In
te
r
n
ation
a
l Jou
rn
al
o
f In
form
at
i
c
s an
d
C
o
m
m
u
n
icati
o
n
Te
ch
n
o
lo
gy
(
IJ-ICT)
V
o
l.8,
N
o.1,
A
p
r
il
20
19,
pp.
19~2
4
IS
S
N
: 2252-
87
76,
D
O
I
:
10.11
59
1
/ij
ic
t.v8
i
1
.p
p1
9-2
4
19
Jou
rn
a
l
h
o
me
pa
ge
:
ht
tp:
//i
a
e
score
.
com
/
j
o
u
r
na
l
s
/
i
n
d
e
x
.
p
hp/IJ
ICT
Enabli
ng s
ocial web for
IoT inducing ontolog
i
es
from social tagging
M
o
ha
m
m
e
d
A
l
ru
qi
mi
,
N
o
ur
a
Ak
ni
n
Infor
m
a
t
io
n
Te
c
h
no
lo
gy
a
nd
M
o
d
e
li
ng
S
ystem
s
Research
Un
it, Ab
d
e
l
m
al
ek
E
ssaadi
Un
iv
e
r
sit
y
, M
o
rocco
Art
i
cl
e In
fo
ABSTRACT
A
r
tic
le hist
o
r
y
:
R
e
c
e
i
v
e
d
Oct
1
2
,
2
018
Re
vise
d D
e
c 20,
201
8
A
c
c
e
pte
d
J
a
n
5
,
2019
S
e
m
a
ntic
d
o
m
ain
on
to
lo
g
i
e
s
a
re
i
ncreasi
n
g
l
y
s
een
a
s
th
e
key
fo
r
e
n
a
bli
ng
interoperabi
l
ity
across
he
t
e
roge
neous
syst
ems
and
s
e
nsor-bas
ed
applications.
Th
e
o
n
t
o
l
ogies
d
eplo
yed
i
n
t
hese
s
yst
e
m
s
a
nd
a
p
p
l
i
cation
s
a
re
d
ev
elop
ed
b
y
rest
ricted
g
rou
p
s
of
dom
ain
exp
e
rts
and
n
o
t
by
s
em
antic
w
eb
e
xp
e
r
ts
.
L
a
te
l
y
,
f
o
l
k
s
onom
ies
are
i
n
creas
ing
l
y
exp
l
oited
in
d
evel
op
in
g
on
to
lo
gie
s.
T
he
“
c
ol
lecti
v
e
i
n
t
e
lli
g
en
ce”,
w
h
i
ch
e
m
e
rge
f
r
o
m
c
o
l
l
a
bo
rativ
e
tagg
in
g
can
b
e
s
e
e
n
a
s
a
n
a
l
t
e
r
n
a
t
i
v
e
f
o
r
t
h
e
c
u
r
r
e
n
t
e
f
f
o
r
t
a
t
s
e
m
a
n
t
i
c
w
e
b
o
n
t
ol
ogies
.
Ho
wev
e
r,
t
he
u
nco
n
t
r
olled
nat
u
re
o
f
so
cia
l
t
ag
gi
ng
s
y
s
t
e
m
s
l
ead
s
to
m
any
ki
nd
s
o
f
n
o
i
s
y
a
nno
tatio
ns
,
s
u
ch
a
s
m
i
ss
pellin
gs,
im
preci
si
on
a
nd
a
mbiguity.
Th
us
,
t
h
e
con
s
t
r
ucti
on
o
f
f
o
rma
l
ont
ol
ogies
f
ro
m
so
cial
t
aggi
ng
d
ata
rem
a
ins
a
re
al
c
hal
l
en
ge.
M
o
st
o
f
research
es
h
av
e
f
o
cus
e
d
on
h
o
w
t
o
di
s
co
ver
relat
e
dnes
s
b
et
ween
t
ag
s
rat
h
er
t
han
p
r
od
ucin
g
on
tol
ogies
,
m
u
ch
l
e
s
s
d
o
m
a
i
n
ontolog
i
es.
This
p
aper
p
roposed
an
a
l
g
orithm
t
h
at
u
t
ili
ses
ta
gs
i
n
s
o
c
i
a
l
tag
g
i
ng
sy
s
t
e
m
s
to
a
u
t
om
ati
c
all
y
g
en
erate
u
p
-to
-
dat
e
s
p
e
c
i
f
i
c-d
om
a
i
n
on
to
lo
g
i
es
.
T
h
e
evalu
a
ti
on
o
f
the
alg
o
rithm,
u
s
i
n
g
a
d
at
a
s
et
e
x
tra
c
t
e
d
from
Bib
S
onom
y,
d
e
m
o
n
strat
e
d
th
at
t
h
e
a
l
gorithm
co
ul
d
eff
ectively
l
e
arn
a
do
m
a
in
t
erm
i
nolo
g
y
,
a
n
d
i
d
e
ntify
m
o
re
m
eanin
g
f
ul
s
em
antic
i
nfor
mation
f
o
r
th
e
do
m
a
in
t
ermi
no
log
y
.
F
u
rt
herm
ore,
t
h
e
p
ro
posed
a
lg
orit
h
m
i
nt
rod
u
ced
a
sim
p
l
e
a
nd
eff
ectiv
e m
e
th
od f
o
r d
i
s
a
m
b
i
guatin
g tags
.
K
eyw
ord
:
Interne
t
of
th
in
gs
So
ci
al
t
aggi
ng
So
ci
al
we
b
Co
pyri
gh
t © 2
019 In
stit
u
t
e
of Advanced
En
gi
neeri
n
g
an
d
Scien
ce.
All
rights
res
e
rv
ed.
Corres
pon
d
i
n
g
Au
th
or:
Mo
ham
m
e
d
A
lruq
imi,
Infor
m
a
tio
n Te
chn
o
l
o
g
y
and
M
o
de
lin
g S
y
ste
m
s
Re
sear
ch U
n
i
t,
Abde
lm
alek
E
ssa
ad
i Uni
v
ersi
t
y
,
Bloc2
App58, Mixta,
Mar
t
il,
Mo
r
o
cco.
Em
ail:
m.
alruqimi@
u
a
e
.m
a
1.
I
N
TR
OD
U
C
TI
O
N
Se
ma
n
t
i
c
d
o
m
a
i
n
ont
ol
o
g
i
e
s
a
r
e
i
n
c
r
ea
si
ng
l
y
s
een
a
s
a
k
e
y
f
act
o
r
in
a
utom
at
i
on
of
i
n
f
o
r
ma
tion
proce
ssi
ng.
R
e
c
e
n
t
l
y,
s
e
m
a
n
t
i
c
w
eb
t
ec
hn
ol
o
g
ie
s
are
i
n
tegr
a
t
i
n
g
t
o
I
n
t
e
r
n
e
t
o
f
T
h
i
n
g
s
.
T
h
e
s
e
o
n
t
o
l
o
g
i
e
s
p
l
a
y
an
e
sse
n
tia
l
role
f
or
i
n
t
e
g
ra
ti
ng
IoT
da
ta
a
nd
w
e
b
i
n
for
m
a
t
io
n
s
y
s
t
e
m
s.
A
pplyi
n
g
s
u
c
h
o
n
t
o
log
i
es
t
o
Io
T
w
o
u
l
d
be
tter
e
n
ab
le
“
th
in
gs”
t
o
w
ork
in
c
o-oper
a
tio
n
an
d
als
o
w
ou
l
d
e
na
bl
e
a
u
t
onom
o
u
s
i
n
t
era
c
t
i
o
n
b
e
t
w
e
e
n
"th
i
ng
s"
[
1
-6].
H
ow
eve
r
,
on
tol
o
gie
s
d
e
v
e
l
o
p
m
e
nt
b
y
d
o
m
a
in
e
xper
ts
i
s
a
t
i
m
e
-
c
ons
umin
g
a
nd
ex
pen
s
i
v
e
proce
s
s.
M
oreove
r,
t
he
o
n
t
o
l
o
g
i
es
d
e
p
lo
ye
d
i
n
t
he
c
urre
nt
s
e
n
so
r
-
base
d
ap
pl
i
c
a
t
ions
a
re
d
evel
ope
d
b
y
restric
t
e
d
g
r
o
ups
o
f
d
o
m
a
in
e
xper
t
s
a
nd
n
o
t
b
y
sem
a
nt
ic
w
eb
e
x
p
e
r
ts.
In
t
h
i
s
c
o
n
t
e
x
t,
s
oc
i
a
l
ta
g
g
i
n
g
da
ta
con
t
ri
b
u
te
d
b
y
mill
i
o
ns
o
f
onl
i
n
e
users
repre
s
en
t
a
n
e
ss
e
n
tia
l
an
d
co
n
tin
uo
us
s
ourc
e
for
t
h
e
“
c
ol
l
ecti
v
e
in
t
e
l
l
i
g
e
n
ce
”,
w
hic
h
a
re
i
ncr
e
asi
ngl
y
se
e
n
a
s
an
a
l
t
e
r
na
t
i
v
e
to
t
he
c
ur
rent
e
ffor
t
a
t
s
e
m
antic
w
e
b
o
n
t
o
l
og
ie
s
[7
–
9
].
T
h
e
o
n
t
ol
o
g
ie
s
deri
ve
d
fr
om
f
ol
ks
o
n
o
mies
c
a
n
g
i
v
e
a
ma
ch
ine-
pr
oc
essab
l
e
for
m
o
f
th
e
S
o
c
i
a
l
t
a
g
g
i
ng
data
r
e
p
rese
n
t
i
n
g
o
n
l
i
n
e
c
o
mm
unit
i
es’
c
o
lle
c
tive
i
n
te
l
l
i
ge
n
ce
r
a
t
her
tha
n
t
he
p
erc
e
p
tio
n
of
a
l
im
ite
d
gr
ou
p
of
expe
r
t
s.
A
s
such,
the
y
w
o
u
l
d
b
e
a
b
l
e
t
o
c
a
pt
ure
cha
n
ge
s
de
ri
v
e
d
f
r
o
m
a
m
o
r
e
d
i
v
e
r
s
e
u
s
e
r
p
o
p
u
l
a
t
i
o
n
.
Th
e
r
efo
r
e,
t
h
e
y
wou
l
d
b
eco
me
s
e
m
a
n
ti
ca
l
l
y
ri
ch
er
a
nd
t
hu
s
h
a
nd
ier
for
logi
cal
r
e
asoning
tasks
[10
]
.
U
n
for
t
u
n
a
t
e
l
y,
s
ocia
l
ta
g
g
i
n
g
sys
t
em
s
share
the
pro
b
l
e
m
s
i
n
h
e
r
e
nt
t
o
a
l
l
u
n
c
on
t
r
o
l
l
e
d
vo
c
a
bul
a
r
i
e
s,
s
uch
as
am
big
u
it
y,
s
y
n
o
n
y
my,
a
nd
t
h
e
lack
o
f
hiera
r
chy.
T
hu
s,
k
n
o
w
l
e
dge
e
x
t
ra
cti
o
n
fr
om
t
he
s
o
c
ia
l
ta
gg
in
g
data
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN: 2252-
8776
IJ-ICT
Vol
.
8,
No.
1
, A
pril 20
19
:
19
–
2
4
20
rem
a
ins
a
cha
l
l
e
n
g
e
no
t
s
o
l
v
ed
y
e
t
.
In
t
h
i
s
pa
per,
w
e
i
n
trod
uc
e
a
n
a
l
g
o
r
i
t
h
m
f
o
r
i
nd
u
c
i
n
g
d
o
m
a
i
n
ont
ol
ogy
from
s
oc
i
a
l
ta
gg
i
ng
da
ta.
Ex
pe
rime
nta
l
r
esul
ts,
o
n
a
s
na
p
s
ho
t
o
f
d
a
ta
se
t
from
B
i
b
S
o
n
o
m
y,
s
how
e
d
t
h
a
t
t
h
e
in
t
r
od
uc
ed
a
l
gor
ithm
c
o
u
l
d
e
ffe
c
t
i
v
e
l
y
ca
p
t
ur
e
dom
a
i
n-s
p
ec
if
i
c
c
once
p
ts,
an
d
e
n
r
i
ch
t
hese
c
once
p
ts
w
ith
se
ma
n
t
i
c
i
nfo
r
mat
i
on
ext
ra
c
t
ed
f
ro
m W
i
ki
ped
i
a.
2.
SOCIAL TAG
GING S
Y
S
TEMS
S
o
cial
t
ag
g
i
ng
w
e
b
sites
en
a
b
le
u
se
rs
t
o
ass
i
g
n
free
-
chos
e
n
t
ag
s
t
o
ca
te
gor
ize
the
i
r
di
gi
t
a
l
c
onte
n
t
(such
as
w
e
b
s
ites,
p
ic
t
u
re
s,
v
ide
o
s
e
t
c.)
ove
r
the
We
b,
f
or
ming
t
h
e
s
o-ca
ll
e
d
f
o
l
ks
o
nom
ies
[11].
Cur
r
e
n
t
l
y
,
ma
ny
w
e
b-ba
se
d
serv
ice
s
f
oster
the
c
o
nce
p
t
of
t
a
g
g
i
ng.
T
h
e
se
s
y
s
t
em
s
c
a
n
b
e
d
i
ffe
r
en
ti
ated
a
c
c
ord
i
ng
t
o
t
h
e
ki
nd
of
r
e
s
o
u
r
ces
s
up
por
te
d
i
n
.
F
o
r
ins
t
ance
,
D
e
l
i
c
i
ous
f
or
s
ha
ring
boo
k
m
a
r
k
s
,
Fl
i
c
k
e
r
fo
r
p
hot
os,
BibS
o
nom
y
fo
r
pub
l
i
ca
ti
on
s
a
nd
bo
o
k
ma
rks
a
nd
Y
o
uT
u
b
e
f
or
s
hari
n
g
v
i
de
os.
The
bas
i
c
pri
n
c
i
p
l
e
of
t
hes
e
service
s
i
s
s
i
m
p
l
y
t
o
al
l
o
w
regi
stere
d
u
se
rs
g
ener
a
tin
g
t
h
e
con
t
e
n
t
a
n
d
c
l
a
s
s
i
f
y
i
t
i
n
t
h
e
i
r
o
w
n
u
n
i
q
u
e
w
a
y
b
y
a
ssig
n
i
n
g
arbi
t
r
a
r
y
t
a
gs
t
o
th
i
s
c
on
t
e
nt
.
Rese
arc
h
ers
a
t
t
r
ib
ut
e
d
t
h
e
su
cce
ss
of
t
a
ggin
g
to
t
h
e
f
act
t
ha
t
n
o
s
p
e
c
i
f
i
c
p
r
i
o
r
k
n
o
w
l
e
d
g
e
i
s
r
e
q
u
i
r
e
d
t
o
t
a
g
,
a
n
d
t
h
e
i
m
m
e
d
i
a
t
e
bene
f
i
t
o
f
t
a
g
g
i
ng
[
12]
,
[13]
.
F
r
om
a
kn
ow
le
d
g
e
or
g
a
ni
z
a
t
i
o
n
po
i
n
t
of
v
iew
,
f
o
l
kson
om
ies
ha
ve
t
w
o
m
ai
n
ad
va
n
t
a
g
es
:
S
oc
ial
ta
gg
i
ng
syste
m
s
pro
v
i
d
e
a
vas
t
a
m
o
u
n
t
num
b
e
r
of
u
ser
-
ge
n
e
rated
a
n
no
t
a
tions
a
nd
dire
ct
ly
r
e
f
lec
t
u
ser
s
’
voca
bula
r
ie
s
an
d
in
t
e
rest
s;
t
he
y
ar
e
re
lative
l
y
c
h
ea
p
to
d
eve
l
o
p
a
n
d
h
ar
ves
t
a
s
the
y
e
me
rge
from
e
nd
use
r
s’
t
a
ggin
g
[
1
2–1
4].
These
ad
va
nta
g
e
s
h
a
v
e
t
u
rne
d
S
oc
ia
l
tag
g
i
ng
s
yst
e
ms
i
n
t
o
a
n
i
nt
e
r
est
i
ng
d
at
a
sou
r
ce
s
fo
r
Seman
t
i
c
W
e
b
app
l
ica
t
i
o
ns [7
]
, [8],
[14],
[15]
.
3.
REL
A
TE
D
WORK
S
Mu
ch w
ork
ha
s be
en d
one
to in
tr
o
d
u
ce sem
a
nt
ics in fo
l
ks
o
nom
y [1
6-
19]
, and t
o
i
nve
st
i
g
ate me
t
h
o
d
s
of
d
e
p
l
o
y
i
n
g
t
hi
s
se
ma
n
t
ic
s
for
ta
sks
suc
h
a
s
inf
o
r
m
a
tion
retri
e
v
a
l
[
2
0
-
2
2
]
,
rec
o
mm
e
nde
r
sys
t
em
s
[
2
3
-
26]
,
and
o
n
t
o
l
og
ie
s
de
ve
l
o
pme
n
t
[27-
29].
A
s
w
e
ll,
q
u
i
te
a
n
u
m
be
r
of
w
ork
s
h
as
b
ee
n
d
one
t
o
ex
trac
t
str
u
c
t
ure
d
kn
ow
le
d
g
e
a
n
d
deve
lop
o
n
t
olo
g
ie
s
fr
o
m
s
oc
ial
ta
g
g
i
n
g
sys
t
em
s.
Th
e
ea
rly
st
u
d
ie
s
ex
pl
ored
m
eans
of
lever
a
gi
n
g
t
he
c
o-oc
c
u
rre
nce
sta
t
is
t
i
c
s
o
f
t
a
gs
a
nd
t
h
e
tri
p
ar
t
ite
s
truc
tur
e
o
f
f
o
lks
o
n
o
m
ie
s
to
m
ea
sur
e
t
a
g
rela
t
e
dnes
s
(
e.g.
,
[30-3
3
])
.
M
ore
r
ece
n
tly rese
a
rc
hers
(
e
.
g.,
[28],
[34],
[35])
pro
pose
d
to
ma
ke
t
a
g
s
se
m
a
ntic
s
exp
l
ici
t
b
y
gr
ou
n
d
in
g
t
h
em
t
o
cor
r
esp
o
n
d
i
ng
e
n
tr
i
e
s
i
n
o
nli
n
e
k
n
o
wl
e
d
g
e
b
a
s
es,
s
u
ch
a
s
Wo
rd
Ne
t
and
DBp
e
di
a
.
A
l
t
h
o
ugh
t
h
e
se
a
p
p
ro
a
c
h
e
s
a
r
e
mo
re
p
rec
i
si
on
[
36
],
b
ut
a
p
p
roac
hes
he
a
v
i
l
y de
pe
n
d
e
n
t on
W
o
r
dN
et
get
po
or
r
e
cal
l
d
u
e
t
o
t
he
f
ac
t
t
h
a
t
m
any
of
t
he
t
a
g
s
fr
om
f
olk
s
o
nom
i
e
s
do
no
t
ex
is
t
in
W
or
dN
et.
I
n
g
e
n
era
l
,
there
is
a
l
a
c
k
o
f
m
e
t
h
o
d
s
th
at
e
x
t
ra
ct
dom
ai
n-spec
ific
o
n
t
o
l
o
g
ies
from
f
o
lks
o
nomies.
Our
algorithm
pro
duce
s
b
ase
line
dom
ai
n
o
n
t
o
l
og
ie
s
fr
om
t
a
g
s
i
n
f
ol
k
s
on
om
ies.
The
pr
op
ose
d
a
l
g
orithm
c
o
l
l
e
c
t
s
d
o
m
ain-
rele
va
nt
t
e
r
m
s
from
tags
r
e
l
y
i
ng
on
a
se
t
of
doma
i
n
keyw
o
r
d
s
e
x
t
r
ac
ted
a
u
tom
a
t
i
ca
ll
y
fr
om
W
i
k
i
p
e
d
i
a
p
age
s
tit
les.
T
hen,
it i
d
en
t
i
fies t
he
e
x
a
c
t
m
eani
ng o
f
the
t
e
r
m
s
and
retrie
v
e
se
m
a
n
t
i
c
inf
o
rm
ation
abo
u
t
e
ach
t
e
r
m
.
4.
INDUCING
D
OMAIN
ONT
O
LO
GY
Ou
r
al
g
o
r
i
t
h
m
t
ak
es
t
h
e
n
ame
o
f
a
s
p
ecif
i
c
do
mai
n
a
nd
a
p
rep
a
re
d
fo
lk
sono
my
d
at
as
e
t
a
s
i
npu
t
s
a
nd
pro
duce
s
a
c
orre
spon
di
n
g
doma
i
n
ter
m
ino
l
ogy
a
s
o
ut
pu
t.
T
h
i
s
a
l
g
orithm
f
i
r
st
r
epresent
s
folksonomy
re
so
u
r
c
e
s
a
s
an
u
ndi
re
ct
ed
w
e
i
gh
t
e
d
g
r
a
p
h
.
N
e
x
t
,
i
t
co
ll
ec
ts
a
d
o
ma
in
t
er
min
o
l
o
gy
t
h
ro
ug
h
trave
r
s
i
n
g
t
he
resour
ces gra
p
h
rel
yi
n
g
o
n a se
t
of doma
i
n keyw
ords e
xtr
a
c
t
ed
a
u
t
om
ati
c
a
l
l
y
from
t
itle
s of Wik
ipe
d
i
a
e
nt
r
i
e
s
.
Fi
n
a
lly
,
we
e
x
t
ra
ct
s
ema
n
ti
cs
i
nfo
r
mat
i
on
a
b
o
u
t
th
e
col
l
e
c
t
ed
d
oma
i
n
te
rmino
l
og
ies
by
l
i
n
k
i
n
g
the
m
t
o
t
h
eir
appr
opr
i
a
t
e
W
iki
p
e
d
ia
e
n
t
rie
s
.
This
i
nc
lude
s
ide
n
tif
yin
g
t
h
e
i
n
te
nde
d
m
e
a
n
in
g,
a
t
t
r
i
b
u
t
es
a
nd
sy
n
o
n
y
m
s
of
the d
o
m
a
i
n
te
r
mino
l
o
g
y
.
The ge
nera
l
m
e
tho
d
is sh
ow
n i
n
F
ig
ure
1
.
4.1.
Pre
-
p
r
ocessin
g
The
pre-
proce
ssi
ng
a
c
t
i
v
i
t
y
is
a
n
im
p
o
r
t
a
n
t
ta
sk
a
s
it
g
u
ara
n
te
e
s
t
he
qua
li
ty
o
f
t
h
e
data
ove
r
w
h
ic
h
the
proc
ess
is
goi
ng
to
b
e
c
a
rr
ied
ou
t.
T
h
i
s
ac
t
i
vi
t
y
i
nc
l
u
d
e
s
dele
t
i
ng
spe
c
ial
c
h
a
r
ac
ters,
du
plica
t
e
d
t
a
g
s
an
d
prep
osi
t
i
o
n
s.
F
urtherm
o
re
,
w
e
u
sed
a
le
xica
l
vec
t
or
t
o
exc
l
ud
e
n
o
n
-o
bjec
tive
ta
gs
t
ha
t
c
a
u
se
d
n
o
i
s
y
con
n
ec
tio
ns be
t
w
een
t
he
re
s
ource
s on t
h
e
resourc
e
s
g
r
ap
h [
17],
[23].
4.2.
R
esou
r
ces
g
rap
h
ge
n
e
r
ati
o
n
A
folks
o
nom
y
c
a
n
be
s
ee
n
a
s
a
t
u
p
l
e
A
:
=
(
U,
T
,
R),
where
U
,
T
,
a
n
d
R
,
a
r
e
f
i
n
i
t
e
s
e
t
s
,
w
h
o
s
e
elem
en
ts
a
r
e
c
a
l
l
e
d
users,
t
a
g
s
and
re
sour
ce
s,
r
espec
tive
l
y.
F
o
lks
o
nomy
c
a
n
be
r
epre
se
n
t
e
d
a
s
a
n
u
n
d
i
re
cte
d
tri-par
t
i
t
e h
ype
r-gr
a
ph
G
=
(V
, E)
w
he
re V
=
U
∪
T
∪
R,
is
t
he
s
e
t
of
ve
r
tic
es and
E =
{
(
u, t, r)
|
(
u, t
,
r
)
∈
A
}
is
t
he
s
e
t
o
f
e
d
ge
s
;
t
he
t
r
i
-
p
arti
t
e
g
ra
ph
c
a
n
be
f
o
l
ded
i
n
t
o
t
w
o
a
n
d
o
ne-mode
g
ra
phs
[
7]
,
[3
7].
In
t
his
w
o
r
k
,
w
e
a
d
o
p
t
t
h
i
s
d
e
f
i
n
i
t
i
o
n
u
s
i
n
g
t
h
e
o
n-m
ode
l
grap
h
G
’
=
(
V
’,
E
’)
i
n
w
h
i
c
h
V
’
r
e
p
r
e
s
e
n
t
s
t
h
e
s
e
t
o
f
r
e
s
o
u
r
c
e
s
,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
-
I
CT
I
S
S
N
:
2252-
87
76
En
abl
in
g
soc
i
a
l
we
b f
o
r I
o
T
ind
u
c
i
n
g
o
n
t
o
lo
gie
s
f
r
om
so
c
i
al t
a
g
g
i
n
g
(
M
oh
am
m
e
d Al
r
u
q
i
m
i
)
21
a
nd
E’
r
epr
e
se
nt
s
the
set
o
f
w
eig
h
t
e
d
ed
ge
s.
T
w
o
r
esour
ce
s
(
r
i
,
r
j)
w
il
l
be
c
o
n
n
e
c
t
ed
i
f
t
h
e
y
s
ha
re
a
t
lea
s
t
o
n
e
t
a
g
.
I
n
t
h
e
f
o
l
l
o
w
i
n
g
s
e
c
t
i
o
n
,
w
e
d
e
s
c
r
i
b
e
h
o
w
t
o
t
r
a
v
e
r
s
e
t
h
i
s
g
ra
ph
in
o
r
d
e
r
t
o
col
l
ec
t
the
r
e
leva
n
t
dom
ai
n
ter
m
s.
T
o
imp
l
em
en
t
th
is
p
hase,
we
u
se
J
Grap
hT
libra
ry,
w
h
i
c
h
i
s
a
f
r
e
e
J
a
v
a
c
l
a
s
s
l
i
b
r
a
r
y
t
h
a
t
pr
ovi
des
ma
the
m
ati
c
a
l
g
r
a
ph-
t
h
eor
y
o
b
j
e
c
t
s
a
nd
al
g
o
r
i
thms
(
ht
tp
://j
g
r
a
pht
.org
/
)
.
4.
3.
Doma
in Term
in
olo
gy
Colle
cti
o
n
I
n
t
his
act
i
v
i
t
y
,
our
a
lgor
it
h
m
s
ta
r
t
s
b
y
e
x
t
r
a
c
tin
g
a
list
of
D
o
m
a
i
n
K
e
y
w
o
r
d
s
f
r
o
m
t
h
e
t
i
t
l
e
s
o
f
Wi
ki
p
e
di
a
a
r
t
i
c
l
es
a
nd
r
e
d
i
r
ec
t
i
on
p
ag
es
c
o
n
t
a
i
n
ed
i
n
the
mai
n
Wi
k
i
p
e
dia
ca
te
gor
y
cor
r
espo
n
d
ed
t
o
t
h
e
dom
ai
n
a
t
h
a
n
d.
Based
on
t
h
is
D
om
ain
K
e
y
w
or
ds
l
is
t
,
w
e
selec
t
a
s
e
t
o
f
r
e
sour
ce
s
as
s
t
a
r
ting
po
i
n
ts
(
S
P
)
f
o
r
tr
a
v
er
si
n
g
t
he
g
r
a
p
h
.
F
or
a
r
eso
u
r
c
e,
t
o
be
m
ar
ked
as
s
tar
tin
g
p
o
i
n
t
,
at
l
e
a
st
t
w
o
-
t
h
i
r
d
s of t
he
t
ag
s
assi
gne
d
t
o
th
is
r
e
s
o
u
r
ce
sho
u
ld
b
e
f
o
u
n
d
i
n
the
Dom
a
in
K
eyw
o
rds.
Ne
xt
,
ou
r
al
g
o
ri
thm
t
r
aver
se
s
the
g
r
ap
h
G’
m
any
time
s
(
st
a
r
t
i
ng
f
r
o
m
S
P
s
)
look
ing
for
r
e
so
ur
ce
s
t
h
a
t
a
r
e
r
e
l
e
v
a
nt
t
o
t
h
e
doma
i
n
a
t
h
a
nd.
T
he
t
a
g
s
t
h
a
t
a
r
e
ass
o
c
i
at
ed
t
o
th
e
ret
u
rn
ed
r
e
s
o
u
r
c
e
s
wil
l
b
e
c
o
ll
ect
e
d
a
s
do
ma
i
n
te
r
m
in
o
l
o
g
i
e
s
.
I
n
m
ore
deta
i
l
s,
t
hr
o
u
gho
ut
the
tr
a
v
e
r
si
n
g
p
r
o
cess,
w
e
a
pplie
d
a
r
a
nk
in
g
f
u
n
c
t
i
on
o
v
er
e
ach
v
i
s
i
t
e
d
v
e
r
t
e
x
.
T
h
e
r
a
n
k
i
n
g
f
u
n
c
t
i
o
n
r
a
t
e
s
t
h
e
r
e
leva
nc
e
o
f
a
v
er
te
x
t
o
t
he
g
i
v
e
n
d
oma
i
n
base
d
o
n
t
h
e
n
um
ber
a
n
d
w
e
i
gh
t
o
f
t
he
p
a
t
hs
c
om
in
g
f
r
o
m
t
he
di
ff
er
e
n
t
se
e
d
s
t
o
i
t
(
S
e
e
F
i
g
ur
e
(
1
)
,
a
dap
t
e
d
f
r
o
m
[
28]
)
.
Res
o
ur
ce
s
tha
t
h
a
v
e
a
r
a
nk
in
g
va
lue
gr
e
a
t
e
r
t
h
a
n
a
de
f
i
ned
h
t
h
r
e
s
h
o
l
d
ha
ve
b
ee
n
m
a
r
k
e
d
a
s
do
m
a
in-
r
eleva
n
t
r
e
sour
c
e
s
,
a
n
d
h
e
n
c
e
all
t
h
ei
r
as
so
c
i
at
ed
t
a
g
s
h
a
v
e
be
en ga
t
h
e
r
e
d
as d
oma
i
n-
r
e
le
va
n
t
t
er
m
s
.
To
t
r
a
ver
s
e
t
h
e
gr
aph,
we
u
se
t
he
b
readt
h
first
s
earch
(BFS
)
met
h
od;
onc
e
t
h
e
gr
a
p
h
be
i
n
g
tr
a
v
er
se
d
s
t
ar
tin
g
fr
om
a
p
a
r
t
i
c
u
lar
se
e
d
,
t
he
t
r
a
v
e
r
s
i
n
g
pr
oce
ss
s
t
o
p
s
w
h
e
t
her
r
e
ac
h
i
n
g
a
n
o
t
he
r se
ed or
re
achin
g a
ter
m
ina
l
v
er
tex.
=
|
∈
|
,
,
∈
∩
∈
|
,
,
∈
|
|
∈
|
,
,
∈
|
∗
(
1
)
Let
us
c
o
n
si
de
r
i
s
t
h
e
pr
e
v
ious
l
y
v
isite
d
ver
t
e
x
f
r
o
m
whic
h
w
e
r
e
ache
d
,
d
i
s
t
he
d
ista
nc
e
be
t
w
ee
n
t
h
e
c
u
r
r
ent
ver
t
ex
a
n
d
s
e
e
d.
F
i
gur
e
1.
A
r
c
hit
e
ct
ur
e
of
t
he
p
r
o
p
o
sed
al
g
o
r
i
t
h
m
4.
4.
C
on
cep
t
s
Id
e
nt
if
ic
at
i
o
n
B
y
c
on
c
e
p
t
s
i
d
en
t
i
f
i
cati
o
n
,
w
e
me
an
t
o
i
d
ent
i
f
y
fo
r
e
a
c
h
t
erm
t
he
a
p
p
ro
pr
i
a
te
W
i
k
ipe
d
ia
a
rtic
le
t
h
a
t
r
e
pr
e
s
e
n
ts
i
t
s
i
nte
n
de
d
m
e
a
n
ing
s
o
t
hat
w
e
c
an
s
ta
n
d
a
r
diz
e
nam
e
s
of
t
he
t
er
ms
a
nd
e
n
r
i
ch
t
he
m
b
y
a
ddi
n
g
t
h
ei
r
cat
e
g
o
r
i
e
s
a
n
d
t
h
ei
r
po
ssib
l
e
sy
no
ny
ms
a
s
we
ll
.
Se
e
t
h
e
e
xam
p
l
e
d
e
p
i
c
te
d
i
n
F
ig
ur
e
2.
T
h
i
s
ac
tiv
i
t
y
a
l
s
o
inc
l
ude
s
di
sam
b
ig
ua
ti
ng
t
e
r
ms
a
n
d
e
x
t
r
a
c
tin
g
sem
a
n
t
ic
i
n
f
or
m
a
ti
o
n
a
b
o
u
t
t
h
e
m
a
s
w
e
l
l
.
T
h
e
a
d
v
a
n
t
a
g
e
o
f
us
in
g
W
i
k
i
ped
i
a
a
s
a
r
efe
r
e
n
c
e
t
o
m
a
p
t
e
r
m
s
is
t
ha
t
Wik
i
pe
dia
i
s
a
c
o
mm
un
i
t
y-
dr
i
v
e
n
k
now
le
dge
b
a
s
e
,
m
uch
li
ke
f
ol
ks
on
omie
s
ar
e
,
s
o
t
h
at
i
t
r
a
p
i
dl
y
a
d
a
p
t
s
t
o
ac
com
m
oda
t
e
n
e
w
t
ermi
n
o
lo
gy
.
M
a
n
y
o
f
t
h
e
pop
ul
ar
t
ags
oc
cur
r
ing in
f
o
l
k
s
o
n
o
mies
d
o
no
t
ap
pea
r
i
n g
r
amm
a
r
dict
i
o
nar
i
es
,
such
a
s Wor
d
Ne
t,
b
eca
use
the
y
c
orrespo
n
d
to
p
r
o
per
no
u
n
s,
m
oder
n
t
ec
hnic
a
l
w
or
ds,
or
a
r
e
w
ide
l
y
use
d
a
cr
o
nym
s
.
I
n
add
i
tio
n,
t
he
r
ed
ir
ec
t
pa
ge
s
i
n
Wik
i
pe
d
i
a
pr
o
v
ide
s
y
n
o
nyms
a
nd
mor
p
hol
o
g
ica
l
v
a
r
ia
ti
o
n
s
f
or
a
c
o
n
c
e
p
t
.
F
o
r
e
x
am
pl
e
,
w
he
n
sea
r
ch
i
ng
th
e
tag
‘
n
yc’
in
W
iki
p
e
d
ia,
th
e
ent
r
y
for
N
e
w
Y
o
r
k
C
i
t
y
i
s
r
e
t
ur
ned
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N: 2
2
5
2
-
87
76
I
J
-
I
CT
V
ol.
8,
N
o.
1
,
A
p
r
i
l
20
19
:
1
9
–
24
22
F
i
gur
e
2.
A
n
exam
ple
o
f
a
ppl
yi
n
g
C
once
p
ts
I
den
tif
ic
at
i
on
p
r
oc
es
s for
the
te
rm “
C
S
S”
T
o
p
e
r
f
o
r
m
t
h
i
s
t
a
s
k
,
w
e
u
s
e
d
G
o
o
g
l
e
a
s
a
n
i
n
t
e
r
m
e
d
i
a
r
y
t
o
r
e
t
r
i
e
ve
t
h
e
a
ppr
o
p
r
i
a
t
e
c
o
r
r
e
spo
ndi
n
g
Wik
i
pe
d
i
a
a
r
tic
le
f
or
e
a
c
h
ter
m
.
F
i
r
s
t
l
y,
w
e
passe
d
t
o
G
oo
gle
a
t
e
r
m
e
n
cl
osin
g
be
tw
e
e
n
t
h
e
dom
ai
n
na
m
e
(
in
th
is
e
xam
p
l
e
:
“
W
eb
D
ev
e
l
o
p
m
ent”)
a
s
a
c
ont
e
x
t
an
d
t
h
e
word
(
“
W
i
k
i
pe
dia
”
)
t
o
b
r
i
n
g
W
i
k
i
p
e
d
i
a
p
a
g
es
t
o
the
t
o
p
.
T
he
n
,
w
e
l
ook
f
o
r
a
m
o
r
p
hol
ogi
ca
l
ma
tch
i
ng
b
e
t
we
e
n
t
h
e
t
er
m
a
nd
the
ti
tle
s
of
t
h
e
t
o
p
f
our
r
e
t
r
i
e
v
ed
Wik
i
pe
d
i
a
pag
e
s.
T
he
s
im
ple
s
t
case
oc
c
u
r
s
w
hen
a
ter
m
can
b
e
m
a
tche
d
dire
c
t
l
y
t
o
the
fi
r
s
t
Go
og
le
r
e
s
u
lt.
I
n
ot
her
case
s
,
a
t
e
r
m
c
ould
be
m
a
t
c
h
ed
d
ir
ec
t
l
y
t
o
a
p
a
g
e
t
i
t
l
e,
t
o
a
p
a
r
t
o
f
t
h
e
titl
e,
o
r
t
o
o
n
e
o
f
t
h
e
re
direc
t
ed
pa
ge
s.
A
s
w
e
ll
,
ter
m
s cou
l
d
b
e
m
atche
d
t
o
a
bbr
e
v
ia
t
i
o
n
s t
h
at
c
o
me
w
it
h
t
h
e
Wi
k
i
pe
d
i
a e
n
tr
ies’ t
i
t
les e
n
clo
s
e
d
be
t
w
ee
n
par
e
n
t
he
se
s.
I
n
som
e
case
s
,
m
a
tc
hin
g
t
o
Wi
kipe
d
i
a
e
n
t
r
i
es f
a
i
l
s
.
I
n
f
a
c
t,
q
ue
r
y
ing
W
i
k
i
ped
i
a
thr
o
u
g
h
G
o
o
g
l
e
a
l
l
o
w
s
t
ak
i
n
g
ad
va
nt
ag
e
of
t
echn
i
qu
e
s
e
mb
e
d
d
e
d
i
n
G
o
o
g
le,
suc
h
a
s
stem
min
g
a
nd
lem
m
a
t
iz
a
tio
n,
s
o
t
h
a
t
w
e
have
a
h
ig
h
c
h
ance
o
f
fi
ndi
ng
t
h
e
corr
ect
co
rres
p
ond
ing
W
i
ki
p
e
di
a
a
r
tic
l
e
s.
A
s
i
t
shown
i
n
F
i
g
u
r
e
2
,
p
a
s
s
in
g
t
h
e
t
e
r
m
‘
C
S
S’
t
o
G
o
og
le
r
e
s
u
l
te
d
in
r
e
triev
i
n
g
t
he
W
ik
ipe
d
ia
a
r
tic
l
e
e
n
tit
le
d
‘Casca
d
i
n
g
S
t
y
le
S
he
e
ts’
si
n
c
e
CSS
re
prese
n
t
s
a
r
e
d
i
r
ec
t
pag
e
t
o
t
h
is
a
r
ti
c
l
e
i
n
W
i
k
i
p
ed
ia.
I
n
t
he
c
ase
of
d
i
s
am
bigua
t
e
d
ter
m
s,
(
for
i
n
st
a
n
ce
the
t
e
rm
“
Aj
a
x
”
t
h
at
c
o
u
ld
r
ef
er
t
o
a
pr
ogr
am
ming
lan
g
u
a
g
e
or
a
m
yth
o
l
o
g
ic
al
G
r
e
ek
h
e
r
o)
,
the
Wik
i
pe
d
i
a
a
r
t
ic
le
t
hat
re
prese
n
t
s
i
ts
i
n
t
en
de
d
me
an
i
n
g
c
o
mes
fi
r
s
t
i
n
t
h
e
G
o
ogl
e
re
su
lt
s
du
e
to
u
si
ng
t
he
d
o
m
a
in
n
am
e
as
c
ont
e
x
t
in
G
oog
l
e
s
ea
rc
h
b
o
x
.
H
o
w
e
ver
,
w
e
use
in
for
m
a
tio
n a
v
ai
la
b
l
e o
n
t
h
e
r
etur
ne
d Wik
i
pe
d
i
a
ar
t
ic
le
s
t
o
e
nr
ic
h
t
h
e ter
m
s.
These
i
ncl
ude
s
r
e
dir
e
c
t
p
a
g
es
a
s
al
ter
n
a
tive
nam
e
s,
a
nd
Wi
k
i
pe
d
i
a
c
a
t
egor
ies
c
on
t
a
i
n
in
g
tha
t
p
a
g
e
t
h
at
a
r
e
l
ist
e
d
on
t
h
e
bo
t
t
om
o
f
e
a
c
h
a
r
t
ic
l
e
.
5.
DIS
C
US
S
I
O
N
Th
e
l
a
c
k
o
f
eva
l
u
a
ti
on
fra
me
w
o
rk
s
an
d
t
h
e
l
a
c
k
/i
nco
m
pl
ete
of
e
l
ec
tron
ic
r
esource
s
t
h
a
t
c
a
n
b
e
use
d
a
s
a
g
o
l
d
st
an
da
r
d
m
a
k
es
t
h
e
p
r
o
c
e
ss
of
e
vo
lut
i
on
a
te
r
m
ino
l
og
y
d
i
f
f
i
c
u
lt
[1
7],
[38].
Bes
ides,
fo
lks
o
n
o
m
y
tag
s
a
r
e
unc
on
tr
o
lle
d
v
o
c
a
b
u
l
a
r
ies
tha
t
c
on
tai
n
m
an
y
sla
ng
w
o
r
d
s
a
nd
a
bbr
e
v
ia
t
i
o
n
s,
w
hi
le
t
he
e
l
e
c
t
r
o
nic
r
e
sour
ce
s
of
te
n
use
f
o
r
m
a
l
a
nd
c
om
po
und
t
er
m
s
.
H
o
w
e
ver
,
t
he
e
xp
e
r
ime
n
ts
w
er
e
p
e
rform
e
d
o
n
d
ata
s
e
t
,
c
a
pt
ur
ed
f
r
o
m
B
i
bS
o
n
o
m
y
[
3
9]
,
com
pose
d
o
f
2
0
,
0
0
0
r
es
our
c
e
s
a
n
not
a
t
ed
b
y
85
,
0
06
t
a
g
s
(
1
1
,
865
u
n
i
qu
e
t
a
gs).
T
h
r
ee
d
o
m
ai
n
s
o
f
c
o
mpu
t
er
s
c
i
en
ce
h
av
e
b
e
en
s
e
l
ecte
d
r
an
dom
l
y
f
or
t
he
e
x
p
er
im
ent
s
:
S
e
m
a
nt
ic
W
e
b
,
C
o
mpu
t
e
r
N
et
w
o
r
k
s,
a
nd
We
b
D
e
ve
l
opm
en
t.
T
o
e
v
a
l
ua
t
e
t
he
o
b
t
a
i
ne
d
ter
m
ino
l
og
y,
w
e
use
d
m
a
j
or
i
t
y
vo
tin
g
of
f
ive
r
e
sea
r
che
r
s
w
ho
w
e
r
e
a
ske
d
t
o
m
a
ke
j
u
d
g
me
n
t
s
o
f
dom
ai
n
r
e
le
va
ncy
(
how
s
t
r
o
ngl
y
a
ter
m
i
s
r
e
leva
n
t
to
t
he
g
i
v
e
n
d
o
m
a
i
n)
f
or
a
l
l
t
he
o
bt
a
i
ne
d
t
e
r
m
s
by
a
s
soc
i
a
t
i
n
g
a
l
a
b
e
l
“re
l
ev
a
n
t
”
,
“
i
rre
l
e
v
a
nt
”
,
o
r
“u
n
c
ert
a
in
”
wit
h
e
ac
h
te
rm.
Ta
bl
e
1
s
h
ows
re
sults
w
e
o
b
ta
i
n
ed
;
w
h
er
e
t
h
e
“
D
istinc
t
T
e
r
m
s
”
c
o
lum
n
s
hows
a
l
l
ob
tai
n
e
d
t
e
rms
af
t
e
r
remo
v
i
ng
d
upl
i
c
a
t
e
d
i
t
e
ms,
a
n
d
t
h
e
“R
el
e
v
an
t
T
e
rms”
c
ol
um
n
s
h
ow
s
t
h
e
ter
m
s
m
a
r
k
e
d
a
s
dom
ai
n-
r
e
le
va
nt
t
er
ms.
We
c
alcu
la
te
d
t
h
e
p
r
ec
i
s
i
o
n
o
f
t
he
obta
i
n
e
d
re
sult
s
a
s
f
o
llo
ws:
P
r
e
c
i
s
i
on=
(|rel
e
v
a
nt
|)
*1
0
0
/
(
|
d
ist
i
n
c
t
ter
m
s|
),
w
here
d
is
t
i
nc
t
ter
m
s
refer
to
t
h
e
a
l
l
un
i
q
ue
t
e
r
m
s
w
e
obta
i
n
e
d.
F
or
m
a
lly,
di
st
i
n
ct
t
e
rm
s
=
re
l
e
v
a
n
t
∪
i
r
r
eleva
n
t
∪
u
ncer
t
a
in
w
here
“
releva
n
t
”
re
fer
s
t
o
t
h
e
ter
m
s
tha
t
w
ere
m
a
rke
d
as
d
om
ain
re
l
e
v
a
n
t
t
e
r
ms;
“i
rre
l
e
v
an
t
”
r
ef
ers
t
o
t
h
e
t
erms
t
h
a
t
were
m
a
r
k
e
d
a
s
no
t
dom
ai
n
r
e
le
va
nt
t
er
ms
a
nd
“
uncer
t
a
in”
f
o
r
uno
b
v
i
o
us
t
er
m
s
.
Ta
ble
1.
S
ta
t
i
s
t
i
c
s
of
t
he
r
esul
ts
Do
m
a
in
D
isti
nc
t
Te
r
m
s
Re
l
e
v
a
nt Te
r
m
s
P
re
c
i
sion
N
e
t
w
orks
149
6
2
4
1
.
61%
We
b
D
e
v
e
lop
m
e
n
t
165
9
3
5
6
.
36%
S
e
m
a
nti
c
We
b
116
7
7
6
6
.
38%
Tab
l
e
2
s
h
ow
s
num
ber
of
t
e
r
m
s
t
h
a
t
hav
e
b
e
e
n
c
o
r
r
e
la
t
e
d
s
u
cce
s
s
fu
l
l
y
to
W
iki
p
edia
a
rt
i
c
l
e
s.
H
o
w
e
ver
,
w
e
no
t
i
ce
d
em
p
i
r
i
c
a
ll
y
t
h
a
t
m
ap
pi
n
g
t
a
g
s
to
W
ik
ipe
d
ia
e
n
t
r
ies
ca
n
be
u
se
d
as
a
g
o
od
w
a
y
f
o
r
e
x
cl
u
d
in
g
no
n-
ob
j
e
c
t
i
v
e
ta
gs
s
inc
e
,
by
na
tur
e
,
the
non
-
o
bje
c
ti
v
e
ta
gs
d
o
no
t
ha
ve
c
or
r
e
spon
d
i
n
g
a
r
t
ic
l
e
s
i
n
Wik
i
pe
d
i
a.
U
sin
g
G
o
o
g
le
t
o
que
r
y
i
n
g
Wi
ki
pe
dia
i
n
cr
ease
s
t
he
p
r
o
ba
b
i
l
i
t
y
o
f
p
os
it
iv
e
m
a
tchi
n
g
t
er
m
s
t
o
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ-ICT
I
S
S
N
:
2252-
87
76
Ena
b
li
ng
so
ci
al
w
e
b
f
o
r
Io
T
i
n
du
ci
ng
on
tol
o
g
i
e
s
f
r
o
m
so
cial
t
agg
ing
(Mo
h
a
mme
d
Al
ru
qimi
)
23
W
i
ki
p
e
di
a
e
n
t
i
t
i
e
s,
a
s
Goog
le
c
an
r
ec
og
ni
ze
w
o
r
d
s
m
o
r
p
hol
o
g
y
.
N
ev
ert
h
el
e
ss,
s
o
m
e
t
e
rms
c
oul
d
n
o
t
be
corr
elate
d
t
o
a
Wi
k
i
pe
d
i
a a
r
t
i
c
l
e d
u
e
to m
i
ssin
g
c
omple
t
e
d
con
t
e
x
t (e.
g
. “usa
b
i
lit
y”
t
erm
c
a
nn
o
t
be
l
i
n
k
e
d, bu
t
“
w
e
b
u
sab
ili
ty
”
c
a
n
be)
.
I
n
ot
he
r
ca
ses,
t
e
r
ms
c
ann
o
t
be
m
atc
h
e
d
du
e
to
t
h
e
v
ari
a
n
t
s
t
r
u
c
tu
re
s
o
f
c
omp
ound
term
s.
F
or
i
ns
ta
nce,
m
atch
i
n
g
t
h
e
term
“
D
H
ML”
t
o
t
he
a
r
tic
l
e
l
a
b
e
ll
e
d
“
D
y
n
a
mi
c
Ht
ml”
fai
l
s
al
t
h
o
ugh
t
he
y
refer to
t
he sam
e conce
p
t
.
S
ome
te
rms,
s
uc
h
as W3
C
, X
M
L, coul
d
be c
o
n
s
i
dere
d
re
le
va
nt t
o se
ver
a
l
d
o
m
a
in
s.
A
s
w
e
a
ddress
e
d
i
n
e
a
r
ly,
ge
ne
rat
i
n
g
t
he
d
o
m
a
i
n
ke
y
w
o
r
ds
p
la
y
s
a
n
i
m
porta
nt
r
o
l
e
for
o
b
t
a
ini
n
g
a
go
od
rec
a
ll.
I
n
th
is
c
on
te
xt,
re
d
i
r
ect
p
a
g
e
s
i
n
Wi
ki
ped
i
a
ser
v
e
as
g
o
o
d
doma
i
n
k
e
yw
ords,
as
f
o
l
k
s
o
n
o
mies
c
ont
a
i
n
muc
h
ne
o
lo
gi
s
m
s and a
c
ron
y
m
s. H
ow
eve
r
, for ge
n
e
r
at
in
g ano
t
her
d
om
ai
n c
onc
ep
ts b
y
ou
r
a
l
gor
it
hm
, do
m
a
i
n
expe
r
t
s
m
a
y
be
i
nv
ol
ve
d
i
n
s
e
l
e
c
t
i
ng
t
h
e
s
u
i
t
a
b
l
e
d
a
t
a
s
e
t
f
o
r
a
g
iv
e
n
dom
a
i
n.
U
se
rs
u
t
i
liz
e
fo
lk
so
n
o
m
i
es
w
i
t
h
vari
ous
i
n
t
e
n
ti
ons.
F
o
r
ins
t
a
n
ce,
D
elici
ous
i
s
used
f
or
g
e
n
e
r
al
purp
o
se
w
he
rea
s
B
ibS
o
n
o
m
y
p
rima
ri
l
y
s
e
r
ves
ac
adem
ic
a
n
d
s
c
i
e
n
t
i
fi
c
in
t
e
r
e
sts.
C
ompa
red
t
o
g
ene
r
a
l
f
o
l
ks
on
o
my,
aca
dem
i
c
fo
l
k
son
o
my
h
a
s
a
m
ore
com
p
le
x
na
t
u
re
i
n
te
rms
of
s
em
ant
i
cs
a
n
d
spars
ity
o
f
th
e
da
ta
[
4
0
-
4
2
]
.
Ther
efore
,
a
c
a
dem
i
c
fo
l
k
s
o
nomie
s
w
o
u
l
d
be
m
ore
use
f
u
l
f
or
b
u
i
ldi
n
g
o
n
t
o
l
og
ie
s
(pa
r
tic
ular
ly,
on
t
ol
o
g
i
e
s
for
scie
n
tific
dom
ain
s
).
N
ever
t
h
e
l
ess,
gene
ra
l
fo
l
k
so
nom
ie
s
w
o
u
l
d
be
m
or
e
s
u
ita
b
l
e
for
e
x
t
r
act
in
g
c
onc
ep
ts
o
f
gener
a
l
dom
ai
ns
s
uc
h
a
s
M
o
v
ie
s,
Tra
n
sp
ort.
T
his
is
su
e
ma
y
b
e
c
o
n
s
i
dere
d
in
our
f
ut
ure
w
o
rk.
Bes
id
e
s
d
ev
el
opi
ng
a
m
e
t
h
od
t
h
a
t
l
oo
k
s
f
o
r
c
o
rre
s
p
ond
i
ng
e
n
t
r
i
e
s
on
t
he
d
i
f
f
e
ren
t
onl
i
n
e
k
n
o
w
l
e
dge
s
o
u
r
c
e
s
f
or
t
er
ms
t
ha
t
c
a
nn
o
t
b
e
m
a
p
p
ed
t
o
W
i
ki
p
e
di
a
.
Tab
l
e
2.
N
umbe
r of ter
ms
c
orre
late
d suc
cess
f
u
l
ly t
o W
i
k
i
pe
dia
e
nt
ri
e
s
D
o
m
a
in
C
o
m
pute
r
N
e
t
w
o
r
k
s
We
b
D
e
v
e
l
o
pm
e
n
t
Se
m
a
ntic
We
b
Do
ma
in C
onc
e
p
ts
4
3
55
5
2
6.
CONCL
U
S
ION
Ta
g-base
d
s
y
s
t
em
s
ha
ve
b
ec
om
e
w
i
de
l
y
a
vai
l
a
b
l
e
t
ha
nk
s
to
t
he
ir
ad
va
nta
g
es,
w
h
ich
inc
l
ude
s
e
l
f-
orga
niza
t
i
o
n
,
curr
ency,
a
n
d
e
a
se
o
f
use
.
T
he
b
o
t
t
o
m
-
up
n
a
t
ure
o
f
the
s
e
sys
t
em
s
has
pro
v
e
d
t
o
be
a
n
in
t
e
rest
i
ng
k
n
o
w
l
e
dge
s
our
ce,
s
ince
t
he
y
pro
v
ide
a
ric
h
t
erm
i
n
o
lo
gy
gener
a
ted
by
po
te
nt
i
a
lly
l
a
r
ge
u
ser
com
m
uni
t
i
es.
Th
is
p
a
p
er
a
d
d
r
essed
t
h
e
pro
b
lem
o
f
h
ow
t
o
harve
s
t
a
nd
exp
l
oi
t
e
m
be
d
d
ed
s
em
an
t
i
cs
i
n
s
o
c
i
a
l
tag
g
i
n
g
s
yste
m
s
f
or
d
eve
l
o
p
i
n
g
sem
a
nt
ic
o
n
t
o
l
og
ie
s.
T
he
e
va
lua
t
i
o
n
of
t
he
a
l
g
ori
t
hm
,
usi
n
g
a
data
se
t
extra
c
t
ed
f
ro
m
Bi
bS
on
om
y
,
d
em
on
s
t
ra
t
e
d
tha
t
t
he
a
lgor
i
t
hm
c
o
u
l
d
e
ff
e
c
t
i
v
ely
l
e
a
r
n
do
main
o
n
t
ol
ogy
c
o
n
cept
s
,
and
i
d
ent
i
f
y
me
a
n
in
gfu
l
s
eman
t
i
c
r
e
l
at
io
ns
f
o
r
t
h
e
e
x
trac
t
e
d
c
onc
e
p
ts.
F
u
rth
e
rm
ore,
t
he
p
r
o
p
o
se
d
alg
o
ri
t
h
m
coul
d hel
p
i
n
reduc
i
n
g
com
m
on
pr
ob
l
e
ms
r
elate
d
to ta
g
a
m
b
ig
ui
t
y
and
s
yno
nym
ous
t
ags.
REFE
RENCES
[
1
]
L
.
A
t
z
o
r
i
,
A
.
I
e
r
a
,
G
.
M
o
r
a
b
i
t
o
,
a
n
d
M
.
N
i
t
t
i
,
“
T
h
e
S
o
c
i
a
l
Internet
o
f
Things
(
SI
oT)
–
W
h
en
s
o
c
ial
netw
orks
m
e
e
t
the
Int
e
rnet
o
f
Thi
n
g
s
:
Co
ncept
,
a
rchit
ectu
r
e
an
d
net
w
ork
c
h
ara
cteri
z
a
t
io
n,”
Comput. Networks
,
vol
.
56,
n
o
.
1
6
,
pp.
359
4–
36
08
,
N
o
v
.
201
2.
[2]
P
.
B
arn
a
gh
i,
W
.
W
a
ng
,
C.
H
enson
,
a
nd
K
.
Tay
l
or,
“S
em
antics
for
the
Internet
o
f
Thing
s
,”
Int.
J.
S
e
ma
nt.
W
e
b
In
f
.
Syst
.
, vo
l
.
8,
no
.
1
, pp
.
1
–2
1
, 2
01
2.
[
3
]
A
.
G
y
r
a
r
d
,
M
.
S
e
r
r
a
n
o
,
a
n
d
G
.
A
.
A
t
e
m
e
z
i
n
g
,
“
S
e
m
a
n
t
i
c
w
e
b
m
eth
odo
logi
es,
b
e
st
p
ract
ices
a
nd
ont
ol
o
g
y
engi
neeri
ng
ap
pl
ied
t
o
I
nt
ernet
of
T
h
i
ngs,
”
i
n
2015 IE
EE 2
nd W
o
rl
d F
o
ru
m on
Int
e
rnet
of
T
h
in
g
s
(
W
F-IoT)
,
20
1
5
,
pp.
4
1
2
–
417
.
[4]
E.
P
somakelis,
F.
A
i
s
opos,
A.
Litke,
K
.
Tserpes,
M
.
Kardara
,
a
nd
P
.
M
.
Cam
po,
“
Bi
g
IoT
and
So
cial
N
etw
o
rking
Data
f
or
S
mart
C
iti
e
s
-
Al
go
rith
mic
Im
prov
em
ents
o
n
Big
Dat
a
A
n
alysis
i
n
t
h
e
Co
n
t
ext
of
R
AD
ICAL
C
it
y
App
l
i
cati
o
n
s
,”
i
n
Pr
oceedi
n
g
s
of
th
e
6
t
h Int
e
rn
ation
a
l
Con
f
er
ence on Clo
ud Com
p
u
tin
g an
d
S
e
rvi
c
es
S
c
ien
c
e
,
20
16,
pp.
3
9
6
–
405
.
[5]
I.
S
zil
a
gy
i
an
d
P
.
W
ira,
“
Ontolog
i
es
a
nd
S
em
a
n
t
i
c
W
e
b
f
o
r
t
he
I
nt
ern
e
t
of
T
hin
g
s
-
a
s
u
rv
e
y
,”
i
n
IE
CON 2
016 -
42n
d Ann
ual
Conf
erence of
th
e
IEEE
Ind
u
s
t
r
i
al E
l
ect
ronics
So
ciet
y
,
2
016
,
p
p
.
6949–
69
54
.
[
6
]
N
.
L
i
n
,
F
.
T
i
a
n
,
E
.
S
u
n
,
a
n
d
C
.
W
a
n
g
,
“
S
w
a
r
m
I
n
t
e
l
l
i
g
e
n
c
e
f
o
r
P
ercept
i
o
n
L
ayer
D
esi
g
n
o
f
I
nt
ern
e
t
of
T
hi
ngs,”
In
s
t
.
Adv.
En
g.
Sci.
,
v
o
l
. Vo
l
12
,
N
o
, 2
01
4.
[7]
T.
G
ruber,
“
Ont
o
logy
o
f
Folk
so
n
o
m
y
:
A
Ma
sh
-
U
p
of
A
pp
le
s
a
n
d
O
ranges,”
Int.
J
.
Se
ma
nt.
We
b In
f. Sy
st
.
,
vo
l
.
3
,
n
o
.
1
, pp
. 1
–1
1
, 2
00
5.
[8]
C.
S
h
i
rky,
“
Onto
l
o
gy
i
s
O
v
e
rrated
--
Categories
,
Links,
a
nd
T
ag
s
,
”
20
05
.
[O
n
lin
e
]
.
A
v
ail
a
bl
e:
htt
p
://www
.sh
i
rk
y.
com/
wri
tings
/
o
ntology_overrat
e
d.
ht
m
l
?
g
oback
=.
gd
e_18
38
70
1_
mem
b
er_17
97
29
76
6.
[
A
c
c
e
s
s
ed
:
29-D
ec-20
16]
.
[9]
I.
P
eters
an
d
W.
G
.
Stock
,
“
F
o
lk
so
no
my
a
n
d
i
n
f
o
r
m
a
ti
on
retr
ie
v
a
l
,
”
Pro
c
.
Am.
So
c
.
In
f.
Sc
i. Te
c
h
no
l.
,
vo
l.
4
4
,
n
o.
1
,
pp.
1
–
2
8
,
O
c
t
.
200
7.
[10
]
A
.
M
i
kro
y
an
nidis
,
“
T
o
w
a
rd
a
S
o
c
i
a
l
Se
m
a
n
t
i
c
W
eb
,
”
Co
mp
ut
e
r
(L
o
n
g
.
B
e
ach.
Ca
l
i
f)
.
,
vol.
40,
n
o.
1
1
,
p
p.
1
1
3
–
115
,
No
v. 20
0
7
.
[11
]
T
.
V
a
nd
er W
al,
“
F
o
l
ks
onom
y
Co
in
age
an
d
D
e
fin
i
ti
on.
”
J
un-2
0
07.
[12
]
A
.
M
a
t
h
es
,
“
F
o
l
kso
nom
ies
-
Coo
p
erati
v
e
Cl
assificat
io
n
an
d
Co
mmun
i
c
a
tion
Th
rou
g
h
S
h
a
re
d
Me
ta
da
ta
,”
200
4.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN: 2252-
8776
IJ-ICT
Vol
.
8,
No.
1
, A
pril 20
19
:
19
–
2
4
24
[Onlin
e].
A
v
ailab
l
e:
h
ttp:/
/w
ww
.adam
m
a
th
es.com
/academ
i
c
/
c
o
m
p
u
t
e
r-m
e
diat
e
d
-com
m
u
n
i
c
a
t
i
on/f
o
lkso
no
mies.
h
t
m
l.
[Access
e
d:
02
-
J
u
l-20
16]
.
[13]
A
.
Hotho,
R
.
Jäschke,
C
.
Sc
hm
it
z
,
a
nd
G
.
Stu
m
me
,
“In
f
o
r
ma
t
ion
Retriev
a
l
in
F
olk
s
o
nom
ies:
S
e
a
rch
an
d
Rank
in
g,
”
in
Proceed
in
gs
o
f
the 3r
d
Eu
ro
p
e
a
n
conf
eren
ce o
n
T
h
e Sema
nt
ic
W
e
b: res
e
ar
ch an
d app
li
ca
t
i
ons
,
S
p
ring
er-Verl
a
g,
2
0
0
6
,
p
p.
4
11
–4
26
.
[1
4]
M
.
Sz
o
m
sz
or
et
a
l
.
,
“
F
olk
s
on
o
m
ies,
t
he
S
em
anti
c W
e
b,
a
nd
M
ov
ie
R
eco
m
m
en
dation,
”
2
00
7.
[15]
H
.
S.
A
l-K
h
al
if
a
and
H.
C
.
D
a
v
i
s,
“
T
o
w
a
rds
be
tter
underst
a
nd
in
g
of
f
olk
s
o
n
o
m
i
c
p
at
terns
,
”
i
n
Proceed
in
gs o
f
t
h
e
18t
h
c
o
nf
eren
ce
on
Hyp
e
rtext
and
hyper
m
ed
ia -
HT
’07
,
20
07,
p
.
16
3.
[16
]
A
.
G
a
rcía-Sil
va,
O.
C
o
r
cho,
H
.
Alan
i,
a
nd
A.
G
óm
ez-P
é
rez,
“Revi
e
w
of
t
he
s
t
a
te
o
f
the
art:
di
scovering
and
asso
ciat
in
g
s
e
m
a
n
t
i
c
s
t
o
tags
i
n
fo
l
ks
on
o
m
i
e
s
,
”
K
n
owl
.
Eng
.
Rev.
,
v
o
l.
2
7
, n
o.
1,
p
p
. 57
–
8
5
, Mar.
20
12
.
[17
]
M
.
Al
ruq
i
mi
a
nd
N
.
Aknin
,
“
S
e
m
a
ntic
E
m
e
rgence
F
r
om
S
o
c
ial
Tagg
in
g
S
y
st
e
m
s,
”
In
t.
J. O
r
g
a
n.
Collect
. In
t
e
ll.
,
vol
.
5
,
n
o.
1
,
p
p
.
16
–3
1,
2
01
5.
[18
]
F
.
J
a
been,
S
.
K
h
u
sro
,
A
.
M
a
j
i
d
,
a
nd
A
.
Rau
f
,
“
S
em
ant
i
cs
d
i
sco
v
ery
in
s
ocial
t
a
gg
ing
s
y
stem
s
:
A
r
evi
e
w,”
Mul
time
d
.
T
ool
s App
l
.
,
vo
l.
7
5
,
no.
1
,
p
p
.
5
7
3
–6
05
,
J
an
.
2
016.
[19
]
H
.
Li
u
,
H
.
Ch
en,
M.
L
i
n
,
and
Y.
W
u,
“
Comm
un
it
y
Det
e
ct
ion
B
ased
o
n
Topi
c
Distance
in
S
o
c
ial
T
a
gging
Netw
orks
,”
TE
LKOM
NIKA
Ind
o
n
e
s.
J. El
ectr.
En
g.
,
vol.
12,
no.
5
,
M
ay
201
4
.
[
2
0
]
M
.
N
.
U
d
d
i
n
,
T
.
H
.
D
u
o
n
g
,
N
.
T
.
N
g
u
y
e
n
,
X
.
M
.
Q
i
,
a
n
d
G
.
S
.
J
o
,
“S
em
anti
c
sim
i
larity
m
easu
r
es
f
o
r
e
nhan
c
in
g
inform
at
i
o
n
retri
e
val i
n
folksonomi
e
s
,
”
Expe
rt Sy
st.
A
p
pl.
, 2
01
3.
[21
]
A
.
Zub
i
ag
a,
V
.
F
r
es
no,
R
.
Ma
rt
in
ez,
a
n
d
A
.
P
.
G
arci
a-Plaza
,
“
H
arnes
s
in
g
Fol
k
so
nom
ies
to
P
rod
u
ce
a
S
o
ci
al
Classif
i
catio
n
o
f
Res
ources
,”
IEE
E
Tr
an
s.
Knowl
.
D
a
t
a
E
ng.
, vo
l
. 2
5,
no
.
8
,
pp
. 1
80
1–
1
8
1
3
, Aug
.
2
0
1
3
.
[22
]
A
.
Tomm
as
e
l
a
nd
D.
G
o
d
o
y
,
“S
em
a
n
ti
c
g
r
ou
ndi
ng
o
f
so
cia
l
a
n
no
tati
on
s
f
o
r
en
han
c
in
g
res
o
u
r
ce
clas
sificati
on
i
n
folksonomies
,
”
J. Intel
l
.
Inf
.
Sys
t
.
,
vo
l
.
4
4,
n
o
.
3
,
p
p
.
4
15–
44
6,
J
un.
2
01
5.
[23
]
I
.
Can
t
ad
or,
I.
K
o
n
s
t
as,
an
d
J
.
M
.
J
o
se,
“
C
at
e
g
orisi
n
g
s
o
c
i
a
l
tag
s
t
o
i
m
prov
e
f
o
lk
so
no
m
y
-bas
ed
r
e
c
om
m
e
nd
ati
ons,”
We
b
S
e
man
t
.
Sc
i
.
Se
rv
.
Ag
e
n
t
s
W
o
rld
Wide
We
b
,
v
o
l
. 9
, n
o.
1
, p
p.
1
–
1
5
, Mar.
2
0
1
1
.
[24
]
I
.
Chin
g
Hsu
,
“In
t
eg
ratin
g
ontolo
gy
tech
no
log
y
w
ith
f
olks
o
n
o
m
i
es
f
o
r
p
erso
n
a
li
zed
s
o
c
ial
tag
reco
m
m
en
dat
i
on,”
App
l
.
Soft
Comput
.
,
vo
l
.
1
3,
n
o
.
8
,
pp
.
3
745
–3
75
0,
A
ug
.
2
013
.
[2
5]
F
.
Fo
n
t
,
J
.
S
e
r
r
à
,
a
n
d
X.
S
e
rra,
“Ana
ly
si
s
of
t
he
I
mp
a
c
t
o
f
a
Tag
Recom
m
e
nd
atio
n
Sy
s
t
em
i
n
a
Real
-W
orl
d
F
o
lk
so
nomy,
”
ACM
T
r
ans. In
te
l
l
.
Syst
. T
ech
nol
.
,
201
5.
[2
6]
A
.
M
.
El-k
o
ra
n
y
a
n
d
S.
M
.
Kha
t
a
b
,
"
O
ntolog
y
-
b
a
se
d
So
c
i
a
l
R
ec
o
m
m
e
n
d
e
r
S
y
s
t
em
,
"
In
st
.
Adv
.
En
g. Sc
i.
,
v
o
l
.
Vo
l
1,
no.
3
,
2
0
12.
[27
]
S
.
Ham
d
i
,
A
.
Lo
pes
G
a
nc
arsk
i,
A
.
Bo
uzegh
o
u
b
,
an
d
S
.
B
en
Y
a
hi
a,
“
En
rich
ing
On
tolo
gies
f
rom
F
o
l
k
s
o
n
o
m
i
es
f
or
eLearni
ng
:
DBpedi
a
Case,
”
i
n
2
012
IEEE 1
2
th In
tern
atio
nal
Con
f
er
ence on
Ad
van
ced L
e
ar
n
i
ng
Tech
no
lo
gie
s
,
2
0
1
2
,
p
p.
2
93
–2
97
.
[28]
A
.
G
a
rcía-Sil
va,
L.
J
.
Garc
ía-Cas
tro,
A
.
Garcí
a
,
and
O.
C
o
rcho,
“
S
o
cial
T
ag
s
an
d
Lin
ked
Dat
a
f
or
O
n
t
o
l
ogy
Devel
o
p
m
e
n
t:
A
C
ase
S
t
u
dy
in
t
he
F
inan
cial
D
om
ain
,
”
i
n
Pro
ceedi
n
g
s
of th
e 4t
h
Int
e
rnati
o
n
a
l
}
Conf
eren
ce on
W
e
b
Intelli
gen
c
e
,
Mini
ng
an
d
S
e
ma
nti
c
s (W
IM
S
14)
, 2
01
4, p.
3
2
:
1--3
2:
1
0
.
[29
]
S
.
Wang
,
W.
W
an
g,
Y
.
Zh
uang
,
and
X.
F
ei,
“
A
n
on
tolog
y
e
v
o
l
uti
on
m
e
tho
d
b
a
s
ed
on
f
o
lkso
nomy
,
”
J.
App
l
.
Re
s
.
Te
c
h
n
o
l.
,
v
o
l
.
13,
n
o
.
2
,
p
p
.
1
7
7
–
1
87,
A
p
r
.
2
015.
[30
]
G
.
Be
g
e
lm
an
,
P.
K
el
ler,
a
nd
F
.
S
m
a
d
j
a,
“
Aut
o
m
a
t
e
d
T
a
g
Clus
tering
:
Im
p
r
oving
s
earchi
ng
and
ex
p
l
o
r
ati
on
in
t
he
t
a
g
space,
” in
WW
W20
06
, 20
0
6
.
[31
]
P
.
S
c
hmit
z,
“
In
d
u
ci
ng
O
nto
l
og
y
f
r
om
F
li
ckr
Tags,
”
i
n
Pro
ceedi
n
g
s
of
t
h
e
W
o
r
ksho
p o
n
Co
l
l
a
bor
a
tive T
a
ggin
g
at
WWW
20
06
, 2
00
6.
[32
]
P
.
Hey
m
ann
and
H.
G
arci
a-Mo
li
na,
“
C
o
l
l
a
bo
rativ
e
Creati
o
n
o
f
Comm
un
al
H
ierarch
i
cal
T
axonomies
in
S
o
c
i
a
l
Tagging
Sys
t
ems.” 2006.
[3
3]
P
.
Mika
,
“
O
nto
l
og
ie
s
Are
Us:
A
U
ni
fie
d
M
o
d
e
l
o
f
Soc
i
a
l
N
e
t
w
o
rk
s
an
d
S
e
m
a
ntics
,
”
S
p
ring
er
B
e
r
li
n
H
e
id
e
l
b
e
rg
,
2
0
0
5
,
p
p.
5
22
–5
36
.
[34
]
S
.
An
gel
e
tou,
“
S
e
m
a
ntic
E
nri
c
hm
en
t
of
F
ol
k
s
on
om
y
Tags
paces
,”
i
n
T
h
e Se
ma
ntic
W
e
b
- ISW
C
200
8
,
Berlin,
Heid
elb
e
rg:
S
p
ringer
Berl
in
Hei
d
e
l
b
erg,
2
0
08,
p
p
.
8
89
–89
4.
[35
]
I
.
Cant
ador,
M
.
S
zom
s
z
o
r,
H
.
Al
ani,
M
.
F
e
rnan
dez,
a
n
d
P
.
C
a
s
te
lls,
“
E
nr
ic
hi
ng
o
n
t
o
l
og
ic
a
l
u
se
r
pr
ofi
l
e
s
w
i
t
h
tagging
hist
ory
f
o
r
multi-do
main
r
e
c
ommendat
i
ons,”
in
1st
Inte
rn
at
ion
a
l
W
o
rk
sh
op
o
n
Colle
c
tiv
e
Se
ma
ntic
s:
Coll
ecti
ve Int
e
lligen
ce
&
a
m
p
; t
h
e S
e
ma
nti
c
W
e
b
(
C
IS
W
e
b 2
0
0
8
)
, 2
00
8.
[36
]
J
.
W
e
i
and
F
.
M
eng
,
“
Is
c
o
l
lect
iv
e
i
n
telli
gen
ce
hel
p
s
m
o
re
i
n
p
o
lys
e
m
y
t
ag
optimazed
a
lg
ori
t
hm
t
han
com
m
o
n
s
en
se
tool
,
”
Cl
u
s
t
e
r
Com
p
u
t
.
,
Ju
l. 2
01
7.
[3
7]
P
.
Mika
,
“
O
nto
l
og
ie
s
Are
Us:
A
U
ni
fie
d
M
od
e
l
o
f
Soc
i
a
l
N
e
t
w
o
rks
and
Sem
a
ntics,”
Spri
nger
B
e
r
li
n
Heide
l
ber
g
,
2
0
0
5
,
p
p.
5
22
–5
36
.
[38
]
K
.
Dellsch
aft
and
S
.
S
taab
,
“On
Ho
w
to
P
erf
o
rm
a
G
ol
d
Stan
dard
B
ased
E
va
l
u
ati
o
n
o
f
O
n
t
o
l
ogy
L
earni
ng,
”
i
n
Pro
ceedi
ngs
of
the 5
t
h
inter
n
a
t
iona
l co
nf
erence
o
n
T
h
e S
e
manti
c
W
e
b
,
S
p
ring
er-Verl
a
g
,
2
00
6,
p
p.
22
8
–
241
.
[39
]
“
K
n
o
w
l
e
dge
&
D
ata
En
gi
neering
Gro
u
p
,
U
ni
ve
rs
it
y
of
K
assel:
B
en
chm
a
rk
F
o
l
ks
on
om
y
Dat
a
f
ro
m
Bi
b
S
o
n
o
m
y,
v
e
rsio
n o
f
S
e
p
te
mbe
r
3
0s
t, 2
00
8.
” [Onl
i
n
e]. Avai
l
able:
h
t
tps:
//
w
w
w
.
kd
e.
cs.
u
n
i
-kassel.
de/bib
so
nomy/d
u
m
p
s/.
[40
]
H
.
Du,
S
.
K
.
W.
C
hu,
a
nd
F
.
T.
Y
.
La
m
,
“
Soci
al
b
o
okm
arkin
g
a
nd
t
aggi
ng
b
ehav
io
r:
a
n
empiri
cal
a
nal
y
s
i
s
o
n
deli
cio
u
s
and
co
n
n
o
t
ea,” i
n
Pro
ceedin
gs
of
t
h
e 2
0
0
9
In
tern
a
t
io
nal
Co
nfer
ence o
n
Kn
owl
e
dge
M
a
n
a
g
e
ment
, 20
0
9
.
[41
]
D
.
H
.
L
ee,
“
Co
mp
arati
v
e
An
alysi
s
o
f
In
dex
Term
s
an
d
S
o
cial
T
ags
,
”
J.
KOR
E
AN Soc.
Li
br
.
In
f
.
S
c
i.
,
vo
l.
4
9,
p
p.
291
–3
11
,
2
0
1
5
.
[42
]
H
.
Do
ng,
W
.
W
a
ng
,
an
d
F
.
C
o
e
nen,
“
Deri
v
i
ng
D
y
n
am
i
c
K
nowled
ge
f
rom
A
cadem
ic
S
oci
a
l
Tag
g
ing
Data:
A
No
vel
Re
s
earch D
irect
ion
,
” i
n
i
C
onfere
n
ce 20
17
Pr
oceed
ing
s
,
20
17,
pp.
6
61
–66
6.
Evaluation Warning : The document was created with Spire.PDF for Python.