio
n
:
NM
F
f
ac
to
r
izes
t
h
e
d
o
cu
m
e
n
t
-
ter
m
m
atr
i
x
V
in
to
t
w
o
lo
wer
-
d
i
m
e
n
s
io
n
al
m
a
tr
ices
W
an
d
H
s
u
c
h
th
at:
≈
w
h
er
e:
V
is
t
h
e
d
o
cu
m
e
n
t
-
ter
m
m
a
tr
ix
;
W
is
t
h
e
d
o
cu
m
en
t
-
to
p
ic
m
atr
i
x
,
an
d
H
is
t
h
e
to
p
i
c
-
ter
m
m
atr
i
x
.
4.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
e
DQE
A
f
r
a
m
e
w
o
r
k
w
as
te
s
ted
u
s
i
n
g
a
lar
g
e
d
ataset
o
b
tain
ed
f
r
o
m
T
u
m
b
lr
,
an
d
its
p
er
f
o
r
m
a
n
ce
w
a
s
v
al
id
ated
ag
ai
n
s
t
h
u
m
a
n
co
d
er
s
f
r
o
m
Am
az
o
n
Me
c
h
an
ical
T
u
r
k
.
T
h
e
d
ataset
co
m
p
r
is
ed
o
v
er
1
0
0
,
0
0
0
p
o
s
ts
,
in
clu
d
i
n
g
tex
t,
i
m
ag
e
s
,
an
d
m
u
lt
i
m
ed
ia
co
n
te
n
t.
T
h
e
i
m
p
le
m
e
n
tat
io
n
e
n
v
ir
o
n
m
en
t
i
n
clu
d
ed
P
y
t
h
o
n
f
o
r
d
ata
p
r
o
ce
s
s
in
g
,
N
L
P
,
an
d
M
L
ta
s
k
s
,
w
it
h
lib
r
ar
ies
s
u
ch
a
s
P
an
d
as,
Sci
k
it
-
lear
n
,
Sp
aC
y
,
an
d
T
en
s
o
r
Flo
w
.
P
y
t
h
o
n
s
er
v
ed
as
t
h
e
co
r
e
p
r
o
g
r
am
m
i
n
g
lan
g
u
ag
e
f
o
r
i
m
p
le
m
e
n
ti
n
g
t
h
e
DQE
A
f
r
a
m
e
w
o
r
k
d
u
e
to
its
v
er
s
atili
t
y
a
n
d
r
o
b
u
s
t s
u
p
p
o
r
t f
o
r
d
ata
an
aly
tics
a
n
d
ML
.
4
.
1
.
Senti
m
ent
a
na
ly
s
is
perf
o
r
m
a
nce
T
h
e
s
en
ti
m
e
n
t
a
n
al
y
s
i
s
m
o
d
e
ls
—
n
aiv
e
B
a
y
es
,
SVM,
a
n
d
DQE
A
(
p
r
o
p
o
s
ed
)
—
o
p
er
ate
o
n
tex
tu
al
d
ata
ex
tr
ac
ted
f
r
o
m
T
u
m
b
lr
.
T
u
m
b
lr
s
er
v
es
a
s
t
h
e
p
r
i
m
ar
y
d
ata
s
o
u
r
ce
,
co
n
tai
n
i
n
g
a
d
iv
er
s
e
r
an
g
e
o
f
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4864
Op
timiz
in
g
s
o
cia
l m
ed
ia
a
n
a
ly
tics
w
ith
th
e
d
a
ta
q
u
a
lity e
n
h
a
n
ce
men
t a
n
d
a
n
a
lytics
…
(
B.
K
a
r
th
ick
)
477
u
s
er
-
g
e
n
er
ated
co
n
te
n
t
i
n
clu
d
i
n
g
b
lo
g
p
o
s
ts
,
co
m
m
e
n
ts
,
a
n
d
m
u
lti
m
ed
ia
ca
p
tio
n
s
.
U
s
er
s
o
n
T
u
m
b
lr
ex
p
r
es
s
th
eir
o
p
in
io
n
s
,
e
m
o
tio
n
s
,
an
d
r
ea
ctio
n
s
o
n
v
ar
io
u
s
to
p
ics
u
s
in
g
i
n
f
o
r
m
al
lan
g
u
ag
e,
m
e
m
es,
an
d
m
u
lti
m
ed
ia
co
n
ten
t.
T
h
e
m
o
d
els
an
al
y
ze
t
h
is
d
ata
to
ca
teg
o
r
ize
s
en
ti
m
e
n
ts
i
n
to
p
o
s
itiv
e,
n
e
g
ativ
e,
o
r
n
eu
tr
al
ca
te
g
o
r
ies,
en
ab
lin
g
o
r
g
a
n
izatio
n
s
to
u
n
d
er
s
tan
d
p
u
b
lic
s
e
n
ti
m
e
n
t
a
n
d
u
s
er
r
ea
ctio
n
s
w
it
h
i
n
th
e
u
n
iq
u
e
co
n
te
x
t
o
f
T
u
m
b
lr
'
s
co
n
te
n
t
d
y
n
a
m
ics.
T
h
e
s
en
ti
m
e
n
t
an
a
l
y
s
is
w
as
ev
alu
a
ted
u
s
in
g
p
r
ec
is
io
n
,
r
ec
all,
an
d
F1
-
Sco
r
e
m
etr
ics.
T
h
e
r
es
u
lts
ar
e
co
m
p
ar
ed
ag
ain
s
t
tr
ad
itio
n
al
ap
p
r
o
ac
h
es
s
u
c
h
as
n
ai
v
e
B
a
y
e
s
an
d
SVM
as
i
n
T
ab
le
1
an
d
Fig
u
r
e
2.
T
ab
le
1
.
Sen
ti
m
e
n
t
an
al
y
s
is
p
e
r
f
o
r
m
an
ce
M
o
d
e
l
P
r
e
c
i
si
o
n
R
e
c
a
l
l
F1
-
S
c
o
r
e
N
a
i
v
e
B
a
y
e
s
0
.
8
1
0
.
7
8
0
.
7
9
S
V
M
0
.
8
4
0
.
8
0
0
.
8
2
R
a
n
d
o
m
f
o
r
e
st
0
.
8
6
0
.
8
2
0
.
8
4
D
Q
EA
(
p
r
o
p
o
se
d
)
0
.
8
9
0
.
8
6
0
.
8
7
E_
B
D
M
S
N
/
A
N
/
A
0
.
8
6
Fig
u
r
e
2
.
Sen
ti
m
e
n
tal
an
a
l
y
s
is
p
er
f
o
r
m
a
n
ce
T
h
e
s
en
ti
m
e
n
t
an
al
y
s
is
p
er
f
o
r
m
an
ce
o
f
v
ar
io
u
s
m
o
d
el
s
,
in
c
lu
d
in
g
n
ai
v
e
B
a
y
es
,
SVM,
th
e
p
r
o
p
o
s
ed
DQE
A
f
r
a
m
e
w
o
r
k
,
an
d
t
h
e
p
r
ev
io
u
s
E
-
B
DM
S
ap
p
r
o
ac
h
.
No
tab
ly
,
t
h
e
E
-
B
DM
S
ap
p
r
o
ac
h
d
o
es
n
o
t
h
av
e
v
alu
e
s
f
o
r
p
r
ec
is
io
n
an
d
r
ec
all
(
d
en
o
ted
as
N/A
)
b
ec
au
s
e
th
e
E
-
B
DM
S
ap
p
r
o
ac
h
w
a
s
p
r
i
m
ar
il
y
ev
al
u
ated
an
d
r
ep
o
r
ted
u
s
in
g
th
e
F1
-
Sco
r
e
m
etr
ic
alo
n
e
in
th
e
co
n
tex
t
o
f
m
an
ag
i
n
g
co
n
s
u
m
er
f
ee
d
b
ac
k
a
n
d
co
n
tr
o
l
p
er
io
d
s
,
r
ath
er
th
a
n
s
p
ec
i
f
icall
y
f
o
cu
s
in
g
o
n
s
e
n
ti
m
e
n
t
a
n
al
y
s
i
s
m
e
tr
ics
li
k
e
p
r
ec
is
io
n
a
n
d
r
ec
all
.
Desp
ite
th
is
,
t
h
e
F1
-
Sc
o
r
e
o
f
th
e
E
-
B
DM
S
ap
p
r
o
ac
h
s
tan
d
s
at
0
.
8
6
,
w
h
ic
h
i
s
m
ar
g
i
n
all
y
lo
w
er
th
a
n
th
e
D
QE
A
f
r
a
m
e
w
o
r
k
’
s
F1
-
Sco
r
e
o
f
0
.
8
7
.
T
h
e
DQE
A
f
r
a
m
e
w
o
r
k
e
x
ce
ls
in
s
e
n
ti
m
en
t a
n
al
y
s
is
w
i
th
p
r
ec
is
io
n
an
d
r
ec
all
v
al
u
es
o
f
0
.
8
9
an
d
0
.
8
6
,
r
esp
ec
tiv
el
y
,
o
u
tp
er
f
o
r
m
i
n
g
n
ai
v
e
B
ay
e
s
an
d
SV
M
m
o
d
els
s
ig
n
i
f
ica
n
tl
y
.
Nai
v
e
B
ay
es
ac
h
iev
ed
a
p
r
ec
is
io
n
o
f
0
.
8
1
an
d
r
ec
all
o
f
0
.
7
8
,
r
esu
ltin
g
in
a
n
F1
-
Sco
r
e
o
f
0
.
7
9
,
w
h
ile
SV
M
p
er
f
o
r
m
ed
b
etter
w
it
h
a
p
r
ec
is
io
n
o
f
0
.
8
4
,
r
ec
all
o
f
0
.
8
0
,
an
d
an
F1
-
Sco
r
e
o
f
0
.
8
2
.
4
.
2
.
T
o
pic
m
o
delin
g
perf
o
r
m
a
nce
T
h
e
to
p
ic
m
o
d
elin
g
p
er
f
o
r
m
a
n
ce
w
as
ev
a
lu
ated
u
s
in
g
co
h
er
e
n
ce
s
co
r
es,
w
h
ic
h
m
ea
s
u
r
e
th
e
s
e
m
an
tic
s
i
m
ilar
it
y
b
et
w
ee
n
h
ig
h
-
s
co
r
i
n
g
w
o
r
d
s
i
n
a
to
p
ic.
T
ex
tu
a
l
d
ata
f
r
o
m
T
u
m
b
lr
p
o
s
ts
was
u
s
ed
f
o
r
to
p
ic
m
o
d
eli
n
g
.
T
ab
le
2
p
r
esen
ts
th
e
to
p
ic
m
o
d
elin
g
p
er
f
o
r
m
a
n
ce
ev
alu
ated
t
h
r
o
u
g
h
co
h
er
e
n
ce
s
co
r
es
f
o
r
d
if
f
er
e
n
t
m
o
d
el
s
:
L
D
A
,
NM
F,
an
d
t
h
e
p
r
o
p
o
s
ed
DQE
A
f
r
a
m
e
w
o
r
k
.
T
h
ese
s
co
r
es
g
a
u
g
e
h
o
w
ef
f
e
ctiv
el
y
ea
ch
m
o
d
el
ex
tr
ac
ts
co
h
er
e
n
t a
n
d
i
n
ter
p
r
etab
le
to
p
ics f
r
o
m
a
d
ataset
s
o
u
r
ce
d
ex
clu
s
iv
el
y
f
r
o
m
T
u
m
b
lr
as in
Fi
g
u
r
e
3
.
Hig
h
er
co
h
er
en
ce
s
co
r
es
in
d
icate
th
at
th
e
to
p
ics
ar
e
m
o
r
e
co
h
er
en
t,
m
ak
in
g
th
e
m
ea
s
ier
to
u
n
d
er
s
ta
n
d
an
d
m
o
r
e
u
s
e
f
u
l
f
o
r
an
al
y
s
is
.
ℎ
=
1
∑
ℎ
(
)
=
1
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4864
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
,
Vo
l.
14
,
No
.
2
,
J
u
l
y
20
25
:
472
-
4
8
0
478
w
h
er
e
is
th
e
s
et
o
f
to
p
w
o
r
d
s
i
n
to
p
ic
iii an
d
NNN
is
t
h
e
to
ta
l n
u
m
b
er
o
f
to
p
ics.
T
ab
le
2
.
T
o
p
ic
m
o
d
elin
g
p
er
f
o
r
m
an
ce
M
o
d
e
l
C
o
h
e
r
e
n
c
e
sco
r
e
L
D
A
0
.
4
8
N
M
F
0
.
5
2
D
Q
EA
(
p
r
o
p
o
se
d
)
0
.
6
3
E_
B
D
M
S
N
/
A
Fig
u
r
e
3
.
T
o
p
ic
m
o
d
elin
g
p
er
f
o
r
m
a
n
ce
T
h
e
DQE
A
f
r
a
m
e
w
o
r
k
ac
h
ie
v
ed
a
co
h
er
e
n
ce
s
co
r
e
o
f
0
.
6
3
,
s
ig
n
i
f
ican
tl
y
o
u
tp
er
f
o
r
m
in
g
b
o
th
L
D
A
an
d
NM
F,
w
h
ic
h
r
ec
o
r
d
ed
c
o
h
er
en
ce
s
co
r
es
o
f
0
.
4
8
an
d
0
.
5
2
,
r
esp
ec
tiv
el
y
.
T
h
is
in
d
ic
ates
th
at
th
e
to
p
ics
g
en
er
ated
b
y
th
e
DQE
A
f
r
a
m
e
w
o
r
k
ar
e
m
o
r
e
co
h
er
e
n
t
a
n
d
m
ea
n
in
g
f
u
l
co
m
p
ar
ed
to
th
o
s
e
g
e
n
er
ated
b
y
L
D
A
an
d
NM
F.
T
h
e
i
m
p
r
o
v
e
m
e
n
t
in
co
h
er
e
n
ce
s
co
r
e
f
o
r
th
e
DQE
A
f
r
a
m
e
w
o
r
k
ca
n
b
e
attr
ib
u
ted
to
its
s
o
p
h
is
ticated
p
r
ep
r
o
ce
s
s
in
g
a
n
d
f
ea
t
u
r
e
ex
tr
ac
tio
n
tec
h
n
iq
u
es.
Th
e
r
esu
lts
in
m
o
r
e
ac
cu
r
a
te
an
d
in
ter
p
r
etab
le
to
p
ics.
L
D
A
,
w
i
th
a
co
h
er
en
ce
s
co
r
e
o
f
0
.
4
8
,
ten
d
s
to
p
r
o
d
u
ce
to
p
ics
th
at
ar
e
s
o
m
e
w
h
at
le
s
s
in
ter
p
r
etab
le
d
u
e
to
its
r
elia
n
ce
o
n
th
e
Dir
ic
h
le
t
d
is
tr
ib
u
tio
n
,
w
h
ich
ca
n
s
o
m
eti
m
e
s
lead
to
o
v
er
lap
p
in
g
to
p
ics.
NM
F,
w
i
th
a
s
lig
h
tl
y
b
etter
co
h
er
en
ce
s
co
r
e
o
f
0
.
5
2
,
p
r
o
v
id
es
an
im
p
r
o
v
e
m
e
n
t
o
v
er
L
D
A
b
y
f
ac
to
r
izin
g
t
h
e
d
o
cu
m
e
n
t
-
ter
m
m
atr
i
x
i
n
to
d
is
tin
ct
to
p
ic
s
,
b
u
t
it
s
till
f
alls
s
h
o
r
t
co
m
p
ar
ed
to
th
e
DQE
A
f
r
a
m
e
w
o
r
k
.
T
ab
le
3
ev
alu
ates
th
e
n
a
m
ed
en
t
it
y
r
ec
o
g
n
itio
n
(
NE
R
)
p
er
f
o
r
m
an
ce
o
f
th
r
ee
m
o
d
els:
Sp
aC
y
,
N
L
T
K,
an
d
th
e
p
r
o
p
o
s
ed
.
T
ab
le
3
.
NE
R
p
er
f
o
r
m
an
ce
M
o
d
e
l
P
r
e
c
i
si
o
n
R
e
c
a
l
l
F1
-
S
c
o
r
e
S
p
a
C
y
0
.
8
5
0
.
8
2
0
.
8
3
N
L
TK
0
.
8
0
0
.
7
7
0
.
7
8
D
Q
EA
(
p
r
o
p
o
se
d
)
0
.
8
8
0
.
8
5
0
.
8
6
e
-
B
D
M
S
N
/
A
N
/
A
0
.
8
5
T
h
e
E
-
B
DM
S
ap
p
r
o
ac
h
h
as
N/
A
f
o
r
p
r
ec
is
io
n
an
d
r
ec
all
b
ec
au
s
e,
s
i
m
ilar
to
its
s
e
n
ti
m
en
t
an
al
y
s
is
ev
alu
a
tio
n
,
it
w
as
p
r
i
m
ar
il
y
as
s
ess
ed
u
s
i
n
g
t
h
e
F1
-
Sco
r
e
m
e
tr
ic
f
o
r
d
if
f
er
en
t
co
n
tex
t
s
an
d
ap
p
licatio
n
s
r
ath
er
th
an
s
p
ec
i
f
icall
y
f
o
r
NE
R
task
s
.
Desp
ite
th
is
,
th
e
E
-
B
DM
S
ap
p
r
o
ac
h
ac
h
iev
ed
an
F1
-
Sco
r
e
o
f
0
.
8
5
,
w
h
ic
h
is
s
lig
h
tl
y
lo
w
er
th
a
n
t
h
e
DQE
A
f
r
a
m
e
w
o
r
k
’
s
F1
-
Sco
r
e
o
f
0
.
8
6
.
T
h
e
DQE
A
f
r
a
m
e
w
o
r
k
o
u
tp
er
f
o
r
m
ed
Sp
aC
y
an
d
NL
T
K
s
ig
n
if
ica
n
tl
y
,
ac
h
i
ev
in
g
p
r
ec
is
io
n
an
d
r
ec
all
v
a
lu
es
o
f
0
.
8
8
an
d
0
.
8
5
,
r
esp
ec
tiv
el
y
.
I
n
co
n
tr
ast,
Sp
aC
y
ac
h
iev
ed
a
p
r
ec
is
io
n
o
f
0
.
8
5
an
d
r
ec
all
o
f
0
.
8
2
,
r
esu
ltin
g
in
an
F1
-
Sco
r
e
o
f
0
.
8
3
,
w
h
ile
N
L
T
K
h
ad
a
p
r
ec
is
io
n
o
f
0
.
8
0
,
r
ec
all
o
f
0
.
7
7
,
an
d
an
F1
-
Sco
r
e
o
f
0
.
7
8
.
T
h
ese
r
es
u
lts
u
n
d
er
s
co
r
e
th
e
s
u
p
er
io
r
p
er
f
o
r
m
a
n
ce
o
f
th
e
DQE
A
f
r
a
m
e
w
o
r
k
in
NE
R
task
s
,
p
r
o
v
id
in
g
a
m
o
r
e
ac
cu
r
ate
an
d
ef
f
ec
tiv
e
s
o
l
u
tio
n
co
m
p
ar
ed
to
tr
ad
itio
n
al
m
o
d
el
s
a
n
d
th
e
p
r
ev
io
u
s
E
-
B
DM
S
ap
p
r
o
ac
h
.
T
ab
le
4
s
u
m
m
ar
izes
th
e
r
es
u
lts
o
b
tain
ed
f
r
o
m
C
N
N
an
al
y
s
is
:
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4864
Op
timiz
in
g
s
o
cia
l m
ed
ia
a
n
a
ly
tics
w
ith
th
e
d
a
ta
q
u
a
lity e
n
h
a
n
ce
men
t a
n
d
a
n
a
lytics
…
(
B.
K
a
r
th
ick
)
479
T
ab
le
4
.
C
NN
an
al
y
s
is
r
es
u
lt
s
M
o
d
e
l
A
c
c
u
r
a
c
y
T
r
u
e
p
o
si
t
i
v
e
r
a
t
e
S
e
n
si
t
i
v
i
t
y
S
p
e
c
i
f
i
c
i
t
y
C
N
N
(
R
e
sN
e
t
)
0
.
9
2
0
.
8
8
0
.
8
7
0
.
9
3
C
N
N
(
V
G
G
1
6
)
0
.
8
8
0
.
8
5
0
.
8
4
0
.
9
0
C
N
N
(
I
n
c
e
p
t
i
o
n
)
0
.
9
1
0
.
8
7
0
.
8
6
0
.
9
2
T
h
e
C
NN
m
o
d
els
in
teg
r
ated
in
to
t
h
e
DQE
A
f
r
a
m
e
w
o
r
k
a
ch
iev
ed
h
i
g
h
ac
cu
r
ac
y
an
d
tr
u
e
p
o
s
iti
v
e
r
ates
in
class
i
f
y
in
g
i
m
a
g
es
ex
t
r
ac
ted
f
r
o
m
s
o
cial
m
ed
ia
p
o
s
ts
.
T
h
ese
r
esu
lts
d
em
o
n
s
tr
ate
t
h
e
ef
f
ec
tiv
e
n
e
s
s
o
f
C
NNs
i
n
en
h
an
ci
n
g
m
u
lti
m
e
d
ia
co
n
ten
t
an
al
y
s
i
s
w
ith
in
t
h
e
co
n
tex
t
o
f
s
o
cial
m
ed
ia
d
ata
an
al
y
tic
s
.
T
h
e
o
v
er
all
p
er
f
o
r
m
a
n
ce
m
etr
ics
i
s
s
h
o
w
n
in
T
ab
le
5.
T
h
e
r
esu
lts
clea
r
l
y
in
d
i
ca
te
th
at
th
e
DQE
A
f
r
a
m
e
w
o
r
k
s
ig
n
i
f
ica
n
tl
y
e
n
h
an
ce
s
t
h
e
q
u
a
lit
y
a
n
d
r
eliab
ilit
y
o
f
s
o
cial
m
ed
ia
d
ata
an
al
y
tics
.
T
ab
le
5
.
Ov
er
all
p
er
f
o
r
m
a
n
ce
m
etr
ics
M
e
t
r
i
c
N
a
i
v
e
B
a
y
e
s
S
V
M
L
D
A
N
M
F
S
p
a
C
y
N
L
TK
D
Q
EA
(
p
r
o
p
o
se
d
)
S
e
n
t
i
me
n
t
a
n
a
l
y
si
s
(
F
1
)
0
.
7
9
0
.
8
2
N
/
A
N
/
A
N
/
A
N
/
A
0
.
8
7
T
o
p
i
c
mo
d
e
l
i
n
g
(
c
o
h
e
r
e
n
c
e
)
N
/
A
N
/
A
0
.
4
8
0
.
5
2
N
/
A
N
/
A
0
.
6
3
N
ER
(
F
1
)
N
/
A
N
/
A
N
/
A
N
/
A
0
.
8
3
0
.
7
8
0
.
8
6
C
N
N
N
/
A
N
/
A
N
/
A
N
/
A
N
/
A
N
/
A
0
.
9
2
5.
CO
NCLU
SI
O
N
T
h
is
p
ap
er
in
tr
o
d
u
ce
s
th
e
DQE
A
f
r
a
m
e
w
o
r
k
,
w
h
ic
h
ad
d
r
ess
es
k
e
y
ch
al
len
g
es
i
n
an
al
y
zi
n
g
T
u
m
b
l
r
d
ata.
B
y
i
n
te
g
r
ati
n
g
ad
v
a
n
ce
d
d
ata
an
al
y
tics
tech
n
iq
u
e
s
w
i
th
ML
a
n
d
N
L
P
alg
o
r
it
h
m
s
,
t
h
e
DQE
A
f
r
a
m
e
w
o
r
k
s
ig
n
i
f
ica
n
tl
y
en
h
a
n
ce
s
d
ata
q
u
alit
y
,
s
e
n
ti
m
en
t
an
al
y
s
is
,
to
p
ic
m
o
d
elin
g
,
an
d
NE
R
.
E
m
p
ir
ical
ev
alu
at
io
n
s
d
em
o
n
s
tr
ate
t
h
at
th
e
DQE
A
f
r
a
m
e
w
o
r
k
s
u
r
p
as
s
es
ex
i
s
ti
n
g
m
e
th
o
d
s
in
p
r
ec
is
io
n
,
r
ec
all,
an
d
co
h
er
en
ce
in
to
p
ic
m
o
d
eli
n
g
,
h
i
g
h
lig
h
ti
n
g
its
ef
f
ec
t
iv
e
n
es
s
i
n
p
r
o
v
id
in
g
ac
cu
r
ate
i
n
s
ig
h
t
s
f
r
o
m
T
u
m
b
lr
d
ataset
s
.
T
h
is
f
r
a
m
e
w
o
r
k
n
o
t o
n
l
y
i
m
p
r
o
v
e
s
d
ec
is
io
n
-
m
a
k
i
n
g
p
r
o
ce
s
s
es b
u
t
also
ad
v
an
ce
s
r
esear
ch
in
s
o
ci
al
m
ed
ia
a
n
al
y
tics
b
y
lev
er
a
g
i
n
g
s
tate
-
of
-
t
h
e
-
ar
t
tech
n
iq
u
es
tailo
r
ed
to
T
u
m
b
lr
's
u
n
iq
u
e
ch
ar
ac
ter
i
s
tics
.
T
h
e
i
m
p
licatio
n
s
o
f
t
h
i
s
w
o
r
k
ar
e
s
u
b
s
ta
n
tial,
o
f
f
er
in
g
a
m
o
r
e
r
ef
in
ed
to
o
l
f
o
r
an
aly
zin
g
s
o
cial
m
ed
ia
d
ata
an
d
p
o
ten
tiall
y
b
en
e
f
iti
n
g
d
ec
is
io
n
-
m
a
k
i
n
g
in
v
ar
io
u
s
co
n
tex
ts
.
F
u
t
u
r
e
r
esear
ch
w
i
ll
f
o
cu
s
o
n
e
x
p
an
d
i
n
g
th
e
f
r
a
m
e
wo
r
k
’
s
ap
p
licatio
n
to
o
th
er
s
o
cial
m
ed
ia
p
latf
o
r
m
s
,
en
h
a
n
ci
n
g
alg
o
r
it
h
m
ac
cu
r
ac
y
,
an
d
ex
p
lo
r
in
g
r
ea
l
-
ti
m
e
d
ata
p
r
o
ce
s
s
in
g
.
T
h
ese
ad
v
an
ce
m
en
ts
w
ill
s
tr
en
g
t
h
en
th
e
f
r
a
m
e
w
o
r
k
’
s
i
m
p
ac
t
o
n
th
e
f
ield
,
co
n
tr
ib
u
ti
n
g
to
m
o
r
e
i
n
s
i
g
h
t
f
u
l
a
n
d
ti
m
el
y
an
al
y
s
es i
n
s
o
cial
m
ed
ia
r
esea
r
ch
.
RE
F
E
R
E
NC
E
S
[
1
]
A
.
A
b
b
a
s
i
,
J.
L
i
,
G
.
C
l
i
ff
o
r
d
,
a
n
d
H
.
T
a
y
l
o
r
,
“
M
a
k
e
‘
f
a
i
r
n
e
ss
b
y
d
e
si
g
n
’
p
a
r
t
o
f
mac
h
i
n
e
l
e
a
r
n
i
n
g
,
”
H
a
rv
a
rd
Bu
si
n
e
ss
R
e
v
i
e
w
,
2
0
1
8
.
[
2
]
G
.
B
e
r
a
r
d
i
,
A
.
Esu
l
i
,
D
.
M
a
r
c
h
e
g
g
i
a
n
i
,
a
n
d
F
.
S
e
b
a
s
t
i
a
n
i
,
“
I
S
TI
@
T
R
EC
M
i
c
r
o
b
l
o
g
T
r
a
c
k
:
Ex
p
l
o
r
i
n
g
t
h
e
u
se
o
f
h
a
sh
t
a
g
se
g
me
n
t
a
t
i
o
n
a
n
d
t
e
x
t
q
u
a
l
i
t
y
r
a
n
k
i
n
g
,
”
in
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
2
0
t
h
T
e
x
t
REt
r
i
e
v
a
l
C
o
n
f
e
re
n
c
e
(
T
RE
C
2
0
1
1
)
,
Vo
l
u
m
e
S
p
e
c
i
a
l
Pu
b
l
i
c
a
t
i
o
n
,
2
0
1
1
.
[
3
]
G
.
A
d
o
mav
i
c
i
u
s
,
J.
B
o
c
k
st
e
d
t
,
a
n
d
S
.
P
.
C
u
r
l
e
y
,
“
B
u
n
d
l
i
n
g
Ef
f
e
c
t
s
o
n
V
a
r
i
e
t
y
S
e
e
k
i
n
g
f
o
r
D
i
g
i
t
a
l
I
n
f
o
r
mat
i
o
n
G
o
o
d
s,”
J
o
u
rn
a
l
o
f
M
a
n
a
g
e
m
e
n
t
I
n
f
o
rm
a
t
i
o
n
S
y
s
t
e
m
s
,
v
o
l
.
3
1
,
n
o
.
4
,
p
p
.
1
8
2
–
2
1
2
,
Ja
n
.
2
0
1
5
,
d
o
i
:
1
0
.
1
0
8
0
/
0
7
4
2
1
2
2
2
.
2
0
1
4
.
1
0
0
1
2
6
6
.
[
4
]
J.
A
g
r
a
w
a
l
a
n
d
W
.
A
.
K
a
mak
u
r
a
,
“
T
h
e
Ec
o
n
o
mi
c
W
o
r
t
h
o
f
C
e
l
e
b
r
i
t
y
E
n
d
o
r
se
r
s:
A
n
Ev
e
n
t
S
t
u
d
y
A
n
a
l
y
si
s,”
J
o
u
rn
a
l
o
f
Ma
r
k
e
t
i
n
g
,
v
o
l
.
5
9
,
n
o
.
3
,
p
p
.
5
6
–
6
2
,
J
u
l
.
1
9
9
5
,
d
o
i
:
1
0
.
1
1
7
7
/
0
0
2
2
2
4
2
9
9
5
0
5
9
0
0
3
0
5
.
[
5
]
A
.
K
r
o
u
sk
a
,
C
.
T
r
o
u
ssas
,
a
n
d
M
.
V
i
r
v
o
u
,
“
C
o
mp
a
r
a
t
i
v
e
e
v
a
l
u
a
t
i
o
n
o
f
a
l
g
o
r
i
t
h
ms
f
o
r
se
n
t
i
me
n
t
a
n
a
l
y
si
s
o
v
e
r
so
c
i
a
l
n
e
t
w
o
r
k
i
n
g
se
r
v
i
c
e
s,”
J
o
u
r
n
a
l
o
f
U
n
i
v
e
rsa
l
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
2
3
,
n
o
.
8
,
p
p
.
7
5
5
–
7
6
8
,
2
0
1
7
.
[
6
]
R
.
A
r
u
n
,
V
.
S
u
r
e
sh
,
C
.
E.
V
.
M
a
d
h
a
v
a
n
,
a
n
d
M
.
N
.
N
.
M
u
r
t
h
y
,
“
O
n
F
i
n
d
i
n
g
t
h
e
N
a
t
u
r
a
l
N
u
m
b
e
r
o
f
T
o
p
i
c
s
w
i
t
h
L
a
t
e
n
t
D
i
r
i
c
h
l
e
t
A
l
l
o
c
a
t
i
o
n
:
S
o
me
O
b
se
r
v
a
t
i
o
n
s
,
”
i
n
I
n
P
a
c
i
fi
c
-
Asi
a
C
o
n
f
e
r
e
n
c
e
o
n
K
n
o
w
l
e
d
g
e
D
i
s
c
o
v
e
r
y
a
n
d
D
a
t
a
Mi
n
i
n
g
,
2
0
1
0
,
p
p
.
3
9
1
–
4
0
2
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
642
-
1
3
6
5
7
-
3
_
4
3
.
[
7
]
R
.
R
e
z
a
p
o
u
r
,
L
.
W
a
n
g
,
O
.
A
b
d
a
r
,
a
n
d
J.
D
i
e
sn
e
r
,
“
I
d
e
n
t
i
f
y
i
n
g
t
h
e
O
v
e
r
l
a
p
b
e
t
w
e
e
n
El
e
c
t
i
o
n
R
e
su
l
t
a
n
d
C
a
n
d
i
d
a
t
e
s’
R
a
n
k
i
n
g
B
a
se
d
o
n
H
a
s
h
t
a
g
-
E
n
h
a
n
c
e
d
,
L
e
x
i
c
o
n
-
B
a
se
d
S
e
n
t
i
me
n
t
A
n
a
l
y
si
s,”
i
n
2
0
1
7
I
E
EE
1
1
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
S
e
m
a
n
t
i
c
C
o
m
p
u
t
i
n
g
(
I
C
S
C
)
,
I
EEE,
2
0
1
7
,
p
p
.
9
3
–
9
6
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
S
C
.
2
0
1
7
.
9
2
.
[
8
]
I
.
S
a
e
n
k
o
a
n
d
I
.
K
o
t
e
n
k
o
,
“
T
o
w
a
r
d
s R
e
si
l
i
e
n
t
a
n
d
Ef
f
i
c
i
e
n
t
B
i
g
D
a
t
a
S
t
o
r
a
g
e
:
Ev
a
l
u
a
t
i
n
g
a
S
I
EM
R
e
p
o
si
t
o
r
y
B
a
se
d
o
n
H
D
F
S
,
”
i
n
2
0
2
2
3
0
t
h
Eu
r
o
m
i
c
r
o
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
P
a
ra
l
l
e
l
,
D
i
s
t
ri
b
u
t
e
d
a
n
d
N
e
t
w
o
rk
-
b
a
s
e
d
Pro
c
e
ss
i
n
g
(
PD
P)
,
I
EEE,
M
a
r
.
2
0
2
2
,
p
p
.
2
9
0
–
2
9
7
,
d
o
i
:
1
0
.
1
1
0
9
/
P
D
P
5
5
9
0
4
.
2
0
2
2
.
0
0
0
5
1
.
[
9
]
P
.
S
h
u
e
t
a
l
.
,
“
e
T
i
me
:
En
e
r
g
y
-
e
f
f
i
c
i
e
n
t
t
r
a
n
smiss
i
o
n
b
e
t
w
e
e
n
c
l
o
u
d
a
n
d
m
o
b
i
l
e
d
e
v
i
c
e
s,”
i
n
2
0
1
3
Pro
c
e
e
d
i
n
g
s
I
EEE
I
N
FO
C
O
M
,
I
EEE,
A
p
r
.
2
0
1
3
,
p
p
.
1
9
5
–
1
9
9
,
d
o
i
:
1
0
.
1
1
0
9
/
I
N
F
C
O
M
.
2
0
1
3
.
6
5
6
6
7
6
2
.
[
1
0
]
P
.
S
i
n
g
h
,
Y
.
K
.
D
w
i
v
e
d
i
,
K
.
S
.
K
a
h
l
o
n
,
A
.
P
a
t
h
a
n
i
a
,
a
n
d
R
.
S
.
S
a
w
h
n
e
y
,
“
C
a
n
t
w
i
t
t
e
r
a
n
a
l
y
t
i
c
s
p
r
e
d
i
c
t
e
l
e
c
t
i
o
n
o
u
t
c
o
me
?
A
n
i
n
s
i
g
h
t
f
r
o
m
2
0
1
7
P
u
n
j
a
b
a
sse
mb
l
y
e
l
e
c
t
i
o
n
s,
”
G
o
v
e
r
n
m
e
n
t
I
n
f
o
rm
a
t
i
o
n
Q
u
a
r
t
e
r
l
y
,
v
o
l
.
3
7
,
n
o
.
2
,
p
.
1
0
1
4
4
4
,
A
p
r
.
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
g
i
q
.
2
0
1
9
.
1
0
1
4
4
4
.
[
1
1
]
R
.
K
.
S
i
n
g
h
a
n
d
H
.
K
.
V
e
r
ma,
“
Ef
f
e
c
t
i
v
e
P
a
r
a
l
l
e
l
P
r
o
c
e
ssi
n
g
S
o
c
i
a
l
M
e
d
i
a
A
n
a
l
y
t
i
c
s
F
r
a
mew
o
r
k
,
”
J
o
u
r
n
a
l
o
f
K
i
n
g
S
a
u
d
U
n
i
v
e
rsi
t
y
-
C
o
m
p
u
t
e
r
a
n
d
I
n
f
o
rm
a
t
i
o
n
S
c
i
e
n
c
e
s
,
v
o
l
.
3
4
,
n
o
.
6
,
p
p
.
2
8
6
0
–
2
8
7
0
,
Ju
n
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
k
s
u
c
i
.
2
0
2
0
.
0
4
.
0
1
9
.
[
1
2
]
C
.
T
r
o
u
ssas,
A
.
K
r
o
u
sk
a
,
a
n
d
M
.
V
i
r
v
o
u
,
“
Ev
a
l
u
a
t
i
o
n
o
f
e
n
se
mb
l
e
-
b
a
se
d
se
n
t
i
me
n
t
c
l
a
ss
i
f
i
e
r
s
f
o
r
Tw
i
t
t
e
r
d
a
t
a
,
”
i
n
2
0
1
6
7
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
I
n
f
o
r
m
a
t
i
o
n
,
I
n
t
e
l
l
i
g
e
n
c
e
,
S
y
st
e
m
s
&
Ap
p
l
i
c
a
t
i
o
n
s
(
I
I
S
A)
,
I
EE
E,
Ju
l
.
2
0
1
6
,
p
p
.
1
–
6
,
d
o
i
:
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4864
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
,
Vo
l.
14
,
No
.
2
,
J
u
l
y
20
25
:
472
-
4
8
0
480
1
0
.
1
1
0
9
/
I
I
S
A
.
2
0
1
6
.
7
7
8
5
3
8
0
.
[
1
3
]
R
.
U
l
M
u
s
t
a
f
a
,
M
.
S
.
N
a
w
a
z
,
M
.
I
.
U
.
L
a
l
i
,
T
.
Z
i
a
,
a
n
d
W
.
M
e
h
mo
o
d
,
“
P
r
e
d
i
c
t
i
n
g
T
h
e
C
r
i
c
k
e
t
M
a
t
c
h
O
u
t
c
o
me
U
si
n
g
C
r
o
w
d
O
p
i
n
i
o
n
s
O
n
S
o
c
i
a
l
N
e
t
w
o
r
k
s:
A
C
o
mp
a
r
a
t
i
v
e
S
t
u
d
y
O
f
M
a
c
h
i
n
e
L
e
a
r
n
i
n
g
M
e
t
h
o
d
s
,
”
Ma
l
a
y
si
a
n
J
o
u
r
n
a
l
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
3
0
,
n
o
.
1
,
p
p
.
6
3
–
7
6
,
M
a
r
.
2
0
1
7
,
d
o
i
:
1
0
.
2
2
4
5
2
/
m
j
c
s.v
o
l
3
0
n
o
1
.
5
.
[
1
4
]
V
.
V
a
n
d
a
n
a
K
o
l
i
se
t
t
y
a
n
d
D
.
S
.
R
a
j
p
u
t
,
“
I
n
t
e
g
r
a
t
i
o
n
a
n
d
c
l
a
ss
i
f
i
c
a
t
i
o
n
a
p
p
r
o
a
c
h
b
a
se
d
o
n
p
r
o
b
a
b
i
l
i
s
t
i
c
se
ma
n
t
i
c
a
sso
c
i
a
t
i
o
n
f
o
r
b
i
g
d
a
t
a
,
”
C
o
m
p
l
e
x
&
I
n
t
e
l
l
i
g
e
n
t
S
y
s
t
e
m
s
,
v
o
l
.
9
,
n
o
.
4
,
p
p
.
3
6
8
1
–
3
6
9
4
,
A
u
g
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
7
/
s
4
0
7
4
7
-
0
2
1
-
0
0
5
4
8
-
x.
[
1
5
]
G
.
V
i
sw
a
n
a
t
h
a
n
d
P
.
V
.
K
r
i
sh
n
a
,
“
H
y
b
r
i
d
e
n
c
r
y
p
t
i
o
n
f
r
a
me
w
o
r
k
f
o
r
sec
u
r
i
n
g
b
i
g
d
a
t
a
st
o
r
a
g
e
i
n
m
u
l
t
i
-
c
l
o
u
d
e
n
v
i
r
o
n
me
n
t
,
”
Ev
o
l
u
t
i
o
n
a
r
y
I
n
t
e
l
l
i
g
e
n
c
e
,
v
o
l
.
1
4
,
n
o
.
2
,
p
p
.
6
9
1
–
6
9
8
,
J
u
n
.
2
0
2
1
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
2
0
6
5
-
0
2
0
-
0
0
4
0
4
-
w.
[
1
6
]
H
.
Y
u
,
Y
.
H
u
,
a
n
d
P
.
S
h
i
,
“
A
P
r
e
d
i
c
t
i
o
n
M
e
t
h
o
d
o
f
P
e
a
k
T
i
me
P
o
p
u
l
a
r
i
t
y
B
a
se
d
o
n
T
w
i
t
t
e
r
H
a
sh
t
a
g
s,
”
I
E
EE
A
c
c
e
ss
,
v
o
l
.
8
,
p
p
.
6
1
4
5
3
–
6
1
4
6
1
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
0
.
2
9
8
3
5
8
3
.
[
1
7
]
S
.
Z
h
a
n
g
,
L
.
Z
h
a
o
,
Y
.
L
u
,
a
n
d
J
.
Y
a
n
g
,
“
D
o
y
o
u
g
e
t
t
i
r
e
d
o
f
so
c
i
a
l
i
z
i
n
g
?
A
n
e
mp
i
r
i
c
a
l
e
x
p
l
a
n
a
t
i
o
n
o
f
d
i
sc
o
n
t
i
n
u
o
u
s
u
s
a
g
e
b
e
h
a
v
i
o
u
r
i
n
so
c
i
a
l
n
e
t
w
o
r
k
se
r
v
i
c
e
s,”
I
n
f
o
rm
a
t
i
o
n
&
Ma
n
a
g
e
m
e
n
t
,
v
o
l
.
5
3
,
n
o
.
7
,
p
p
.
9
0
4
–
9
1
4
,
N
o
v
.
2
0
1
6
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
m
.
2
0
1
6
.
0
3
.
0
0
6
.
[
1
8
]
K
.
M
u
s
i
a
ł
,
P
.
K
a
z
i
e
n
k
o
,
a
n
d
P
.
B
r
ó
d
k
a
,
“
U
se
r
p
o
si
t
i
o
n
me
a
su
r
e
s
i
n
so
c
i
a
l
n
e
t
w
o
r
k
s,”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
3
rd
Wo
r
k
s
h
o
p
o
n
S
o
c
i
a
l
N
e
t
w
o
r
k
Mi
n
i
n
g
a
n
d
A
n
a
l
y
si
s
,
N
e
w
Y
o
r
k
,
N
Y
,
U
S
A
:
A
C
M
,
Ju
n
.
2
0
0
9
,
p
p
.
1
–
9
,
d
o
i
:
1
0
.
1
1
4
5
/
1
7
3
1
0
1
1
.
1
7
3
1
0
1
7
.
[
1
9
]
G
.
P
e
t
z
,
M
.
K
a
r
p
o
w
i
c
z
,
H
.
F
ü
r
sc
h
u
ß
,
A
.
A
u
i
n
g
e
r
,
V
.
S
t
ř
í
t
e
sk
ý
,
a
n
d
A
.
H
o
l
z
i
n
g
e
r
,
“
R
e
p
r
i
n
t
o
f
:
C
o
m
p
u
t
a
t
i
o
n
a
l
a
p
p
r
o
a
c
h
e
s
f
o
r
mi
n
i
n
g
u
se
r
’
s
o
p
i
n
i
o
n
s
o
n
t
h
e
W
e
b
2
.
0
,
”
I
n
f
o
rm
a
t
i
o
n
P
ro
c
e
ss
i
n
g
&
Ma
n
a
g
e
m
e
n
t
,
v
o
l
.
5
1
,
n
o
.
4
,
p
p
.
5
1
0
–
5
1
9
,
J
u
l
.
2
0
1
5
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
p
m.
2
0
1
4
.
0
7
.
0
1
1
.
[
2
0
]
M
.
R
i
c
h
a
r
d
so
n
a
n
d
P
.
D
o
mi
n
g
o
s,
“
M
i
n
i
n
g
k
n
o
w
l
e
d
g
e
-
sh
a
r
i
n
g
si
t
e
s
f
o
r
v
i
r
a
l
mark
e
t
i
n
g
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
T
h
e
E
i
g
h
t
h
A
C
M
S
I
G
K
D
D
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
K
n
o
w
l
e
d
g
e
D
i
sc
o
v
e
ry
and
D
a
t
a
Mi
n
i
n
g
,
N
e
w
Y
o
r
k
,
N
Y
,
U
S
A
:
A
C
M
,
J
u
l
.
2
0
0
2
,
p
p
.
6
1
–
7
0
,
d
o
i
:
1
0
.
1
1
4
5
/
7
7
5
0
4
7
.
7
7
5
0
5
7
.
[
2
1
]
R
.
G
h
o
sh
a
n
d
K
.
L
e
r
man
,
“
P
r
e
d
i
c
t
i
n
g
i
n
f
l
u
e
n
t
i
a
l
u
se
r
s i
n
o
n
l
i
n
e
so
c
i
a
l
n
e
t
w
o
r
k
s,”
a
rX
i
v
,
2
0
1
0
,
d
o
i
:
1
0
.
4
8
5
5
0
/
a
r
X
i
v
.
1
0
0
5
.
4
8
8
2
.
[
2
2
]
J.
G
o
l
b
e
c
k
a
n
d
J.
H
e
n
d
l
e
r
,
“
I
n
f
e
r
r
i
n
g
b
i
n
a
r
y
t
r
u
st
r
e
l
a
t
i
o
n
s
h
i
p
s
i
n
W
e
b
-
b
a
se
d
so
c
i
a
l
n
e
t
w
o
r
k
s,”
A
C
M
T
r
a
n
sa
c
t
i
o
n
s
o
n
I
n
t
e
r
n
e
t
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
6
,
n
o
.
4
,
p
p
.
4
9
7
–
5
2
9
,
N
o
v
.
2
0
0
6
,
d
o
i
:
1
0
.
1
1
4
5
/
1
1
8
3
4
6
3
.
1
1
8
3
4
7
0
.
[
2
3
]
D
.
G
r
u
h
l
,
R
.
G
u
h
a
,
D
.
L
i
b
e
n
-
N
o
w
e
l
l
,
a
n
d
A
.
T
o
mk
i
n
s,
“
I
n
f
o
r
mat
i
o
n
d
i
f
f
u
si
o
n
t
h
r
o
u
g
h
b
l
o
g
sp
a
c
e
,
”
i
n
P
ro
c
e
e
d
i
n
g
s
o
f
t
h
e
1
3
t
h
i
n
t
e
r
n
a
t
i
o
n
a
l
c
o
n
f
e
r
e
n
c
e
o
n
W
o
r
l
d
Wi
d
e
W
e
b
,
N
e
w
Y
o
r
k
,
N
Y
,
U
S
A
:
A
C
M
,
M
a
y
2
0
0
4
,
p
p
.
4
9
1
–
5
0
1
,
d
o
i
:
1
0
.
1
1
4
5
/
9
8
8
6
7
2
.
9
8
8
7
3
9
.
[
2
4
]
H
.
H
a
n
a
n
d
S
.
T
r
i
mi
,
“
A
f
u
z
z
y
T
O
P
S
I
S
m
e
t
h
o
d
f
o
r
p
e
r
f
o
r
man
c
e
e
v
a
l
u
a
t
i
o
n
o
f
r
e
v
e
r
se
l
o
g
i
st
i
c
s
i
n
so
c
i
a
l
c
o
m
me
r
c
e
p
l
a
t
f
o
r
ms,”
Ex
p
e
rt
S
y
st
e
m
s w
i
t
h
Ap
p
l
i
c
a
t
i
ons
,
v
o
l
.
1
0
3
,
p
p
.
1
3
3
–
1
4
5
,
A
u
g
.
2
0
1
8
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
e
sw
a
.
2
0
1
8
.
0
3
.
0
0
3
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
M
r
.
B
.
K
a
r
th
ic
k
is
a
P
h
.
D.
re
se
a
rc
h
sc
h
o
lar
in
th
e
De
p
a
rt
m
e
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
,
A
lag
a
p
p
a
Un
iv
e
rsity
,
Ka
ra
ik
u
d
i,
T
a
m
il
N
a
d
u
,
I
n
d
ia.
He
is
w
o
rk
in
g
a
s
A
ss
istan
t
p
ro
f
e
ss
o
r
in
th
e
De
p
a
rtm
e
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
,
S
y
e
d
Ha
m
e
e
d
h
a
A
rts
a
n
d
S
c
ien
c
e
Co
ll
e
g
e
,
Kilak
a
ra
i,
Ta
m
il
N
a
d
u
,
In
d
ia.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
b
k
a
rth
ick
1
9
8
0
@g
m
a
il
.
c
o
m
.
Dr
.
T.
M
e
y
y
a
p
p
a
n
is
w
o
rk
in
g
a
s
S
e
n
io
r
P
r
o
f
e
ss
o
r,
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
,
A
l
a
g
a
p
p
a
Un
iv
e
rsit
y
,
K
a
ra
ik
u
d
i,
T
a
m
il
N
a
d
u
,
In
d
ia
.
He
h
a
s
v
a
st
e
x
p
e
rien
c
e
s
in
th
e
tea
c
h
in
g
f
ield
.
He
h
a
s
p
u
b
li
sh
e
d
se
v
e
r
a
l
re
se
a
r
c
h
p
a
p
e
rs
in
v
a
rio
u
s
c
o
n
f
e
re
n
c
e
s
a
n
d
jo
u
r
n
a
ls
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
m
e
y
y
a
p
p
a
n
t@a
lag
a
p
p
a
u
n
iv
e
rsi
y
.
a
c
.
i
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.