I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
p
ute
r
E
ng
in
ee
ring
(
I
J
E
CE
)
Vo
l.
7
,
No
.
6
,
Dec
em
b
er
201
7
,
p
p
.
3
7
0
0
~
3
7
0
4
I
SS
N:
2088
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v
7
i
6
.
pp
3
7
0
0
-
3704
3700
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
A Survey
on
A
uto
m
a
tically
Mining
Facets
for Web
Queries
Duhi
t
a
P
a
w
a
r
,
Vina
M
.
L
o
m
t
e
RM
D S
h
i
n
g
a
d
S
c
h
o
o
l
o
f
En
g
in
e
e
rin
g
,
P
u
n
e
,
I
n
d
ia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
an
4
,
2
0
1
7
R
ev
i
s
ed
J
u
n
2
,
2
0
1
7
A
cc
ep
ted
J
u
n
2
6
,
2
0
1
7
In
t
h
is
p
a
p
e
r,
a
d
e
tailed
su
rv
e
y
o
n
d
if
f
e
re
n
t
fa
c
e
t
m
in
in
g
tec
h
n
i
q
u
e
s,
t
h
e
ir
a
d
v
a
n
tag
e
s
a
n
d
d
isa
d
v
a
n
tag
e
s
is
c
a
rried
o
u
t
.
F
a
c
e
ts
a
re
a
n
y
w
o
rd
o
r
p
h
ra
se
w
h
ich
su
m
m
a
riz
e
a
n
i
m
p
o
rtan
t
a
sp
e
c
t
a
b
o
u
t
th
e
w
e
b
q
u
e
ry
.
R
e
se
a
rc
h
e
r
s
p
ro
p
o
se
d
d
if
fe
re
n
t
e
ff
icie
n
t
tec
h
n
iq
u
e
s
w
h
ich
im
p
ro
v
e
s
th
e
u
se
r’s
w
e
b
q
u
e
r
y
se
a
rc
h
e
x
p
e
rien
c
e
s
m
a
g
n
if
ic
e
n
tl
y
.
Us
e
r
s
a
re
h
a
p
p
y
w
h
e
n
th
e
y
f
in
d
t
h
e
re
lev
a
n
t
in
f
o
r
m
a
ti
o
n
to
t
h
e
ir
q
u
e
r
y
in
th
e
to
p
re
su
lt
s.
T
h
e
o
b
jec
ti
v
e
s
o
f
th
e
ir
re
se
a
rc
h
a
re
:
(
1
)
T
o
p
re
se
n
t
a
u
t
o
m
a
ted
so
lu
t
io
n
t
o
d
e
riv
e
th
e
q
u
e
r
y
f
a
c
e
ts
b
y
a
n
a
ly
z
in
g
th
e
tex
t
q
u
e
ry
;
(
2
)
T
o
c
re
a
te
tax
o
n
o
m
y
o
f
q
u
e
ry
re
f
in
e
m
e
n
t
stra
teg
ies
f
o
r
e
ff
icie
n
t
re
su
lt
s
;
a
n
d
(
3
)
T
o
p
e
rso
n
a
li
z
e
se
a
rc
h
a
c
c
o
rd
in
g
t
o
u
se
r
in
tere
st.
K
ey
w
o
r
d
:
W
e
b
c
ra
w
li
n
g
I
n
d
e
x
in
g
QD
m
in
e
r
Co
p
y
rig
h
t
©
2
0
1
7
I
n
stit
u
te o
f
Ad
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
:
Du
h
ita
P
a
w
ar
,
Dep
ar
te
m
en
t o
f
C
o
m
p
u
ter
E
n
g
in
ee
r
i
n
g
,
Sav
itrib
ai
P
h
u
le
P
u
n
e
U
n
i
v
er
s
it
y
,
Nea
r
J
ag
r
u
t Sch
o
o
l,
Ma
Sh
a
k
t
i N
ag
ar
,
W
ar
u
d
E
m
ail:
d
u
h
ita.
p
a
w
ar
2
4
@
g
m
ai
l.c
o
m
1.
I
NT
RO
D
UCT
I
O
N
T
h
er
e
ar
e
d
if
f
er
en
t
w
a
y
s
to
h
elp
u
s
er
s
to
b
etter
d
escr
ib
e
th
eir
q
u
er
y
in
f
o
r
m
atio
n
n
e
ed
.
Qu
er
y
r
ef
o
r
m
u
latio
n
an
d
q
u
er
y
r
ec
o
m
m
en
d
atio
n
(
o
r
q
u
er
y
s
u
g
g
es
tio
n
)
ar
e
t
w
o
p
o
p
u
lar
w
a
y
s
o
u
t
o
f
t
h
e
m
.
T
h
e
m
ai
n
g
o
al
o
f
f
ac
et
s
m
i
n
in
g
is
d
i
f
f
er
en
t
f
r
o
m
q
u
er
y
r
ec
o
m
m
e
n
d
ati
o
n
a
n
d
r
ef
o
r
m
u
latio
n
.
T
h
e
g
o
al
o
f
th
e
f
ac
e
t
m
i
n
in
g
i
s
to
s
u
m
m
ar
ize
t
h
e
k
n
o
w
led
g
e
a
n
d
in
f
o
r
m
atio
n
co
n
tai
n
ed
in
t
h
e
q
u
er
y
.
A
q
u
er
y
f
ac
et
is
a
co
llect
io
n
o
f
r
elate
d
an
d
i
n
f
o
r
m
ati
v
e
wo
r
d
s
w
h
ich
d
escr
ib
es
i
m
p
o
r
tan
t
a
s
p
ec
ts
o
f
t
h
e
q
u
er
y
.
He
r
e
a
f
ac
et
ite
m
i
s
t
y
p
icall
y
a
w
o
r
d
.
A
W
eb
q
u
er
y
h
as
m
u
ltip
le
f
ac
ets
t
h
at
s
u
m
m
ar
ize
th
e
i
n
f
o
r
m
atio
n
ab
o
u
t
t
h
e
q
u
er
y
f
r
o
m
d
if
f
er
e
n
t
p
er
s
p
ec
ti
v
es.
I
f
th
e
u
s
er
e
n
ter
‘
w
atc
h
es
’
as
q
u
er
y
th
e
n
d
i
f
f
er
en
t
asp
ec
ts
o
f
th
e
q
u
er
y
‘
w
atc
h
es
’
ar
e
d
is
p
la
y
ed
as
f
ac
ets
w
h
ich
i
n
cl
u
d
es
g
e
n
d
er
ca
te
g
o
r
ies,
b
r
an
d
s
,
s
t
y
le
s
,
co
lo
r
s
,
etc.
Face
ts
ar
e
ass
ig
n
ed
w
ei
g
h
t
in
o
r
d
er
t
o
d
is
p
lay
t
h
e
f
ac
et
s
p
r
io
r
it
y
w
is
e.
T
ec
h
n
iq
u
es
u
s
ed
in
v
ar
io
u
s
ap
p
r
o
ac
h
as see
n
in
T
ab
le
1
.
T
ab
le
1.
T
he
tech
n
iq
u
e
s
u
s
ed
in
v
ar
io
u
s
ap
p
r
o
ac
h
N
o
.
P
a
p
e
r
N
a
me
T
e
c
h
n
i
q
u
e
s
1)
Q
u
e
r
y
S
u
b
t
o
p
i
c
M
i
n
i
n
g
b
y
C
o
mb
i
n
i
n
g
M
u
l
t
i
p
l
e
S
e
man
t
i
c
s
1
)
C
l
u
st
e
r
i
n
g
Q
u
e
r
y
R
e
f
o
r
mu
l
a
t
i
o
n
2
)
p
h
r
a
se
e
mb
e
d
d
i
n
g
r
e
p
r
e
se
n
t
a
t
i
o
n
a
n
d
q
u
e
r
y
c
a
t
e
g
o
r
y
d
i
st
r
i
b
u
t
i
o
n
a
l
r
e
p
r
e
se
n
t
a
t
i
o
n
2)
S
e
a
r
c
h
R
e
s
u
l
t
D
i
v
e
r
si
f
i
c
a
t
i
o
n
B
a
se
d
o
n
Q
u
e
r
y
F
a
c
e
t
s
1
)
I
n
t
e
n
t
-
a
w
a
r
e
d
i
v
e
r
si
f
i
c
a
t
i
o
n
a
l
g
o
r
i
t
h
ms
t
h
a
t
s
u
se
r
i
n
t
e
n
t
s
a
s
su
b
t
o
p
i
c
s.
2
)
f
a
c
e
t
e
d
d
i
v
e
r
si
f
i
c
a
t
i
o
n
a
p
p
r
o
a
c
h
e
s
3)
B
e
y
o
n
d
b
a
s
i
c
f
a
c
e
t
e
d
se
a
r
c
h
i
n
g
1
)
M
u
l
t
i
f
a
c
e
t
e
d
se
a
r
c
h
2
)
O
n
l
i
n
e
a
n
a
l
y
t
i
c
a
l
p
r
o
c
e
ssi
n
g
t
o
e
f
f
i
c
i
e
n
t
l
y
a
n
d
i
n
t
u
i
t
i
v
e
l
y
su
p
p
o
r
t
a
n
a
l
y
s
i
s o
f
mu
l
t
i
-
d
i
me
n
s
i
o
n
a
l
d
a
t
a
a
t
m
u
l
t
i
p
l
e
a
g
g
r
e
g
a
t
i
o
n
l
e
v
e
l
s
4)
Ex
t
r
a
c
t
i
n
g
Q
u
e
r
y
F
a
c
e
t
s
f
r
o
m Se
a
r
c
h
R
e
su
l
t
s
U
R
L
e
x
t
r
a
c
t
i
o
n
,
c
o
n
t
e
n
t
e
x
t
r
a
c
t
i
o
n
,
f
a
c
e
t
s c
l
u
s
t
e
r
i
n
g
,
f
a
c
e
t
s r
a
n
k
i
n
g
5)
S
e
a
r
c
h
i
n
g
D
o
c
u
me
n
t
s
1
)
R
e
l
e
v
a
n
c
e
M
o
d
e
l
F
r
a
mew
o
r
k
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2088
-
8708
A
S
u
r
ve
y
o
n
A
u
to
ma
tica
lly
Mi
n
in
g
F
a
ce
ts
fo
r
W
eb
Qu
eries
(
Du
h
ita
P
a
w
a
r
)
3701
B
a
se
d
o
n
R
e
l
e
v
a
n
c
e
a
n
d
Ty
p
e
2
)
Ty
p
e
M
o
d
e
l
F
r
a
mew
o
r
k
6)
F
a
c
e
t
e
d
p
e
d
i
a
:
D
y
n
a
mi
c
G
e
n
e
r
a
t
i
o
n
o
f
Q
u
e
r
y
-
D
e
p
e
n
d
e
n
t
F
a
c
e
t
e
d
I
n
t
e
r
f
a
c
e
s fo
r
W
i
k
i
p
e
d
i
a
f
a
c
e
t
e
d
i
n
t
e
r
f
a
c
e
d
i
s
c
o
v
e
r
y
a
l
g
o
r
i
t
h
ms
t
h
a
t
o
p
t
i
m
i
z
e
t
h
e
r
a
n
k
i
n
g
me
t
r
i
c
7)
Q
u
e
r
y
R
e
c
o
mm
e
n
d
a
t
i
o
n
u
si
n
g
Q
u
e
r
y
L
o
g
s i
n
S
e
a
r
c
h
E
n
g
i
n
e
s
1
)
Q
u
e
r
y
c
l
u
st
e
r
i
n
g
p
r
o
c
e
ss b
y
w
h
i
c
h
g
r
o
u
p
s o
f
q
u
e
r
i
e
s a
r
e
i
d
e
n
t
i
f
i
e
d
w
h
i
c
h
a
r
e
se
man
t
i
c
a
l
l
y
si
mi
l
a
r
2
)
U
se
s t
h
e
c
o
n
t
e
n
t
o
f
h
i
s
t
o
r
i
c
a
l
p
r
e
f
e
r
e
n
c
e
s o
f
u
se
r
s i
n
t
h
e
q
u
e
r
y
l
o
g
s.
8)
T
r
a
n
sl
a
t
i
n
g
Q
u
e
r
i
e
s i
n
t
o
S
n
i
p
p
e
t
s
f
o
r
I
mp
r
o
v
e
d
Q
u
e
r
y
Ex
p
a
n
s
i
o
n
.
1
)
T
r
a
n
sl
a
t
i
o
n
M
o
d
e
l
-
b
a
se
d
o
n
t
h
e
se
q
u
e
n
c
e
o
f
a
l
i
g
n
me
n
t
mo
d
e
l
s w
h
i
c
h
c
o
n
t
a
i
n
n
u
l
l
w
o
r
d
s
2
)
L
a
n
g
u
a
g
e
M
o
d
e
l
-
A
ssi
g
n
s p
r
o
b
a
b
i
l
i
t
y
t
o
st
r
i
n
g
o
f
w
o
r
d
s
9)
O
p
t
i
mal
A
l
g
o
r
i
t
h
ms f
o
r
C
r
a
w
l
i
n
g
a
H
i
d
d
e
n
D
a
t
a
b
a
se
i
n
t
h
e
W
e
b
1
)
M
a
t
c
h
i
n
g
F
u
n
c
t
i
o
n
.
2
)
L
a
b
e
l
M
a
t
c
h
i
n
g
.
10)
D
y
n
a
mi
c
f
a
c
e
t
e
d
se
a
r
c
h
f
o
r
D
i
sco
v
e
r
y
-
d
r
i
v
e
n
a
n
a
l
y
si
s
1
)
C
a
se
b
a
se
d
r
e
a
so
n
i
n
g
2
)
F
e
d
e
r
a
t
e
d
S
e
a
r
c
h
2.
M
E
T
H
O
DS U
SE
D
Fo
llo
w
i
n
g
m
et
h
o
d
s
co
llecti
v
el
y
ca
lled
QD
m
i
n
er
ar
e
u
s
ed
:
1.
UR
L
E
x
tr
ac
tio
n
:
T
h
is
m
e
th
o
d
is
u
s
ed
to
e
x
tr
ac
t
t
h
e
s
ee
d
s
ite
s
f
r
o
m
s
o
u
r
ce
s
li
k
e
Go
o
g
le,
y
ah
o
o
,
B
in
g
,
etc.
W
h
en
t
h
e
en
d
u
s
er
en
ter
s
h
is
q
u
er
y
t
h
e
s
ea
r
ch
e
n
g
in
e
s
o
u
r
c
es
d
is
p
la
y
s
h
u
n
d
r
ed
s
o
f
th
e
li
n
k
w
it
h
r
ef
er
e
n
ce
to
th
e
en
ter
ed
q
u
er
y
.
T
o
p
m
a
tch
i
n
g
U
R
L
’
s
co
n
tain
in
g
t
h
e
q
u
er
y
w
o
r
d
s
in
t
h
e
m
ar
e
ex
tr
ac
ted
b
y
s
m
ar
t
cr
a
w
li
n
g
.
Fo
r
th
at,
r
ev
er
s
e
s
ea
r
ch
in
g
al
g
o
r
ith
m
is
u
s
ed
.
2.
C
o
n
te
n
t
E
x
tr
ac
tio
n
:
T
h
is
m
et
h
o
d
is
u
s
ed
to
ex
tr
ac
t
th
e
w
e
b
co
n
ten
ts
f
r
o
m
ex
tr
ac
ted
UR
L
’
s
.
Do
cu
m
en
t
p
ar
s
in
g
i
s
d
o
n
e
to
ex
tr
ac
t
th
e
co
n
ten
ts
.
I
n
Do
cu
m
e
n
t
p
ar
s
i
n
g
all
t
h
e
w
o
r
d
ele
m
e
n
ts
i
n
HT
ML
tag
s
(
lik
e
s
elec
t,
u
l,
o
l,
T
ab
le)
o
f
th
e
web
p
ag
es
ar
e
ex
tr
ac
ted
.
Fro
m
ea
c
h
d
o
cu
m
e
n
t
w
e
e
x
tr
ac
t
t
h
e
s
et
o
f
co
n
te
n
t
lis
ts
.
3.
Mir
r
o
r
W
e
b
s
ites
R
e
m
o
v
al:
I
n
th
i
s
m
eth
o
d
t
w
o
w
eb
s
ites
w
it
h
t
h
e
d
if
f
er
en
t
U
R
L
’
s
m
a
y
co
n
tai
n
th
e
d
u
p
licated
co
n
te
n
ts
.
I
t
g
en
er
a
tes
d
u
p
licated
ex
tr
ac
ted
co
n
t
en
t
l
is
t.
Fi
n
e
g
r
ain
ed
s
i
m
ilar
i
t
y
i
s
ca
lc
u
lated
b
et
w
ee
n
t
h
e
t
w
o
lis
t
s
b
y
b
ase
d
o
n
Ha
m
m
in
g
Dis
ta
n
ce
b
et
wee
n
t
h
eir
co
n
ten
t
s
.
O
n
e
o
f
t
h
e
d
u
p
licated
lis
t
i
s
th
en
r
e
m
o
v
ed
s
o
th
a
t r
esu
l
ts
ar
e
m
o
r
e
f
in
ed
a
n
d
w
i
th
o
u
t r
ed
u
n
d
an
c
y
.
4.
L
is
t W
eig
h
ti
n
g
: So
m
e
o
f
th
e
e
x
tr
ac
ted
f
ac
et
s
li
s
ts
ca
n
b
e
n
o
is
y
o
r
u
n
i
m
p
o
r
tan
t.
Go
o
d
lis
t
s
m
o
r
e
f
r
eq
u
e
n
tl
y
o
cc
u
r
in
m
an
y
w
eb
s
ite
s
a
n
d
c
o
n
tain
t
h
e
i
n
f
o
r
m
ati
v
e
ite
m
s
.
T
h
er
ef
o
r
e
w
e
ca
lcu
late
w
e
ig
h
t
ag
e
o
f
ea
ch
lis
t
b
ased
o
n
t
w
o
co
m
p
o
n
e
n
ts
1
)
F
r
eq
u
en
c
y
o
f
Occ
u
r
r
en
ce
2
)
I
D
F (
I
n
v
er
s
e
Do
c
u
m
e
n
t Fr
eq
u
e
n
c
y
)
5.
L
is
t Cl
u
s
ter
i
n
g
: I
n
t
h
is
m
et
h
o
d
f
ac
et
lis
t
s
co
n
tai
n
i
n
g
th
e
s
i
m
il
ar
ite
m
s
ar
e
clu
s
ter
ed
to
g
eth
er
.
Fo
r
th
is
,
QT
(
Qu
alit
y
T
h
r
esh
o
ld
)
alg
o
r
ith
m
is
u
s
ed
.
6.
I
te
m
r
an
k
in
g
a
n
d
d
is
p
la
y
: I
n
t
h
is
m
et
h
o
d
ite
m
s
ar
e
r
an
k
ed
ac
co
r
d
in
g
to
th
eir
f
r
eq
u
e
n
c
y
o
f
o
cc
u
r
r
en
ce
.
Fin
all
y
,
h
ig
h
l
y
r
an
k
ed
ite
m
s
a
r
e
d
is
p
lay
ed
b
ef
o
r
e
lo
w
r
a
n
k
i
t
e
m
s
i
n
d
escen
d
i
n
g
o
r
d
er
as “
f
ac
ets”.
3.
F
I
NDIN
G
S
Qu
er
y
f
ac
et
ex
tr
ac
tio
n
is
e
v
al
u
ated
w
it
h
d
if
f
er
en
t p
er
s
p
ec
tiv
e
s
:
1)
Qu
alit
y
o
f
cl
u
s
ter
i
n
g
2)
Face
t r
an
k
in
g
ef
f
ec
t
iv
e
n
es
s
3)
E
f
f
ec
tiv
e
n
es
s
i
n
f
i
n
d
in
g
f
ac
ets
Usi
n
g
Di
f
f
er
en
t
m
etr
ics
t
h
e
a
ll
t
h
e
ab
o
v
e
f
ac
et
e
x
tr
ac
tio
n
p
er
s
p
ec
tiv
es
ar
e
ev
al
u
ated
in
o
r
d
er
to
g
et
g
o
o
d
q
u
alit
y
f
ac
ets.
E
x
i
s
ti
n
g
f
ac
et
m
in
i
n
g
s
y
s
te
m
s
f
o
cu
s
ed
o
n
to
g
en
er
ate
th
e
s
u
m
m
ar
ies
b
y
u
s
i
n
g
s
e
n
te
n
ce
s
ex
tr
ac
ted
f
r
o
m
th
e
d
o
cu
m
en
ts
w
h
ile
QD
m
i
n
er
s
y
s
te
m
g
e
n
er
ates
f
ac
ets
b
ased
o
n
f
r
e
q
u
e
n
t
lis
t
s
.
QD
m
i
n
i
n
g
ap
p
r
o
ac
h
is
d
if
f
er
en
t th
a
n
t
h
e
ex
is
t
in
g
ap
p
r
o
ac
h
in
t
w
o
w
a
y
s
:
1)
Op
en
d
o
m
ain
:
Qu
er
ies ar
e
n
o
t r
elate
d
to
s
p
ec
if
ic
d
o
m
ai
n
.
2)
Qu
er
y
d
ep
en
d
an
t: Fac
ets ar
e
e
x
tr
ac
ted
f
r
o
m
to
p
r
etr
iev
ed
d
o
cu
m
e
n
t
s
f
o
r
ea
ch
q
u
er
y
Ag
ai
n
QD
m
i
n
i
n
g
ap
p
r
o
ac
h
u
s
e
s
th
r
ee
p
atter
n
s
to
ex
tr
ac
t
o
u
t
th
e
f
ac
et
li
s
t
f
r
o
m
t
h
e
w
e
b
p
ag
es.
T
h
e
th
r
ee
tech
n
iq
u
es
ar
e
f
r
ee
tex
t
p
atter
n
,
HT
ML
tag
p
atter
n
,
an
d
r
ep
ea
t
r
eg
io
n
p
atter
n
.
R
esu
lts
s
h
o
w
s
t
h
at
co
m
b
i
n
atio
n
o
f
t
h
ese
t
h
r
ee
p
atter
n
s
g
i
v
es
th
e
b
est p
er
f
o
r
m
a
n
ce
p
r
o
v
i
n
g
t
h
at
QD
m
i
n
i
n
g
ap
p
r
o
ac
h
is
m
o
r
e
ef
f
icie
n
t t
h
a
n
t
h
e
ex
is
t
in
g
f
ac
e
ts
m
i
n
in
g
ap
p
r
o
ac
h
es.
4.
RE
S
E
ARCH
M
E
T
H
O
D
[
1
]
P
ap
er
n
am
e:
Au
to
m
atica
ll
y
M
in
i
n
g
Face
ts
f
o
r
Qu
er
ies
f
r
o
m
th
eir
s
ea
r
ch
r
es
u
lt
s
-
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2088
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
6
,
Dec
em
b
er
201
7
:
3
7
0
0
–
3
7
0
4
3702
T
h
is
S
u
r
v
e
y
p
r
o
p
o
s
ed
th
e
s
y
s
t
e
m
atic
s
o
lu
tio
n
f
o
r
f
ac
et
s
m
i
n
in
g
.
Face
t
s
ar
e
e
x
tr
ac
ted
f
r
o
m
th
e
s
ee
d
s
ites
.
T
h
ese
s
ee
d
s
ites
ar
e
th
e
s
ite
s
w
h
ic
h
w
e
g
et
as
r
es
u
lt
w
h
en
w
e
d
o
w
eb
s
ea
r
ch
f
o
r
o
u
r
q
u
er
ies.
Fro
m
t
h
es
e
to
p
s
ee
d
s
ites
f
ac
e
ts
ar
e
ex
t
r
ac
ted
b
y
d
o
cu
m
en
t
p
ar
s
in
g
,
w
ei
g
h
ti
n
g
,
cl
u
s
ter
in
g
an
d
r
an
k
i
n
g
o
f
t
h
e
ex
tr
ac
ted
f
ac
et
s
.
[
2
]
P
ap
er
n
am
e:
Q
u
er
y
Su
b
to
p
ic
Min
i
n
g
b
y
C
o
m
b
i
n
i
n
g
Mu
lt
ip
le
Se
m
a
n
tics
-
T
h
e
f
r
a
m
e
w
o
r
k
o
f
t
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
is
d
iv
id
ed
in
to
t
h
r
ee
p
ar
ts
,
A
s
p
ec
t
P
h
r
ase
E
x
tr
a
ctio
n
,
Se
m
a
n
ti
c
R
ep
r
esen
tat
io
n
s
a
n
d
C
l
u
s
ter
i
n
g
&
S
u
b
to
p
ic
Min
i
n
g
.
I
n
th
e
f
ir
s
t
p
ar
t,
th
e
r
elate
d
q
u
er
i
es
o
f
th
e
to
p
ic
(
o
r
ig
in
al
q
u
er
y
)
ar
e
e
x
tr
ac
ted
f
r
o
m
th
e
q
u
er
y
lo
g
a
n
d
d
en
o
te
th
e
q
u
er
y
w
i
th
m
u
lt
i
-
w
o
r
d
p
h
r
ase.
T
h
en
,
n
o
v
el
s
e
m
an
tic
r
ep
r
esen
tatio
n
s
an
d
co
m
b
in
at
io
n
s
ar
e
u
s
e
d
to
r
ep
r
esen
t
t
h
e
q
u
er
y
asp
ec
t
p
h
r
ases
f
o
r
d
is
tin
g
u
is
h
i
n
g
t
h
e
s
e
m
a
n
tic
s
o
f
w
o
r
d
s
,
s
u
c
h
as,
th
e
s
y
n
o
n
y
m
o
u
s
w
it
h
s
p
ec
ial
-
s
h
ap
es
o
r
w
o
r
d
s
w
it
h
d
if
f
er
e
n
t
m
ea
n
i
n
g
s
.
Fi
n
all
y
,
th
e
y
ad
o
p
t
th
e
clu
s
ter
in
g
ap
p
r
o
a
ch
to
g
en
er
ate
t
h
e
s
u
b
to
p
ics
a
n
d
ea
ch
clu
s
ter
d
en
o
tes o
n
e
s
u
b
to
p
ic
o
f
th
e
i
n
i
tial q
u
er
y
.
[
3
]
P
ap
er
n
am
e:
Sear
c
h
R
e
s
u
l
t D
i
v
er
s
i
f
icatio
n
B
ased
o
n
Qu
er
y
Face
ts
-
I
n
th
i
s
p
ap
er
r
esear
ch
er
s
p
r
o
p
o
s
e
th
r
ee
f
ac
eted
m
o
d
el
s
wh
ich
d
iv
er
s
i
f
y
s
ea
r
c
h
r
es
u
lts
b
ased
o
n
th
e
f
ac
eted
s
u
b
to
p
ics.
T
h
e
y
a
g
ain
ad
o
p
t th
e
d
iv
er
s
if
ica
tio
n
al
g
o
r
ith
m
w
h
ich
i
m
p
r
o
v
e
t
h
e
r
esu
l
t
d
iv
er
s
it
y
.
[
4
]
P
ap
er
n
am
e:
B
e
y
o
n
d
b
asic f
ac
eted
s
ea
r
ch
-
T
h
is
p
ap
er
d
escr
ib
es
tw
o
ex
t
en
s
io
n
s
to
th
e
b
asic
f
ac
eted
s
ea
r
ch
s
y
s
te
m
.
T
h
e
ex
ten
s
io
n
s
ad
d
s
to
th
e
f
ac
eted
ap
p
licatio
n
s
b
y
f
lex
ib
le
an
d
d
y
n
a
m
ic
b
u
s
i
n
e
s
s
d
ata
co
llectio
n
an
d
t
h
is
e
n
ab
le
t
h
e
u
s
er
s
to
g
ai
n
in
s
i
g
h
t
in
to
t
h
eir
d
ata
w
h
ich
i
s
v
er
y
r
ich
q
u
a
l
it
y
o
f
d
ata
b
ec
au
s
e
it
is
m
u
c
h
m
o
r
e
th
a
n
j
u
s
t
k
n
o
w
led
g
e
o
f
th
e
d
o
cu
m
e
n
t
s
b
elo
n
g
in
g
to
ea
ch
f
ac
et
[
5
]
P
ap
er
n
am
e:
D
y
n
a
m
ic
f
ac
eted
s
ea
r
ch
f
o
r
d
is
co
v
er
y
d
r
iv
en
a
n
al
y
s
i
s
-
T
h
is
p
ap
er
im
p
le
m
e
n
t
s
OL
A
P
s
t
y
le
d
is
co
v
er
y
d
r
iv
e
n
an
al
y
s
is
o
n
b
ig
s
et
o
f
u
n
s
tr
u
ctu
r
ed
an
d
s
tr
u
ct
u
r
ed
d
ata.
R
esear
ch
er
s
ag
ai
n
p
r
o
p
o
s
e
a
n
e
w
tech
n
iq
u
e
to
m
ea
s
u
r
e
th
e
in
ter
es
tin
g
n
es
s
an
d
n
o
v
el
n
av
ig
atio
n
m
et
h
o
d
to
s
et
t
h
e
u
s
er
s
e
x
p
ec
tatio
n
.
Ag
ai
n
it
tak
e
s
f
ee
d
b
ac
k
f
r
o
m
u
s
er
a
n
d
f
ee
d
b
ac
k
s
u
r
v
e
y
r
e
s
u
l
ts
v
alid
ate
t
h
at
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
m
ee
ts
e
x
p
ec
tatio
n
s
an
d
is
p
r
o
m
i
s
in
g
.
T
h
e
y
b
u
ild
th
e
ef
f
icien
t
r
u
n
ti
m
e
e
n
g
i
n
e
o
n
t
h
e
to
p
o
f
th
e
i
n
v
er
ted
i
n
d
ex
b
y
e
x
p
lo
itin
g
co
d
es a
n
d
b
it set tr
ee
.
[
6
]
P
ap
er
n
am
e:
E
x
tr
ac
tin
g
Q
u
er
y
Face
ts
f
r
o
m
Sear
ch
R
es
u
lts
-
T
h
is
s
u
r
v
e
y
p
r
o
p
o
s
ed
n
e
w
ev
alu
atio
n
m
etr
ic
f
o
r
t
h
is
t
as
k
t
o
co
m
b
in
e
r
ec
all
a
n
d
p
r
ec
is
io
n
o
f
f
ac
et
ter
m
s
w
it
h
g
r
o
u
p
i
n
g
q
u
a
lit
y
.
T
o
lear
n
q
u
er
y
f
ac
ets
e
x
p
er
i
m
e
n
ta
l
r
esu
lts
s
h
o
w
s
th
a
t
t
h
e
s
u
p
e
r
v
is
ed
m
et
h
o
d
class
i
f
ies o
t
h
er
u
n
s
u
p
er
v
i
s
ed
m
et
h
o
d
s
.
[
7
]
P
ap
er
n
am
e:
Op
ti
m
al
A
l
g
o
r
it
h
m
s
f
o
r
C
r
a
w
l
in
g
a
Hid
d
en
D
atab
ase
in
t
h
e
W
eb
-
T
h
is
p
ap
er
r
eso
lv
es
th
e
p
r
o
b
le
m
o
f
r
ele
v
a
n
t
s
ea
r
c
h
f
o
r
t
h
e
u
s
er
i
n
o
r
d
er
to
m
in
e
o
u
t
all
t
h
e
t
u
p
les
f
r
o
m
d
atab
ase
b
y
i
m
p
le
m
en
t
in
g
s
o
m
e
e
f
f
icie
n
t
al
g
o
r
ith
m
s
w
h
ic
h
d
o
es
t
h
e
tas
k
to
f
i
n
d
t
h
e
r
elev
an
t
s
ea
r
c
h
r
esu
lt
s
ev
e
n
in
t
h
e
w
o
r
s
t
s
c
en
ar
io
b
y
t
a
k
i
n
g
o
n
l
y
s
m
all
n
u
m
b
er
q
u
er
ie
s
as
in
p
u
t
.
R
esear
ch
er
s
also
p
r
o
p
o
s
e
th
e
th
eo
r
etica
l r
es
u
lts
w
h
ic
h
in
d
icate
s
th
a
t th
e
s
e
alg
o
r
ith
m
s
ar
e
o
p
ti
m
al.
[
8
]
P
ap
er
n
am
e:
A
T
w
o
-
s
ta
g
e
C
r
a
w
ler
f
o
r
E
f
f
ic
ien
t
l
y
Har
v
esti
n
g
Dee
p
-
W
eb
I
n
ter
f
ac
e
s
-
T
h
is
p
ap
er
p
r
o
p
o
s
ed
th
e
m
ec
h
a
n
i
s
m
s
i
n
t
w
o
s
ta
g
e
cr
a
w
le
r
ca
lled
s
m
ar
t
cr
a
w
ler
is
u
s
e
d
f
o
r
ef
f
icien
t
m
i
n
in
g
o
f
th
e
d
ee
p
w
eb
p
ag
e
s
.
T
h
er
e
a
r
e
tw
o
s
tep
s
in
v
o
l
v
ed
to
m
in
e
d
ee
p
w
eb
p
ag
es.
I
n
f
ir
s
t
s
tep
th
is
s
m
ar
t
cr
a
w
ler
d
o
es t
h
e
s
ite
b
as
ed
s
ea
r
ch
i
n
g
f
o
r
t
h
e
ce
n
tr
e
p
a
g
es
w
it
h
t
h
e
h
elp
o
f
an
y
s
ea
r
c
h
e
n
g
in
e
w
h
ich
is
ab
le
to
av
o
id
t
h
e
v
is
iti
n
g
o
f
lar
g
e
n
u
m
b
er
o
f
w
eb
p
ag
es
.
I
n
o
r
d
er
to
ac
h
iev
e
t
h
e
m
o
r
e
ac
cu
r
ate
an
d
co
r
r
ec
t
r
esu
lts
f
o
r
f
o
cu
s
ed
cr
a
w
ler
th
i
s
s
m
ar
t
cr
a
w
ler
g
i
v
es
r
an
k
i
n
g
ac
co
r
d
in
g
to
th
e
p
r
io
r
ity
.
Hig
h
p
r
io
r
ity
s
ites
ar
e
r
an
k
ed
to
p
.
I
n
th
e
s
ec
o
n
d
s
ta
g
e
s
m
ar
t
cr
a
w
ler
d
o
es
f
ast
s
ea
r
c
h
i
n
g
b
y
e
x
tr
ac
ti
n
g
m
o
s
t
r
elev
an
t l
in
k
s
.
R
esear
ch
er
s
d
esig
n
t
h
e
lin
k
tr
ee
d
ata
s
tr
u
ct
u
r
e
in
o
r
d
er
to
ac
h
iev
e
th
e
b
r
o
ad
er
co
v
er
ag
e
[
9
]
P
ap
er
n
am
e:
Sear
c
h
in
g
Do
cu
m
en
ts
B
ased
o
n
R
ele
v
a
n
ce
an
d
T
y
p
e
-
T
h
e
p
ap
er
im
p
le
m
e
n
ts
t
h
e
p
r
o
b
lem
i
n
a
g
e
n
er
al
f
r
a
m
e
w
o
r
k
co
n
s
is
t
s
o
f
‘
t
y
p
e
m
o
d
el
’
an
d
‘
r
ele
v
a
n
ce
m
o
d
el
’
.
T
h
e
r
elev
a
n
ce
m
o
d
el
in
d
icate
s
w
h
et
h
er
o
r
n
o
t
a
d
o
cu
m
e
n
t
is
r
ele
v
an
t
to
a
q
u
er
y
.
T
h
e
ty
p
e
m
o
d
el
s
h
o
w
s
th
a
t
w
h
eth
er
th
e
d
o
cu
m
en
t
d
o
es
b
elo
n
g
to
t
h
e
d
es
ig
n
a
ted
d
o
cu
m
e
n
t
t
y
p
e
o
r
n
o
t.
R
esear
c
h
er
s
co
n
s
id
er
th
r
ee
m
et
h
o
d
s
f
o
r
co
m
b
in
g
t
h
e
m
o
d
els:
li
n
ea
r
co
m
b
in
atio
n
o
f
s
co
r
es
th
r
es
h
h
o
ld
i
n
g
o
n
t
h
e
t
y
p
e
s
co
r
e
an
d
h
y
b
r
id
o
f
t
h
e
p
r
ev
i
o
u
s
t
w
o
m
et
h
o
d
s
.
I
t
tak
e
s
co
u
r
s
e
p
ag
e
s
ea
r
ch
an
d
i
n
s
tr
u
ct
io
n
d
o
cu
m
e
n
t
s
ea
r
ch
as e
x
a
m
p
le
s
an
d
th
e
y
h
av
e
co
n
d
u
cted
a
s
er
ies o
f
th
e
e
x
p
er
i
m
e
n
t.
[
1
0
]
P
ap
er
n
am
e:
P
er
s
o
n
aliza
tio
n
o
n
E
-
C
o
n
te
n
t Retr
ie
v
al
B
ased
o
n
Se
m
a
n
tic
W
eb
Ser
v
ice
s
-
T
h
is
m
o
d
el
p
r
o
p
o
s
es
a
ap
p
r
o
a
ch
f
o
r
f
ilter
in
g
t
h
e
ed
u
ca
tio
n
a
l
co
n
ten
t
b
ased
o
n
C
ase
B
ase
d
R
ea
s
o
n
i
n
g
.
I
t
is
b
ased
o
n
th
e
m
o
d
el
A
r
c
h
itect
u
r
e
f
o
r
I
n
telli
g
e
n
t
R
ec
o
v
er
y
o
f
t
h
e
E
d
u
ca
tio
n
al
c
o
n
ten
t
in
th
e
Hete
r
o
g
en
eo
u
s
E
n
v
ir
o
n
m
e
n
t.
Mu
lti
ag
e
n
t
ar
c
h
itect
u
r
e
s
ea
r
ch
an
d
i
n
te
g
r
ate
h
eter
o
g
e
n
e
o
u
s
ed
u
ca
t
io
n
al
co
n
ten
t
t
h
r
o
u
g
h
t
h
e
r
ec
o
v
er
y
m
o
d
el
w
h
ic
h
u
s
e
s
th
e
f
ed
er
ated
s
ea
r
ch
.
T
h
e
tech
n
o
lo
g
i
es
an
d
m
o
d
el
w
h
ic
h
ar
e
p
r
esen
ted
i
n
t
h
i
s
r
esear
c
h
p
ap
er
ex
e
m
p
li
f
y
t
h
e
p
o
te
n
tial
f
o
r
d
ev
e
lo
p
in
g
p
er
s
o
n
a
lized
r
ec
o
v
er
y
s
y
s
te
m
s
f
o
r
d
ig
ita
l
co
n
te
n
t
b
a
s
ed
o
n
t
h
e
p
ar
ad
ig
m
o
f
th
e
v
ir
tu
al
o
r
g
a
n
izatio
n
s
o
f
a
g
en
ts
.
T
h
e
ad
v
an
ta
g
es
o
f
th
e
ar
ch
i
tectu
r
e
p
r
o
p
o
s
ed
in
th
is
p
ap
er
ar
e
its
f
le
x
ib
ilit
y
,
cu
s
to
m
izatio
n
,
an
d
ef
f
icie
n
c
y
.
[
1
1
]
P
ap
er
n
am
e:
Face
ted
p
ed
ia:
D
y
n
a
m
ic
Ge
n
er
atio
n
o
f
Qu
er
y
-
Dep
en
d
en
t Fac
eted
I
n
ter
f
ac
es
f
o
r
W
ik
ip
ed
ia
-
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2088
-
8708
A
S
u
r
ve
y
o
n
A
u
to
ma
tica
lly
Mi
n
in
g
F
a
ce
ts
fo
r
W
eb
Qu
eries
(
Du
h
ita
P
a
w
a
r
)
3703
I
n
th
is
p
ap
er
s
h
ar
in
g
,
p
u
b
lis
h
in
g
,
an
d
co
n
n
ec
tin
g
d
ata
o
n
th
e
W
eb
p
r
o
v
id
es
n
e
w
alter
n
ati
v
e
f
o
r
d
ata
in
te
g
r
atio
n
a
n
d
i
n
ter
o
p
er
ab
ilit
y
.
Ho
w
e
v
er
,
p
r
o
lif
er
atio
n
o
f
d
is
tr
ib
u
ted
a
n
d
i
n
ter
co
n
n
ec
ted
d
ata
s
o
u
r
ce
s
o
n
th
e
W
eb
cr
ea
tes
s
ig
n
i
f
ica
n
t
n
e
w
c
h
alle
n
g
es
f
o
r
co
n
ti
n
u
o
u
s
l
y
m
a
n
ag
in
g
t
h
e
lar
g
e
n
u
m
b
er
o
f
v
ast
d
ata
s
et
s
an
d
th
e
ir
in
ter
d
ep
en
d
en
cies.
I
n
t
h
e
s
e
ar
tic
le
r
esear
ch
er
s
f
o
cu
s
es
o
n
t
h
e
m
ain
p
r
o
b
lem
o
f
p
r
eser
v
i
n
g
ev
o
lv
i
n
g
s
tr
u
ctu
r
ed
i
n
ter
li
n
k
e
d
d
ata.
T
h
e
y
p
r
o
p
o
s
e
th
at
a
n
u
m
b
er
o
f
is
s
u
e
s
w
h
ich
h
i
n
d
er
ap
p
licatio
n
s
an
d
u
s
er
s
ar
e
r
elate
d
to
th
e
te
m
p
o
r
al
asp
ec
t
w
h
ic
h
is
i
n
tr
i
n
s
ic
i
n
L
i
n
k
ed
Data
.
T
h
is
w
o
r
k
p
r
o
p
o
s
e
th
r
ee
u
s
e
ca
s
es
in
o
r
d
er
to
m
o
tiv
ate
t
h
e
ap
p
r
o
ac
h
an
d
also
d
is
cu
s
s
is
s
u
es
t
h
at
o
cc
u
r
an
d
p
r
o
p
o
s
e
w
a
y
to
th
e
s
o
lu
tio
n
[
1
2
]
P
ap
er
n
am
e:
Q
u
er
y
R
ec
o
m
m
e
n
d
atio
n
u
s
i
n
g
Q
u
er
y
L
o
g
s
i
n
S
ea
r
ch
E
n
g
in
e
s
–
T
h
is
s
u
r
v
e
y
p
r
esen
ts
a
n
o
v
e
l
q
u
er
y
p
r
o
ce
s
s
in
g
tec
h
n
iq
u
e
wh
ich
m
ai
n
tai
n
s
h
i
g
h
ac
c
u
r
ac
y
an
d
s
ca
lab
ilit
y
,
an
d
ag
ai
n
it
m
a
n
ag
e
s
to
m
i
n
i
m
ize
t
h
e
laten
c
y
to
g
r
ea
t
e
x
te
n
t
i
n
a
n
s
w
er
in
g
lo
ca
tio
n
b
ase
d
s
p
atial
q
u
er
ies.
P
r
o
p
o
s
ed
ap
p
r
o
ac
h
d
ep
en
d
s
o
n
p
ee
r
-
to
-
p
ee
r
s
h
ar
in
g
,
w
h
ic
h
en
ab
les
to
p
r
o
ce
s
s
q
u
er
ies
w
it
h
o
u
t
d
ela
y
at
a
m
o
b
ile
h
o
s
t b
y
u
s
i
n
g
q
u
er
y
r
e
s
u
lt
s
ca
ch
ed
i
n
its
n
ei
g
h
b
o
r
in
g
m
o
b
ile
p
ee
r
s
.
[
1
3
]
P
ap
e
r
n
a
m
e:
T
r
an
s
latin
g
Qu
er
ies i
n
to
S
n
ip
p
ets f
o
r
I
m
p
r
o
v
ed
Qu
er
y
E
x
p
an
s
io
n
-
P
r
o
p
o
s
ed
w
o
r
k
u
s
e
s
t
h
e
ap
p
r
o
ac
h
o
f
k
e
y
w
o
r
d
m
in
i
n
g
.
I
n
d
ex
in
g
ap
p
r
o
ac
h
is
ap
p
lied
o
v
er
s
ea
r
ch
d
ata.
Sp
atial
in
v
er
ted
in
d
e
x
ex
ten
d
s
th
e
s
ta
n
d
ar
d
in
v
er
ted
in
d
ex
w
h
ic
h
ad
d
r
ess
m
u
lt
id
i
m
e
n
s
io
n
al
i
n
f
o
r
m
atio
n
.
I
t c
o
m
e
s
w
it
h
alg
o
r
it
h
m
s
w
h
ic
h
an
s
w
er
th
e
n
ea
r
est n
e
ig
h
b
o
r
q
u
er
ies
w
it
h
k
e
y
w
o
r
d
s
.
5.
E
XI
ST
I
N
G
SYS
T
E
M
O
VE
RVIE
W
User
s
n
ee
d
to
f
r
eq
u
e
n
tl
y
m
o
d
if
y
t
h
eir
s
ea
r
ch
q
u
er
y
i
n
o
r
d
er
to
g
et
d
esire
d
r
esu
lt
f
o
r
th
eir
w
eb
q
u
er
ies.
T
h
is
s
tr
ate
g
y
o
f
q
u
e
r
y
m
o
d
i
f
icatio
n
is
ca
lled
a
s
q
u
er
y
r
ef
o
r
m
u
la
tio
n
.
Di
f
f
er
en
t
k
i
n
d
s
o
f
e
x
i
s
ti
n
g
s
y
s
te
m
s
h
av
e
p
r
o
p
o
s
ed
d
if
f
er
en
t
ap
p
r
o
ac
h
es
to
g
et
t
h
e
d
esire
s
q
u
er
y
r
e
s
u
l
ts
.
B
u
t
a
u
to
m
atic
f
ac
et
Mi
n
i
n
g
ap
p
r
o
ac
h
is
d
if
f
er
en
t a
n
d
m
o
s
t
ef
f
ec
t
iv
e
ap
p
r
o
ac
h
to
g
et
d
esi
r
es r
esu
lt
s
f
o
r
th
e
u
s
er
s
e
n
ter
e
d
q
u
er
ies.
E
x
is
ti
n
g
s
y
s
te
m
s
u
s
ed
f
o
llo
w
i
n
g
d
i
f
f
er
en
t
k
in
d
s
o
f
s
tr
ate
g
ies
:
1)
C
o
m
p
u
ter
g
en
er
ated
r
ef
o
r
m
u
l
atio
n
s
:
B
y
u
s
in
g
q
u
er
y
lo
g
s
n
e
w
q
u
er
y
r
ef
o
r
m
u
latio
n
w
a
y
s
h
as
b
ee
n
d
is
co
v
er
ed
.
Ag
ai
n
b
y
u
s
i
n
g
click
b
e
h
av
io
u
r
au
to
m
at
icall
y
g
e
n
er
ated
r
ef
o
r
m
u
latio
n
s
w
er
e
d
is
co
v
er
ed
2)
Qu
er
y
s
ess
io
n
b
o
u
n
d
ar
y
d
etec
tio
n
:
Ses
s
io
n
i
s
s
er
ie
s
o
f
i
n
ter
ac
ti
o
n
s
d
o
n
e
b
y
t
h
e
u
s
er
in
o
r
d
er
to
g
et
th
eir
d
esire
d
in
f
o
r
m
atio
n
.
Ses
s
io
n
b
o
u
n
d
ar
y
d
e
tectio
n
is
d
o
n
e
to
d
is
co
v
er
d
if
f
er
e
n
t q
u
er
y
r
e
f
o
r
m
u
latio
n
s
tr
ateg
ie
s
.
3)
C
lic
k
d
ata
an
al
y
s
i
s
:
C
lic
k
d
ata
i
n
d
icate
s
t
h
e
s
ea
r
ch
r
es
u
lt
p
r
ef
er
en
ce
.
So
clic
k
d
ata
a
n
al
y
s
i
s
i
s
d
o
n
e
in
o
r
d
er
to
i
m
p
r
o
v
e
s
ea
r
ch
r
elev
a
n
ce
.
D
is
ad
v
a
n
ta
g
es:
-
1)
Hig
h
co
m
p
u
ta
tio
n
al
ti
m
e.
2)
R
es
u
lts
w
i
th
le
s
s
ac
c
u
r
ac
y
a
n
d
ef
f
ic
ien
c
y
.
6.
AP
P
L
I
CA
T
I
O
N
S
Face
t
m
in
in
g
tec
h
n
iq
u
e
ca
n
b
e
u
s
ed
f
o
r
d
if
f
er
en
t
k
in
d
s
o
f
ap
p
licatio
n
s
.
T
h
is
tec
h
n
iq
u
e
i
s
u
s
ed
f
o
r
h
u
g
e
lib
r
ar
y
d
atab
ase
ap
p
lic
atio
n
s
a
n
d
in
f
o
r
m
a
tio
n
s
cien
ce
r
esear
ch
ap
p
licatio
n
s
a
n
d
to
s
o
m
e
co
m
p
u
ter
s
cien
ce
r
esear
c
h
ap
p
licatio
n
s
an
d
co
m
m
er
cial
s
ea
r
c
h
ap
p
licatio
n
s
E
g
.
Am
az
o
n
.
co
m
n
ee
d
f
ac
ets
m
i
n
i
n
g
ap
p
licatio
n
in
o
r
d
er
to
g
et
r
eq
u
ir
ed
d
ata
in
ef
f
icie
n
t
m
an
n
er
7.
CO
NCLU
SI
O
N
T
h
is
s
u
r
v
e
y
is
p
er
f
o
r
m
ed
w
i
t
h
i
n
ten
t
to
co
llect
v
ar
io
u
s
f
ac
et
m
i
n
i
n
g
tech
n
iq
u
e
s
.
Di
f
f
er
en
t
t
y
p
e
s
o
f
f
ac
et
m
in
in
g
m
ec
h
a
n
i
s
m
ar
e
an
al
y
ze
d
.
A
q
u
er
y
f
ac
e
t
is
s
in
g
le
w
o
r
d
o
r
s
et
o
f
w
o
r
d
s
w
h
ic
h
s
u
m
m
ar
ize
s
i
m
p
o
r
tan
t
i
n
f
o
r
m
atio
n
ab
o
u
t
t
h
e
q
u
er
y
.
Face
t
m
in
i
n
g
m
ec
h
a
n
is
m
p
r
o
v
es
v
er
y
u
s
e
f
u
l
as
i
t
s
av
e
s
th
e
s
ea
r
ch
i
n
g
ti
m
e
o
f
t
h
e
u
s
er
.
I
t
i
m
p
r
o
v
es
t
h
e
s
ea
r
ch
i
n
g
ex
p
er
ien
ce
s
o
f
th
e
u
s
er
aid
in
g
h
i
m
to
h
a
v
e
all
t
h
e
r
elev
an
t
lin
k
s
o
f
th
e
w
eb
s
i
tes
co
n
tai
n
in
g
m
o
s
t
r
elev
an
t
in
f
o
r
m
a
tio
n
f
o
r
h
i
s
e
n
ter
ed
q
u
er
y
o
n
t
h
e
s
a
m
e
p
a
g
e.
T
h
is
f
ac
et
m
i
n
i
n
g
tech
n
iq
u
e
is
m
o
s
tl
y
u
s
e
f
u
l
f
o
r
e
-
co
m
m
er
ce
ap
p
licatio
n
s
,
s
e
a
r
ch
en
g
in
e
s
,
h
u
g
e
r
esear
c
h
li
b
r
ar
y
ap
p
licatio
n
s
,
etc.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2088
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
6
,
Dec
em
b
er
201
7
:
3
7
0
0
–
3
7
0
4
3704
ACK
NO
WL
E
D
G
M
E
NT
I
tak
e
th
i
s
c
h
an
ce
to
e
x
p
r
ess
m
y
ap
p
r
ec
iatio
n
to
m
y
g
u
id
e
an
d
Hea
d
o
f
th
e
Dep
ar
t
m
e
n
t
o
f
C
o
m
p
u
ter
E
n
g
i
n
ee
r
i
n
g
,
R
MD
S
SOE,
P
r
o
f
.
Vin
a
M.
L
o
m
te
f
o
r
h
er
k
in
d
co
o
p
er
atio
n
an
d
g
u
id
an
ce
d
u
r
in
g
t
h
e
en
t
ir
e
r
esear
ch
w
o
r
k
.
I
w
o
u
ld
also
li
k
e
to
th
a
n
k
o
u
r
P
r
in
cip
al
an
d
Ma
n
ag
e
m
e
n
t
f
o
r
p
r
o
v
id
in
g
lab
an
d
o
th
er
f
ac
ilit
ies.
RE
F
E
R
E
NC
E
S
[1
]
Zh
ich
e
n
g
Do
u
,
M
e
m
b
e
r,
IEE
E,
Zh
e
n
g
b
a
o
Jia
n
g
,
S
h
a
Hu
,
Ji
-
R
o
n
g
W
e
n
,
a
n
d
Ru
ih
u
a
S
o
n
g
.
A
u
to
m
a
ti
c
a
ll
y
M
in
in
g
F
a
c
e
ts
f
o
r
Qu
e
ries
f
ro
m
th
e
ir
se
a
rc
h
re
su
lt
s.
I
EE
E
T
ra
n
s
a
c
ti
o
n
s
o
n
Kn
o
wled
g
e
a
n
d
Da
t
a
E
n
g
i
n
e
e
rin
g
,
v
o
l.
2
8
,
n
o
.
2
,
F
e
b
2
0
1
6
.
[2
]
L
izh
e
n
L
iu
,
W
e
n
b
in
X
u
,
W
e
i
S
o
n
g
,
Ha
n
sh
i
W
a
n
g
a
n
d
C
h
a
o
D
u
.
Qu
e
ry
S
u
b
to
p
ic
M
i
n
in
g
b
y
Co
m
b
in
i
n
g
M
u
lt
i
p
le
S
e
m
a
n
ti
c
s.
In
ter
n
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
M
u
lt
ime
d
ia
a
n
d
Ub
iq
u
it
o
u
s E
n
g
in
e
e
rin
g
,
V
o
l.
1
0
,
N
o
.
1
2
(
2
0
1
5
).
[3
]
S
h
a
Hu
,
Zh
i
-
C
h
e
n
g
Do
u
,
X
iao
-
Jie
W
a
n
g
.
S
e
a
rc
h
Re
su
lt
Div
e
r
sif
ica
ti
o
n
Ba
se
d
o
n
Q
u
e
ry
F
a
c
e
ts.
J
o
u
rn
a
l
o
f
c
o
mp
u
ter
sc
ien
c
e
a
n
d
tec
h
n
o
l
o
g
y
,
3
0
(
4
):
8
8
8
–
9
0
1
Ju
ly
2
0
1
5
.
[4
]
O.
Be
n
-
Yitzh
a
k
,
N.
G
o
lb
a
n
d
i,
N.
Ha
r’E
l,
R.
L
e
m
p
e
l,
A
.
Ne
u
m
a
n
n
,
S
.
Ofe
k
-
Ko
i
fm
a
n
,
D.S
h
e
in
w
a
ld
,
E.
S
h
e
k
it
a
,
B.
S
z
n
a
jd
e
r,
a
n
d
S
.
Yo
g
e
v
.
Bey
o
n
d
b
a
sic
fa
c
e
ted
se
a
rc
h
in
g
.
P
r
o
c
.
In
t.
Co
n
f
.
W
e
b
S
e
a
r
c
h
Da
ta
M
in
in
g
,
2
0
0
8
,
p
p
.
3
3
–
44.
[5
]
D.
Da
sh
,
J.
Ra
o
,
N.
M
e
g
id
d
o
,
A
.
A
il
a
m
a
k
i,
a
n
d
G
.
L
o
h
m
a
n
,
Dy
n
a
mic
fa
c
e
ted
se
a
rc
h
fo
r
Disc
o
v
e
ry
-
d
riv
e
n
a
n
a
lys
is
in
ACM
.
I
n
t.
C
o
n
f
.
In
f
.
Kn
o
w
l.
M
a
n
a
g
e
,
p
p
.
3
–
1
2
,
2
0
0
8
.
[6
]
W
e
iz
e
Ko
n
g
a
n
d
Ja
m
e
s
A
ll
a
n
E
x
tra
c
ti
n
g
Qu
e
ry
Fa
c
e
ts
fro
m
S
e
a
rc
h
Res
u
lt
s.
Ce
n
ter
f
o
r
In
telli
g
e
n
t
In
f
o
rm
a
ti
o
n
Re
tri
e
v
a
l
S
c
h
o
o
l
o
f
Co
m
p
u
ter S
c
ien
c
e
Un
iv
e
rsit
y
o
f
M
a
ss
a
c
h
u
se
tt
s
Am
h
e
rst
Am
h
e
rst,
M
A
0
1
0
0
3
.
[7
]
Ch
e
n
g
S
h
e
n
g
1
Na
n
Z
h
a
n
g
3
Yu
f
e
i
T
a
o
1
,
2
X
i
n
Ji
n
3
,
Op
ti
m
a
l
A
lg
o
rith
ms
f
o
r
Cra
wl
in
g
a
Hid
d
e
n
Da
ta
b
a
se
i
n
t
h
e
W
e
b
.
Ista
n
b
u
l,
T
u
rk
e
y
.
P
r
o
c
e
e
d
in
g
s o
f
th
e
V
L
DB E
n
d
o
wm
e
n
t,
V
o
l
.
5
,
N
o
.
1
1
[8
]
F
e
n
g
Zh
a
o
,
Ji
n
g
y
u
Zh
o
u
,
Ch
a
n
g
Nie
,
He
q
in
g
Hu
a
n
g
,
Ha
i
Jin
,
S
m
a
rtCra
w
ler:
A
Tw
o
-
sta
g
e
Cr
a
w
ler
f
o
r
Eff
ici
e
n
tl
y
Ha
rv
e
stin
g
De
e
p
-
W
e
b
In
terf
a
c
e
s
.
in
IE
EE
T
r
a
n
sa
c
ti
o
n
s
o
n
S
e
rv
ice
s Co
mp
u
ti
n
g
,
V
o
lu
m
e
:
P
P
Ye
a
r:
2
0
1
5
.
[9
]
Ju
n
Xu
1
,
Yu
n
b
o
Ca
o
1
,
Ha
n
g
L
i1
,
Nic
k
Cr
a
s
w
e
ll
2
,
a
n
d
Ya
lo
u
Hu
a
n
g
3
,
S
e
a
rc
h
i
n
g
Do
c
u
me
n
ts
Ba
se
d
o
n
Rele
v
a
n
c
e
a
n
d
T
y
p
e
,
in
ECIR
2
0
0
7
,
L
NCS
4
4
2
5
,
p
p
.
6
2
9
–
6
3
6
,
2
0
0
7
.
[1
0
]
A
.
B.
G
il
1
,
S
.
R
o
d
ríg
u
e
z
1
,
F
.
d
e
la
P
rieta
1
a
n
d
De
P
a
z
J.F
.
,
Per
so
n
a
li
z
a
ti
o
n
o
n
E
-
Co
n
ten
t
Retr
iev
a
l
Ba
se
d
o
n
S
e
ma
n
ti
c
W
e
b
S
e
rv
ice
s
,
in
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
,
Un
iv
e
rsit
y
o
f
S
a
la
m
a
n
c
a
,
P
laz
a
d
e
la
M
e
rc
e
d
,
S
a
lam
a
n
c
a
3
7
0
0
8
,
S
p
a
i
n
.
[1
1
]
Ch
e
n
g
k
a
i
L
i,
Nin
g
Ya
n
,
S
e
n
ju
ti
B.
Ro
y
,
Lek
h
e
n
d
ro
L
ish
a
m
,
G
a
u
tam
Da
Fa
c
e
ted
p
e
d
ia
:
Dy
n
a
mic
Ge
n
e
ra
ti
o
n
o
f
Qu
e
ry
-
De
p
e
n
d
e
n
t
F
a
c
e
ted
In
ter
f
a
c
e
s
fo
r W
ikip
e
d
ia
,
i
n
W
OD
’1
2
,
M
a
y
2
5
2
0
1
0
,
Na
n
tes
,
F
ra
n
c
e
[1
2
]
Rica
rd
o
Ba
e
z
a
-
Ya
tes
1
,
Ca
rlo
s
H
u
rtad
o
1
,
a
n
d
M
a
rc
e
lo
M
e
n
d
o
z
a
,
Qu
e
ry
Rec
o
mm
e
n
d
a
ti
o
n
u
si
n
g
Qu
e
ry
L
o
g
s
i
n
S
e
a
rc
h
E
n
g
i
n
e
s
,
in
ECIR
2
0
0
7
,
L
NCS
4
4
2
5
,
p
p
.
6
2
9
6
3
6
,
2
0
0
9
.
[1
3
]
S
tef
a
n
Riez
ler
a
n
d
Yi
L
iu
a
n
d
A
le
x
a
n
d
e
r
V
a
ss
e
r
m
a
n
,
T
ra
n
sla
ti
n
g
Qu
e
ries
in
t
o
S
n
ip
p
e
ts
f
o
r
I
m
p
ro
v
e
d
Qu
e
r
y
Ex
p
a
n
sio
n
.
In
In
ter
n
a
ti
o
n
a
l
j
o
u
rn
a
l
o
f
c
o
mp
u
ter
sc
ien
c
e
,
Vo
l.
2
,
Iss
u
e
2
,
p
p
:
(8
2
-
9
9
),
M
o
n
th
:
A
p
ril
-
Ju
n
e
2
0
1
4
.
Evaluation Warning : The document was created with Spire.PDF for Python.