Indonesi
an
Journa
l
of El
ect
ri
cal Engineer
ing
an
d
Comp
ut
er
Scie
nce
Vo
l.
24
,
No.
2
,
N
ov
em
ber
20
21
, pp.
1228
~
123
7
IS
S
N: 25
02
-
4752, DO
I: 10
.11
591/
ijeecs
.v
24
.i
2
.
pp
1228
-
123
7
1228
Journ
al h
om
e
page
:
http:
//
ij
eecs.i
aesc
or
e.c
om
A
s
eman
tic web s
ervi
ce
s
discov
ery approa
ch
integr
ating
mu
ltip
le similarity m
ea
su
res and
k
-
means clust
er
i
ng
Moura
d F
aris
s,
N
aou
fa
l E
l
Allali
, Hakim
a Asa
idi
, Mo
h
amed
Bel
louki
Moham
m
ed
Firs
t
Univer
si
t
y
Ouj
da,
FP
D Nador,
LMAS
I,
Nador,
Morocc
o
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
J
u
l
3
,
2021
Re
vised
Sep
17
,
2021
Accepte
d
Se
p
22
,
2021
W
eb
servic
e
(W
S)
discove
r
y
is
an
essenti
a
l
ta
sk
for
implementi
ng
complex
appl
i
ca
t
ions
in
a
servic
e
ori
ented
arc
hi
te
c
ture
(SO
A),
such
as
sele
cting,
compos
ing,
and
pr
ovidi
ng
serv
ices.
Th
is
ta
sk
is
li
m
it
ed
s
emanti
c
al
l
y
in
the
inc
orpora
ti
on
of
the
customer’s
req
uest
and
the
web
servic
es.
Furthermore,
apply
ing
sui
ta
bl
e
sim
il
arit
y
m
ethods
for
the
increa
sing
num
ber
of
W
Ss
is
m
ore
rel
ev
ant
for
eff
icien
t
web
service
disc
over
y
.
To
over
come
the
s
e
li
m
it
ations,
we
p
ropose
a
n
ew
ap
proa
ch
for
web
s
erv
ice
discov
er
y
integra
t
ing
m
ult
ipl
e
sim
il
arit
y
m
ea
sures
and
k
-
m
ea
ns
cl
usteri
ng.
The
appr
o
a
ch
ena
bl
es
m
ore
ac
cur
at
e
s
erv
ices
appr
opri
at
e
to
the
custo
m
er'
s
req
uest
b
y
ca
l
cul
a
ti
ng
diffe
ren
t
sim
il
ar
ity
scor
es
bet
wee
n
the
customer'
s
req
u
est
and
the
web
servic
es.
The
gl
obal
sem
ant
i
c
sim
il
ari
t
y
is
de
te
r
m
ine
d
b
y
app
l
y
i
ng
k
-
m
ea
ns
cl
uster
ing
using
the
obta
ine
d
sim
il
ari
t
y
score
s
.
The
expe
rime
nta
l
result
s
demons
tra
te
d
th
at
the
proposed
sem
ant
ic
web
s
erv
ic
e
discov
er
y
appr
oa
ch
outpe
rform
s
the
stat
e
-
of
-
the
app
roa
che
s
in
t
erms
of
pre
ci
sion
(9
8%),
re
call
(95%),
and
F
-
m
ea
sure
(96%).
T
he
proposed
app
roa
ch
is
eff
i
cient
l
y
d
esigne
d
to
support
and
fac
il
i
ta
t
e
the
sel
ec
t
ion
and
compos
it
ion
of
web
servic
es
phase
s
i
n
complex
app
lications.
Ke
yw
or
ds:
K
-
m
eans
On
t
ology
Sem
antic
m
eas
ur
e
of sim
il
arity
SOA
Web ser
vice
di
sco
ver
y
This
is an
open
acc
ess arti
cl
e
un
der
the
CC
B
Y
-
SA
l
ic
ense
.
Corres
pond
in
g
Au
th
or
:
Mourad
Fa
riss
LMASI
, FPD
Nado
r
Moh
am
m
ed
First U
niv
e
rsity
Nado
r,
B
V
Mo
ham
m
ed
V
I
B.
P. 524
O
ujd
a
6000
0
Mo
r
occo
Em
a
il
:
m
.f
ariss@u
m
p.
ac.m
a
1.
INTROD
U
CTION
W
e
b
s
e
r
v
i
c
e
d
i
s
c
o
v
e
r
y
i
s
t
o
f
i
n
d
t
h
e
r
e
l
e
v
a
n
t
w
e
b
s
e
r
v
i
c
e
s
t
h
a
t
s
a
t
i
s
f
y
t
h
e
s
e
r
v
i
c
e
c
u
s
t
om
e
r
s
'
r
e
q
u
i
r
e
m
e
nt
s
.
O
w
e
d
t
o
t
h
e
i
n
c
r
e
a
s
e
d
n
um
b
e
r
o
f
p
u
b
l
i
s
h
e
d
w
e
b
s
e
r
v
i
c
e
s
(
W
S
s
)
o
n
t
h
e
I
n
t
e
r
n
e
t
,
t
h
e
d
i
s
c
o
v
e
r
y
s
t
a
ge
o
f
f
e
r
s
a
n
i
m
po
r
t
a
n
t
n
um
b
e
r
o
f
c
a
n
d
i
d
a
t
e
W
S
s
f
o
r
a
g
i
v
e
n
r
e
q
u
e
s
t
.
T
h
e
W
S
d
i
s
c
o
v
e
r
y
c
a
n
b
e
p
e
r
f
o
r
m
e
d
s
y
n
t
a
c
t
i
c
a
l
l
y
b
a
s
e
d
on
k
e
y
w
o
r
d
s
o
r
s
e
m
a
nt
i
c
a
l
l
y
b
a
s
e
d
o
n
W
S
d
e
s
c
r
i
p
t
i
o
n
.
T
h
e
s
y
nt
a
c
t
i
c
m
e
c
h
a
n
i
s
m
i
s
l
i
m
i
t
e
d
t
o
r
e
s
p
o
n
d
t
o
t
he
f
u
n
c
t
i
o
n
a
l
c
us
t
om
e
r
r
e
q
u
i
r
e
m
e
nt
s
.
T
h
i
s
m
a
k
e
s
i
n
t
r
o
d
u
c
i
n
g
a
n
e
w
m
e
c
h
a
ni
s
m
n
e
c
e
s
s
a
r
y
,
w
h
i
c
h
i
n
v
o
l
v
e
s
t
h
e
l
o
c
a
l
i
z
a
t
i
o
n
o
f
W
S
s
b
a
s
e
d
o
n
t
h
e
c
a
p
a
b
i
l
i
t
i
e
s
t
he
y
o
f
f
e
r
.
I
n
t
e
g
r
a
t
i
o
n
o
f
s
e
m
a
n
t
i
c
t
e
c
h
n
i
q
u
e
s
i
n
W
S
s
c
a
n
p
l
a
y
a
n
i
m
p
o
r
t
a
n
t
r
o
l
e
i
n
i
n
c
o
r
p
o
r
a
t
i
n
g
d
i
f
f
e
r
e
n
t
t
e
r
m
i
n
o
l
o
g
i
e
s
i
n
W
S
s
.
F
u
r
t
h
e
r
m
o
r
e
,
a
d
o
p
t
i
n
g
t
h
e
r
i
g
h
t
m
e
a
s
u
r
e
s
i
m
i
l
a
r
i
t
y
t
o
f
i
n
d
m
o
s
t
s
i
m
i
l
a
r
W
S
s
d
u
r
i
n
g
t
h
e
d
i
s
c
o
v
e
r
y
p
h
a
s
e
i
s
c
om
p
l
e
x
t
a
s
k
.
The
tra
diti
onal
unive
rsal
de
sc
riptio
n
,
disco
ve
ry,
a
nd
integ
r
at
ion
(
U
DDI
)
on
ly
s
upports
keyw
ords
for
searchi
ng
WSs;
this
search
cannot
find
t
he
w
ho
le
rele
van
t
WSs
f
or
custom
ers
[1]
,
[2]
.
Keyw
ords
are
insuffici
ent
to
express
sem
antic
con
cepts,
s
e
m
antic
al
ly
diff
er
e
nt
con
ce
pt
s
can
hav
e
the
sa
m
e
rep
resent
at
ion
,
wh
ic
h
furthe
r
infl
uen
ces
the
a
ccur
acy
[
3]
.
H
ence,
se
ver
al
appr
oac
hes
have
been
pr
opos
e
d
to
ad
d
the
sem
antic
con
ce
pts
t
o
th
e
W
S
desc
rip
ti
on
as
we
b
s
erv
ic
e
m
od
el
ing
la
ngua
ge
(
WSML
)
,
web
ser
vice
desc
r
ipti
on
la
nguag
e
with
sem
antic
s
(
W
S
DL
-
S
)
,
ontolo
gy
w
eb
la
ngua
ge
f
or ser
vices
(
O
WL
-
S
),
to fa
ci
li
ta
te
the d
isc
ov
e
ry
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
A sema
ntic w
eb
servi
ces
d
isc
overy
appro
ac
h
inte
grati
ng
mu
lt
iple si
mila
rit
y m
eas
ur
es
…
(
Mou
r
ad F
ari
ss
)
1229
and sele
ct
i
on
t
asks
[4]
. Ma
ny
ex
ist
ing we
b
s
erv
ic
e
disco
ve
r
y approac
hes
a
r
e b
a
sed
on ope
rati
on
al
si
gn
i
ng and
sever
al
m
at
ching
m
echan
ism
s.
W
e
dif
fer
e
ntiat
e
betwee
n
log
ic
-
base
d
r
easo
ning
[
5]
and
non
-
lo
gic
-
base
d
te
chn
iq
ues
(i.e
.,
gr
a
ph
m
at
ching
,
sim
i
la
rity
m
easur
es
[
6],
[
7]
,
gr
a
ph
m
at
ching,
a
nd
da
ta
m
ining
[
8]
)
a
nd
hybri
d disco
ve
ry
[9
]
, [1
0]
(i.e
.,
m
et
ho
ds t
hat
util
iz
e log
ic
al
and no
n
-
l
og
ic
a
l
m
at
ching
both)
.
Seve
ral
w
orks
dev
el
op
t
he
noti
on
of
sim
il
a
rity
in
the
dis
cov
e
ry
of
we
b
serv
ic
es
[11]
-
[17]
;
so
m
e
researc
hers
ha
ve
pro
posed
a
nd
stu
died
thei
r
si
m
il
arity
m
e
asur
e
s
as
in
[12],
[18],
[19]
.
Othe
rs
ha
ve
w
orke
d
with
al
rea
dy
existi
ng
m
easure
s
li
ke
in
[20]
-
[22]
.
Eve
n
th
ough
the
e
xisti
ng
of
m
ulti
ple
s
i
m
i
la
rity
m
eas
ur
es
t
o
cal
culat
e
the
si
m
ilit
ud
e
be
twe
en
the
we
b
ser
vice
an
d
t
he
c
us
tom
er
re
qu
e
st,
they
ar
e
s
om
e
par
ti
cularit
ie
s.
I
n
m
or
e
detai
l,
the
web
se
rv
ic
es
are
m
od
el
ed
usi
ng
diff
e
re
nt
f
or
m
at
s,
the
en
ough
pu
blishe
d
ontol
og
ie
s
of
W
Ss
with
O
WL
-
S
,
t
he
acc
ur
at
e
inf
or
m
at
ion
ab
ou
t
the
WS
duri
ng
the
se
rv
ic
e
di
scov
e
ry
is
require
d
.
Acc
ordi
ng
to
these
par
ti
cula
rit
ie
s,
con
ce
ntr
at
ing
on
one
m
easur
e
of
si
m
il
arity
can
lose
the
interest
s
of
oth
e
r
m
eas
ur
es
as
well
as ca
n
in
fl
uen
ce
the
res
ul
ts by it
s lim
it
ation
s
.
T
h
i
s
p
a
p
e
r
p
r
o
p
o
s
e
s
a
n
e
w
s
o
l
u
t
i
o
n
f
o
r
t
h
e
s
i
m
i
l
a
r
i
t
y
m
e
a
s
ur
e
t
o
c
a
l
c
u
l
a
t
e
t
h
e
s
i
m
i
l
a
r
i
t
y
b
e
t
w
e
e
n
a
w
e
b
s
e
r
v
i
c
e
f
r
om
a
d
a
t
a
s
e
t
o
f
s
e
m
a
nt
i
c
w
e
b
s
e
r
v
i
c
e
s
a
n
d
c
u
s
t
om
e
r
s
'
r
e
q
u
e
s
t
s
.
O
n
s
o
l
v
i
n
g
t
h
i
s
p
r
o
b
l
e
m
,
o
t
h
e
r
s
s
y
s
t
e
m
a
t
i
c
a
l
l
y
h
a
v
e
a
n
a
v
e
n
u
e
f
o
r
i
m
p
r
o
v
i
n
g
r
e
s
u
l
t
s
,
s
u
c
h
a
s
s
e
l
e
c
t
i
n
g
w
e
b
s
e
r
v
i
c
e
s
a
n
d
t
he
c
om
p
o
s
i
t
i
o
n
o
f
w
e
b
s
e
r
v
i
c
e
s
.
C
o
n
s
e
q
u
e
n
t
l
y
,
a
n
e
w
a
p
p
r
o
a
c
h
t
o
s
o
l
v
e
t
h
e
p
r
o
b
l
e
m
o
f
s
e
m
a
n
t
i
c
w
e
b
s
e
r
v
i
c
e
s
d
i
s
c
o
v
e
r
y
b
y
d
e
c
r
e
a
s
i
n
g
t
h
e
n
um
b
e
r
o
f
d
i
s
c
o
v
e
r
e
d
W
S
s
t
h
r
o
u
g
h
i
n
c
r
e
a
s
i
n
g
t
h
e
p
r
e
c
i
s
i
o
n
.
O
u
r
a
p
p
r
o
a
c
h
c
o
n
s
i
s
t
s
o
f
c
a
l
c
u
l
a
t
i
n
g
s
i
x
s
i
m
i
l
a
r
i
t
y
m
e
a
s
u
r
e
s
,
n
a
m
e
l
y
;
E
u
c
l
i
d
e
a
n
d
i
s
t
a
n
c
e
,
M
a
n
h
a
t
t
a
n
d
i
s
t
a
n
c
e
,
W
u
a
n
d
P
a
l
m
e
r
’
s
i
m
i
l
a
r
i
t
y
,
C
o
s
i
n
e
s
i
m
i
l
a
r
i
t
y
,
J
a
c
c
a
r
d
'
s
i
n
d
e
x
,
a
n
d
l
o
g
i
c
a
l
c
o
r
r
e
s
p
o
n
d
e
n
c
e
.
a
n
d
u
s
e
t
h
e
k
-
m
e
a
n
s
c
l
u
s
t
e
r
i
n
g
t
o
d
e
s
i
g
n
a
t
e
t
he
d
i
s
c
o
v
e
r
e
d
w
e
b
s
e
r
v
i
c
e
s
m
o
s
t
s
i
m
i
l
a
r
t
o
t
h
e
c
u
s
t
om
e
r
’
s
r
e
q
ue
s
t
b
a
s
e
d
o
n
t
h
e
s
i
m
i
l
a
r
i
t
y
s
c
or
e
s
o
f
e
a
c
h
W
S
.
T
h
e
r
e
m
a
i
n
d
e
r
o
f
t
h
e
p
a
p
e
r
i
s
s
t
r
u
c
t
u
r
e
d
a
s
f
o
l
l
o
w
s
:
s
e
c
t
i
o
n
2
i
n
c
l
u
d
e
s
t
h
e
r
e
l
a
t
e
d
w
o
r
k
s
a
n
d
t
h
e
m
o
t
i
v
a
t
i
o
n
,
s
e
c
t
i
o
n
3
d
e
t
a
i
l
s
t
h
e
p
r
o
b
l
e
m
a
n
d
t
h
e
n
e
c
e
s
s
a
r
y
b
a
c
k
g
r
o
u
n
d
t
o
s
o
l
v
e
i
t
.
S
e
c
t
i
o
n
4
g
i
v
e
s
o
u
r
p
r
o
p
o
s
e
d
c
o
n
t
r
i
b
u
t
i
o
n
.
S
e
c
t
i
o
n
5
d
i
s
c
u
s
s
e
s
t
h
e
e
x
p
e
r
i
m
e
n
t
a
l
r
e
s
u
l
t
s
.
S
e
c
t
i
o
n
6
c
l
o
s
e
s
t
h
e
p
a
p
e
r
w
i
t
h
t
h
e
c
o
n
c
l
u
s
i
o
n
.
2.
RELATE
D
W
ORK A
N
D M
OTIV
ATIO
N
We
can
m
ention
a
set
of
m
at
c
hm
aker
s
for
WSs
de
velo
ped
i
n
the
li
te
ratur
e,
su
ch
as
[
23]
,
that
app
li
e
d
three
functi
on
s
to
cal
culat
e
the
le
xical
sp
e
ci
ficat
ion
sim
i
la
rity
and
s
howed
t
he
best
perform
ance
f
or
t
he
cl
assic
al
vecto
r
s
pace
m
od
el
.
S
ur
pr
isi
ng
ly
,
sem
antic
si
m
il
arit
y
m
e
tric
s
did
no
t
help
i
m
pr
ove
the
s
erv
ic
e
interface
m
app
in
g
accu
racy
and
recall
.
Cl
assic
al
te
r
m
f
reque
ncy
–
in
ve
rse
doc
um
ent
fr
e
qu
e
ncy
(
TF
-
I
DF
)
heurist
ic
s
outp
erfor
m
ed
ot
her
ap
proac
hes
i
n
m
os
t
cases.
Du
e
to
the
e
xc
essive
ge
ner
a
l
it
y
of
the
WordNet
on
t
ology,
m
an
y
false
cor
res
ponde
nces
ha
ve
been
f
ound.
T
he
stud
y
co
ncl
ud
e
d
that
the
se
m
antic
si
m
il
a
rity
has
no g
ai
n
in
prec
isi
on
due to
the
choice
of t
he data
set
, whic
h i
s
W
S
DL fil
es
that i
nf
l
uen
ce
the
resu
lt
s
obta
ined
.
The
C
ondorcet
-
f
us
e
syst
em
[24]
is
base
d
on
a
m
ajo
rity
vo
t
ing
sc
hem
e.
More
e
xpli
ci
tl
y,
a
docum
ent
d1
is
cl
assifi
ed
befor
e
a
no
t
her
do
c
um
ent
d2
in
the
com
bin
e
d
li
st,
if
d1
is
c
la
ssifie
d
be
for
e
d2
m
or
e
tim
e
s
tha
n
d2
is
cl
assifi
e
d
befor
e
d1.
T
he
ou
tr
an
king
a
ppr
oach
[25]
adjusts
the
m
ajorit
y
vo
ti
ng
m
o
del
by
set
ti
ng
a
set
of
thres
ho
l
ds
. E
xperim
ental
ev
al
uation no
te
s t
hat p
er
f
or
m
ance is v
ery deli
cat
e to thr
es
ho
l
d
values
. Th
e
propos
e
d
appr
oach use
d
diff
e
re
nt
sim
i
lar
it
y
m
easur
es t
hro
ugh
t
he
to
p
-
ra
nk
i
ng r
at
io
.
F
e
t
h
a
l
l
a
h
e
t
a
l
.
[26]
f
o
c
u
s
e
d
o
n
a
u
t
om
a
t
i
c
s
e
r
v
i
c
e
d
i
s
c
o
v
e
r
y
b
a
s
e
d
o
n
t
h
e
s
e
m
a
nt
i
c
w
e
b
.
A
s
a
s
o
l
u
t
i
o
n,
t
h
e
y
p
r
o
p
o
s
e
d
a
n
a
p
p
r
o
a
c
h
t
h
a
t
u
t
i
l
i
z
e
s
t
h
e
s
e
r
v
i
c
e
i
n
t
e
r
f
a
c
e
a
n
d
t
h
e
d
om
a
i
n
o
n
t
o
l
o
g
y
t
o
m
o
d
e
l
w
e
b
s
e
r
v
i
c
e
s
.
T
h
e
n
,
t
h
e
y
c
a
l
c
u
l
a
t
e
d
t
h
e
s
i
m
i
l
a
r
i
t
y
s
c
o
r
e
u
s
i
ng
a
m
a
t
c
hi
n
g
a
l
g
o
r
i
t
hm
b
e
t
w
e
e
n
t
h
e
r
e
q
u
e
s
t
a
n
d
t
h
e
w
e
b
s
e
r
v
i
c
e
m
o
d
e
l
,
w
h
i
c
h
i
s
b
a
s
e
d
o
n
t
h
e
W
u
a
n
d
P
a
l
m
e
r
s
i
m
i
l
a
r
i
t
y
m
e
a
s
u
r
e
.
T
h
e
d
a
t
a
s
e
t
u
s
e
d
t
o
e
v
a
l
u
a
t
e
t
h
e
a
p
p
r
o
a
c
h
i
s
s
a
m
pl
e
d
f
r
om
t
h
e
o
n
t
ol
o
g
y
w
e
b
l
a
n
g
u
a
g
e
s
e
r
v
i
c
e
s
-
t
e
s
t
c
o
l
l
e
c
t
i
o
n
(
O
W
L
S
-
TC
)
c
o
r
p
u
s
v
e
r
s
i
o
n
2
.
2
.
1
.
T
h
i
s
a
p
p
r
o
a
c
h
i
s
c
h
a
r
a
c
t
e
r
i
z
e
d
b
y
i
t
s
s
i
m
p
l
i
c
i
t
y
a
nd
i
t
s
w
e
a
k
c
om
p
l
e
x
i
t
y
,
b
u
t
s
t
i
l
l
s
h
o
w
a
w
e
a
k
n
e
s
s
i
n
r
e
c
a
l
l
.
Wu
e
t
a
l
.
[27]
,
a
m
e
t
h
o
d
w
a
s
p
r
o
p
o
s
e
d
t
o
f
a
c
i
l
i
t
a
t
e
c
l
u
s
t
e
r
e
d
w
e
b
s
e
r
v
i
c
e
d
i
s
c
o
v
e
r
y
u
s
i
n
g
W
S
D
L
d
o
c
u
m
e
n
t
m
e
t
a
t
a
g
s
.
T
h
e
e
f
f
i
c
i
e
n
c
y
o
f
c
l
u
s
t
e
r
i
n
g
w
a
s
m
a
i
n
t
a
i
n
e
d
b
y
t
h
e
u
n
s
t
a
b
l
e
d
i
s
t
r
i
b
u
t
i
o
n
a
n
d
c
l
a
s
s
i
f
i
c
a
t
i
o
n
o
f
n
o
i
s
e
t
a
g
s
.
I
n
a
d
d
i
t
i
o
n
,
t
h
e
a
u
t
h
o
r
s
p
r
o
p
o
s
e
d
a
u
t
om
a
t
e
d
w
e
b
s
e
r
v
i
c
e
c
l
u
s
t
e
r
i
n
g
b
y
t
a
g
g
i
n
g
w
e
b
s
e
r
v
i
c
e
s
o
n
a
d
o
m
a
i
n
o
r
U
D
D
I
s
e
a
r
c
h
e
n
g
i
n
e
o
r
o
n
t
o
l
o
g
y
.
H
o
w
e
v
e
r
,
s
y
n
t
a
x
-
b
a
s
e
d
a
p
p
r
o
a
c
h
e
s
w
i
t
h
a
l
o
w
e
r
p
e
r
f
o
r
m
a
n
c
e
d
r
a
w
b
a
c
k
d
u
e
t
o
t
h
e
c
om
p
l
e
x
i
t
y
o
f
n
a
t
u
r
a
l
l
a
ng
u
a
g
e
,
a
v
a
r
i
e
t
y
o
f
s
e
m
a
n
t
i
c
a
p
p
r
o
a
c
h
e
s
a
r
e
a
l
s
o
s
u
g
g
e
s
t
e
d
.
Re
nzis
et
al
.
[
28]
ai
m
s
to
su
ggest
a
s
olu
ti
on b
ase
d
on
apply
ing
ca
se
-
base
d
reasonin
g
i
n
se
le
ct
ing
a
nd
disco
ver
i
ng
w
eb
ser
vices
ta
s
ks
.
A
s
im
il
arity
functi
on
det
erm
ines
the
scor
e
of
sim
il
ari
ty
a
m
on
g
tw
o
cases.
Fu
rt
her
m
or
e,
t
he
ap
proac
h
c
om
bin
es
the
con
ce
pts
of
ca
se
-
ba
sed
reasonin
g
(CBR
)
usi
ng
Wor
dN
et
and
distrib
utional
ly
sim
il
ar
wor
ds
us
i
ng
co
-
occurre
nces
(
DI
SC
O
)
as
a
li
ghtw
ei
gh
t
sem
antic
bas
is.
T
he
res
ult
is
case
m
anag
em
ent,
wh
ic
h
ca
n
increase
t
he
visibil
it
y
of
the
a
ppr
opriat
e
ser
vi
ces
to
accom
plish
certai
n
require
d
featur
e
s;
this
s
erv
ic
e
has
bee
n
ret
urne
d
as
a
pro
po
se
d
s
olut
ion
in
a
ppr
oxim
at
ely
90
%
of
cases.
Howe
ve
r,
th
e
exp
e
rim
ents are
pe
rf
or
m
ed
on
sm
all d
at
aset
o
f
w
e
b
se
r
vices (
62 servic
es)
.
The
w
ork
done
in
[29]
al
locat
ed
t
he
web
ser
vice
disco
ver
y
pro
blem
based
on
inte
gr
at
io
n
al
gorithm
s.
The
c
ontrib
ution
co
ns
ist
s
of
com
pu
ti
ng
t
he
best
WSs
f
ollo
wing
a
n
ou
t
rankin
g
relat
ion
s
hi
p.
T
o
e
valuate
their
appr
oach,
the
auth
or
s
us
e
d
the
O
WLS
-
TC
ver
si
on
2.2
be
nch
m
ark
.
T
he
resu
lt
s
prese
nted
are
im
pr
essive
but
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
24
, N
o.
2
,
N
ove
m
ber
2021
:
12
28
-
123
7
1230
rem
ai
n
low
in
te
r
m
s
of
pr
ec
isi
on
(
66
%
).
F
el
la
h
et
al.
[30]
,
gav
e
a
we
b
ser
vices
dis
cov
e
ry
fr
am
ework
to
perform
the
sem
antic
interoperabili
ty
of
WSs
in
a
m
ulti
-
on
tolo
gical
en
vironm
ent.
A
propose
d
al
gorit
hm
fo
r
com
pu
ti
ng
the
si
m
il
arity
of
con
ce
pts
bet
we
en
ontolo
gies
was
de
velo
pe
d,
wh
ic
h
com
pr
ise
s
a
co
m
bin
at
ion
of
diff
e
re
nt
local
si
m
il
arity
m
ea
su
res
,
ta
ki
ng
into
acc
ount
al
l
it
e
m
s
and
se
m
antic
structu
res
of
t
he
or
i
gi
n
an
d
obj
ect
ive
ont
olo
gies
.
Howe
ve
r,
the
dataset
com
po
ne
nt
is
based
on
sem
antic
ann
otati
ons
fo
r
WSDL
an
d
XML
schem
a
(
SAW
SD
L
)
la
ng
uag
e
, which
is ra
rel
y use
d.
To
so
l
ve
the
pro
blem
of
we
b
ser
vices
disc
ov
e
ry,
resea
rc
her
s
a
dopted
t
he
noti
on
of
t
he
sem
antic
web,
as
well
as
m
easur
es
of
si
m
il
arity
to
determ
ine
the
si
m
i
la
rity
bet
ween
t
he
cust
o
m
er
req
uest
and
th
e
publishe
d
WSs.
These
works
are
div
i
ded
i
nto
two
cat
e
gori
es,
as
show
n
in
Table
1
.
S
om
e
research
e
rs
hav
e
dev
el
op
e
d
thei
r
si
m
il
arity
m
e
asur
e
s;
oth
er
s
hav
e
worked
with
al
read
y
e
xisti
ng
m
easure
s.
Howe
ver,
al
l
these
pro
po
sal
a
ppr
oa
ches
rem
ai
n
lim
it
ed
becau
se
of
seve
ral
fact
or
s
li
ke
the
ch
oice
of
sim
il
ari
ty
m
easur
es
s
ui
ta
ble
for
the
w
eb
se
rv
ic
es
disco
very
pr
oble
m
and
the
pr
eci
si
on
that
can
be
c
onside
red
l
ow
f
or
the
se
ap
pro
aches
.
These
li
m
i
ta
ti
on
s
hav
e
enc
our
aged
us
to
pro
po
s
e
a
sem
antic
disco
ve
ry
we
b
se
rv
ic
es
a
ppr
oach
with
at
tra
ct
ive
pr
eci
sio
n wit
hout
f
ocu
si
ng on a speci
fie
d
si
m
il
arity
m
easur
e.
Table
1.
Functi
on
al
c
om
par
iso
n of o
ur pr
oposa
l wit
h
the
pr
e
vious st
ud
ie
s
Sy
n
tactic/S
e
m
an
ti
c
Dataset us
ed
Si
m
ila
rity
m
e
asu
re
/
m
eth
o
d
[
2
3
]
(20
0
6
)
Sy
n
tactic
-
Ter
m
f
requ
en
cy
-
in
v
erse do
cu
m
en
t f
re
q
u
e
n
cy
[
2
5
]
(20
0
8
)
Se
m
an
tic
Top
ic
Distillatio
n
(T
D
)
task
o
f
T
REC
-
2
0
0
4
W
e
b
tr
ack
M
ajo
rit
y
vo
tin
g
[
2
6
]
(20
1
0
)
Se
m
an
tic
OW
LS
-
TC
V.
2.2
W
u
&
P
al
m
e
r’
s
Si
m
ilarity
[
2
7
]
(20
1
4
)
Sy
n
tactic
STag Datas
et 1.0
Jaccard co
ef
f
icien
t
[
2
8
]
(20
1
6
)
Sy
n
tactic
an
d
Se
m
an
tic
-
Cas
e
-
b
ased
Reaso
n
in
g
[
2
9
]
(20
1
7
)
Se
m
an
tic
OW
LS
-
TC
v
ersio
n
2.2
Co
sin
e,
E
x
ten
d
ed
Jaccard,
Jen
so
n
S
h
an
o
n
,
i
n
f
o
r
m
atio
n
lo
ss
,
lo
g
ic
m
atch
in
g
.
[
3
0
]
(20
1
9
)
Se
m
an
tic
Ben
ch
m
a
rk
o
f
On
to
lo
g
y
Alig
n
m
en
t
Evalu
atio
n
Initiativ
e OA
EI
I
n
d
iv
id
u
al si
m
ila
ri
ty
m
easu
re
Ou
r
ap
p
roach
Se
m
an
tic
OW
LS
-
TC v4
.0.
Euclid
ean
d
istan
ce,
M
an
h
attan
d
istan
ce,
W
u
a
n
d
Pal
m
e
r,
Co
sin
e,
Jaccard's
Ind
ex
,
an
d
Log
ic
al
Co
rr
esp
o
n
d
en
ce
3.
PRELIM
IN
A
RIES
This
sect
io
n
pro
vid
es
the
ne
cessary
bac
kgr
ound
f
or
un
de
rstan
ding
the
rem
ai
nd
er
of
t
his
pa
pe
r,
includi
ng the
web ser
vice
disco
ver
y, t
he w
eb
se
rv
ic
e cl
ust
ering, a
nd the
web ser
vice si
m
il
arity.
3.1.
We
b ser
vice cl
ust
eri
ng
A
data
obj
ect
cl
us
te
r
ca
n
be
treat
ed
to
gethe
r
as
a
group
a
nd
seen
a
s
dat
a
com
pr
essio
n.
Ca
lc
ulati
ng
the
si
m
il
arity
betwee
n
obj
ec
ts
is
us
ually
t
he
first
ste
p
in
a
cl
us
te
rin
g
al
go
rithm
.
Pr
e
-
cl
us
te
ri
ng
ai
m
s
to
decr
ease
th
e
search
a
rea.
W
it
h
com
pu
ta
ti
on
al
ly
intensive
sem
antic
si
m
il
arity
cal
c
ulati
on
,
th
e
serv
ic
e
m
at
ching
proc
ess
can
be
m
ore
be
nef
ic
ia
l
in
a
sp
eci
fic
gro
up
tha
n
i
n
a
la
r
ge
group
of
un
relat
ed
se
rv
ic
e
s.
T
he
si
m
il
arity o
f
th
e
W
S
s is
first c
al
culat
ed,
t
hen
we
ca
n
cal
c
ulate
the d
ist
a
nce
betwee
n
the
tw
o
se
rv
ic
es
.
(
,
)
=
1
_
(
,
)
⁄
(1)
Ma
ny
al
go
rith
m
s
wer
e
propo
sed
to
cl
us
te
r
data
hav
e
rece
ntly
e
m
erg
ed.
They
can
be
cl
assifi
ed
int
o
exclusi
ve
cl
us
te
rin
g,
hiera
rc
hi
cal
cl
us
te
ring
,
ov
erla
pp
i
ng
c
lusterin
g,
an
d
pro
bab
il
ist
ic
clu
ste
rin
g
[
31
]
,
[32]
.
Our
co
ntributi
on
e
xp
l
oits
the
k
-
m
eans
al
go
rithm
;
i
t
is
a
n
exclu
sive
cl
us
te
rin
g
an
d
one
of
th
e
m
os
t
use
d
al
gorithm
s f
or
cl
us
te
rin
g.
Bas
ed on t
he si
m
ilarity
, th
e clu
ste
rin
g
al
gorit
hm
s
are
us
e
d
to
gr
oup WSs
.
K
-
m
eans
is
one
of
the
sim
pl
est
fo
rm
s
of
the
un
s
up
e
r
vised
al
go
rithm
[3
3]
us
ed
to
so
lve
cl
us
te
rin
g
pro
blem
s.
The
al
gorithm
def
ines
ce
rtai
n
dat
a
by
a
s
pecific
nu
m
ber
if
the
c
lusters
a
re
determ
in
ed
a
pri
ori
[14]
.
Desp
it
e
al
l
it
s
adv
a
ntages
[
34
]
,
k
-
m
eans
has
so
m
e
weak
ne
sses,
but
these
la
te
r
on
e
s
do
not
aff
e
ct
our
con
t
rib
ution
t
o
st
ud
yi
ng
the
s
i
m
i
la
rity
scor
e
betwee
n
co
nsum
ers'
qu
eries
and
publishe
d
W
Ss.
O
ur
a
ppro
a
c
h
consi
sts o
f usi
ng clusteri
ng t
hro
ugh
cl
assifi
cat
ion
a
nd f
il
tr
at
ion
rathe
r
tha
n
sim
il
arity.
3.2.
We
b
ser
vice simi
larit
y
Sem
antic
si
m
ilarity
is
a
m
et
r
ic
acro
s
s
a
set
of
data
ba
sed
on
th
e
sim
ilitu
de
of
t
heir
m
eanin
g.
It
pr
ese
nts
t
he
c
orres
pondence
betw
een
tw
o
ontol
og
ic
al
c
on
ce
pts
or
ta
xonom
ie
s
con
c
epts,
a
nd
it
de
fines
a
distance
am
ong
w
ords
by
st
at
ist
ic
al
m
eans.
The
sim
il
arity
betwee
n
co
nc
epts
is
a
qua
ntit
at
ive
m
easur
e
of
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
A sema
ntic w
eb
servi
ces
d
isc
overy
appro
ac
h
inte
grati
ng
mu
lt
iple si
mila
rit
y m
eas
ur
es
…
(
Mou
r
ad F
ari
ss
)
1231
inf
or
m
at
ion
de
te
rm
ined
base
d
on
t
heir
at
tri
bute
s
a
nd
relat
ion
s
.
It
has
a
ce
ntral
ro
le
i
n
a
n
en
vir
on
m
ent
s
uch
as
the
se
m
antic
web,
w
her
e
da
ta
co
m
e
fr
om
m
ul
ti
ple
so
ur
c
es
and
m
us
t
fl
exibly
be
com
bin
e
d
an
d
integrate
d.
Give
n
m
any
sim
il
arity
m
easur
es,
we
f
oc
us
on
t
he
six
m
os
t
us
ed
i
n
the
di
scov
e
ry
we
b
s
erv
ic
es
dom
ain
.
T
he
fo
ll
owin
g
detai
l
of
each
sim
i
la
rity
m
easur
e
resp
ect
i
vely
:
Eucli
dea
n
dist
ance,
Ma
nh
at
t
an
dista
nce,
Wu
a
nd
Pal
m
er,
Cosi
ne
, J
acca
r
d'
s
in
de
x,
a
nd
lo
gical
c
orres
pondence
.
3.2.1.
E
ucli
dean
dist
an
c
e
The
Eucli
dea
n
distance
is
a
geo
m
et
ric
pr
oble
m
sta
nd
ar
d
m
easur
e.
Th
e
us
ua
l
distance
betwee
n
tw
o
po
i
nts
ca
n
be
m
easur
ed
easi
ly
in
tw
o
-
or
t
hree
-
dim
ensional
sp
ace.
T
his
m
easur
e
is
co
m
m
on
ly
e
m
plo
y
ed
t
o
so
lve
cl
us
te
ri
ng
prob
le
m
s
an
d
is
c
onsidere
d
the
m
os
t
desir
able
m
at
ching
m
easur
e
wh
e
n
the
data
is
de
nse
or
con
ti
nu
ous.
It
sat
isfie
s
the
f
our
co
ndit
ion
s
a
bove
an
d
is
,
th
eref
or
e
,
a
n
acc
ur
at
e
scal
e.
T
he
Eucli
de
an
di
sta
nce
is al
so
the
stan
dard
distance
m
e
asur
e
util
iz
ed
in
the
K
-
Me
an
al
go
rithm
.
To
m
easur
e
t
he
Eucli
dea
n
dis
ta
nce
be
twee
n
two
data
el
em
e
nts
d
1
an
d
d
2
re
pr
ese
nted
by
t
he
ir
vect
or
s
1
⃗
⃗
⃗
⃗
an
d
2
⃗
⃗
⃗
⃗
res
pecti
ve
ly
. Th
e e
quat
ion i
s as
f
ollows:
(
1
⃗
⃗
⃗
⃗
,
2
⃗
⃗
⃗
⃗
)
=
√
(
∑
|
,
1
−
,
2
|
2
=
1
)
(2)
wh
e
re t
he
te
rm
set is
T
= {v
1
, ... ,v
m
}
, a
nd t
he
tf
idf
value
is ut
il
iz
ed
as term
weig
hts,
t
hat is
w
t,1=
tf
idf(
d
1
, v
).
3.
2.2. M
an
h
at
ta
n
dist
a
nc
e
The
Ma
nhat
ta
n
distance
ca
n
be
cal
culat
ed
as
the
su
m
of
the
a
bs
ol
ut
e
diff
e
ren
ce
s
betwee
n
the
Ca
rtesi
an
c
oor
din
at
es
of
tw
o
po
i
nts.
We
ca
n
say
that
it
is
the
s
um
of
the
diff
e
re
nce
between
the
c
oor
din
at
es
x
and
y
.
The
dis
ta
nce
betwee
n
t
wo
points
p
1
and
p
2
,
res
pe
ct
ively
with
the
co
ordin
at
es
(x
1
,
y
1
)
and
(x
2
,
y
2
)
,
m
easur
ed
alo
ng the
ax
e
s at th
e right a
ng
le
s
is:
ℎ
=
|
1
−
2
|
+
|
1
−
2
|
(3)
t
he
distance
M
anh
at
ta
n
is
fr
e
qu
e
ntly
us
e
d
in
integrate
d
ci
rc
uits
w
her
e
t
he
wires
run
on
ly
p
arall
el
to
the
X
or
Y
axis.
It
is
al
s
o
nam
ed
Ma
nh
at
ta
n
le
ngth
,
str
a
igh
t
distance
,
L1
distance
Mi
nkowski
or
L1
no
rm
,
ta
xi
sca
le
,
or
blo
c
k dist
ance.
3.2.3. Wu
and
Pa
lm
er
’simi
larit
y
me
a
sure
Wu
a
nd
Palm
e
r'
s
[18]
sim
il
arit
y
m
easur
e
ge
ner
al
ly
us
es
in
form
ation
f
rom
the
sh
or
te
st
path
of
tw
o
con
ce
pts,
t
he
s
pecifici
ty
or
p
r
evalence
of
t
he
two
c
on
ce
pts
i
n
the
on
t
ology
hierar
c
hy,
a
nd
their
relat
io
ns
with
oth
e
r
co
ncep
ts
.
The
auth
ors
pr
op
os
e
a
sim
il
arity
m
easur
e
to
fi
nd
th
e
m
os
t
sp
eci
fied
com
m
on
con
ce
pt
unde
rly
ing
the
two
m
easur
ed
co
nce
pts.
T
he
m
or
e
sp
eci
fic
com
m
on
con
ce
pt
pat
h
le
ng
t
h
is
com
pute
d
by
add
i
ng
the
I
S
-
A
li
nks
t
o
t
he
two
com
par
ed
con
ce
pts.
T
he
fo
ll
owin
g
e
qua
ti
on
pr
e
sent
t
he
f
or
m
ula
of
t
he
Wu
and Palm
er’
sim
il
arity:
&
(
1
,
2
)
=
2
(
1
+
2
+
2
⁄
)
(4)
wh
e
re
N
1
a
nd
N
2
are
the
nu
m
ber
s
of
IS
-
A
li
nk
s
of
C
1
a
nd
C
2
res
pecti
vely
with
t
he
m
os
t
sp
eci
fic
gen
e
ra
l
con
ce
pt
C
,
an
d
H
is
the
num
ber
of
IS
-
A
li
nks
of
C
to
the
root
of
a
n
onto
log
y.
This
m
ea
su
re
of
sim
il
ar
it
y
is
betwee
n 1 a
nd
0.
3.2.4. C
os
ine
s
im
il
arity
Cosine
sim
il
ar
it
y
is
on
e
of
the
s
uccessful
si
m
il
arity
m
ea
su
res
us
e
d
with
te
xt
doc
um
e
nts
in
m
any
inf
or
m
at
ion
ret
rievals
a
nd
cl
ust
ering
ap
plica
ti
on
s
[21]
,
[
35]
.
The
cosi
ne
si
m
il
arity
al
go
rithm
util
iz
es
the
an
gle
betwee
n
two
ve
ct
or
s
in
the
ve
ct
or
sp
ace
to
def
i
ne
the
difference
in
co
nt
ent
betwee
n
two
ve
ct
ors
[36]
.
It
is
essenti
al
ly
based
on
t
he
c
onsu
m
er'
s
pr
efe
ren
ce
a
nd
t
he
le
vel
of
va
ri
at
ion
bet
ween
prov
i
ded
W
Ss.
It
determ
ines
a
se
m
antic
WS
th
at
sequ
e
ntial
ly
conf
or
m
s
to
the
c
on
s
um
er'
s
con
te
xt
w
he
n
it
feed
s
t
he
se
m
antic
WS
back
t
o
th
e
consum
er
to
m
eet
the
con
s
um
er'
s
diff
ere
nt
needs
in
the
c
on
s
um
er'
s
con
te
xt.
S
uch
a
s
1
⃗
⃗
⃗
⃗
an
d
2
⃗
⃗
⃗
⃗
two d
oc
um
ent
s,
their
cosi
ne sim
il
arity is:
(
1
⃗
⃗
⃗
⃗
,
2
⃗
⃗
⃗
⃗
)
=
1
⃗
⃗
⃗
⃗
.
2
⃗
⃗
⃗
⃗
(
|
1
⃗
⃗
⃗
⃗
|
×
|
2
⃗
⃗
⃗
⃗
|
)
⁄
(5)
s
ince
the
co
sin
e
sim
il
arit
y
al
g
or
it
hm
con
ce
nt
rates
on
the
dif
fer
e
nce
of
vect
or
s
’
directi
o
ns
,
it
is
not
s
us
ce
ptible
to their
size.
H
ence,
c
ons
um
e
rs
m
ai
nly use it t
o determ
ine w
het
her the
W
S content is
int
eresti
ng.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
24
, N
o.
2
,
N
ove
m
ber
2021
:
12
28
-
123
7
123
2
3.2.5. J
acca
r
d
's
ind
ex
The
Jacca
r
d
I
ndex
(also
nam
ed
the
Jacca
rd
Si
m
il
arity
Coeff
ic
ie
nt)
is
a
st
at
ist
ic
utilized
to
com
par
e
the
di
ve
rsity
a
nd
sim
il
arit
y
of
the
sam
ple
set
[37]
.
T
his
I
nd
e
x
is
de
fine
d
as
present
i
n
(6),
as
the
nu
m
ber
of
sh
are
d o
bj
ect
s
div
ide
d by th
e
total
n
um
ber
of
obj
ect
s,
m
inu
s the
num
ber
of c
omm
on
o
bje
ct
s.
(
)
(
,
)
=
(
‖
‖
2
2
+
‖
‖
2
2
−
)
⁄
(6)
3.2.6. L
ogic
al
cor
resp
onden
ce
Ther
e
are
four
ty
pes
of
m
at
c
hes:
exact,
plugin,
subs
um
e
a
nd
fail
.
T
he
de
finiti
on
of
eac
h
ty
pe
is
a
s
fo
ll
ows:
E
X
A
CT
: i
f
t
he
concept R
Q
out and A
Sout a
r
e in the
sam
e o
nto
lo
gy class.
PLUGI
N
:
if
th
e
ASou
t
c
on
ce
pt
in
the
on
t
olo
gy
is
a
s
ub
cl
a
ss
of
R
Q
ou
t,
i
n
this
case,
the
RQout
co
nce
pt
is
m
or
e sp
eci
fic t
han the
searc
he
d AS
ou
t c
onc
ept.
SU
BS
UM
E
:
i
f
the
cl
ass
of
R
Qout
is
m
or
e
ge
ner
al
t
han
A
S
ou
t,
the
ASo
ut
cl
ass
of
t
he
ontolo
gy
is
a
sub
-
cl
ass of RQ
out.
FAIL
: i
f
t
her
e
is n
o
s
ub
s
um
ption
sim
il
arit
y between
R
Q
ou
t
and ASo
ut in
the
on
t
ology.
Fo
r
e
xam
ple,
t
o
com
par
e
al
l
t
he
outp
ut
co
nc
epts
of
t
he
RQ
ou
t
que
ry
wit
h
the
set
of
outp
ut
con
c
epts
of the
ASo
ut note ser
vice,
we
apply
:
(
,
)
=
(
ℎ
1
(
1
,
)
)
ℎ
1
(
,
)
=
(
ℎ
1
(
,
)
)
if
the
two
co
nc
epts
are
declar
ed
com
ple
m
entary
or
disti
nct
,
the
zero
sc
ores
are
giv
e
n
di
rectl
y.
Otherwi
se,
we
check t
he su
bsum
ption
b
el
ow and s
upply t
he
score
for
the
inter
val [0, 1]
a
ccordin
g
t
o
Ta
ble
2
.
Table
2
.
Subs
um
pt
ion
base
d
s
cor
e
assig
nm
e
nts
Relatio
n
sh
ip
Descripti
o
n
Sco
re
EXACT
RQou
t
an
d
ASo
u
t
are
eq
u
iv
alen
t
1
.0
PLUG
IN
RQou
t
is a par
en
t
o
f
ASo
u
t
0
.6
SUBSUM
E
ASo
u
t
is a su
b
class
of
RQou
t
0
.4
FAIL
No
su
b
su
m
p
tio
n
r
e
latio
n
between
RQ
o
u
t
an
d
ASo
u
t
0
4.
PROP
OSE
D APP
ROAC
H
To
s
olv
e
WSs
disc
o
ve
ry,
ou
r
proposal
i
ntr
oduces
the
no
ti
on
of
cl
us
te
r
ing
a
nd
sim
il
a
rity
scor
e
betwee
n
the
use
r'
s
req
uest
an
d
the
avail
able
serv
ic
es.
The
si
m
il
arity
m
ea
su
res
ai
m
to
c
om
pu
te
the
si
m
il
arity
scor
e
bet
wee
n
the
c
us
tom
er
r
equ
e
st
an
d
t
he
serv
ic
es
avail
a
ble
in
the
U
D
DI
re
gister.
Th
is
si
m
il
arity
be
com
es
p
o
s
s
i
b
l
e
w
i
t
h
t
h
e
s
e
m
a
nt
i
c
a
s
p
e
c
t
b
y
u
t
i
l
i
z
i
n
g
O
W
L
-
S
s
e
r
v
i
c
e
s
,
k
n
o
w
i
n
g
t
h
a
t
t
h
e
r
e
a
r
e
s
e
v
e
r
a
l
m
e
t
h
od
s
o
f
s
e
m
a
n
t
i
c
d
e
s
c
r
i
p
t
i
o
n
.
T
h
e
c
h
o
i
c
e
o
f
a
n
o
n
t
o
l
og
y
l
a
n
g
u
a
g
e
b
a
s
e
d
o
n
O
W
L
-
S
i
s
j
u
s
t
i
f
i
e
d
i
n
ou
r
p
r
e
v
i
o
u
s
w
o
r
k
[4]
,
the m
os
t st
andard
iz
e
d
a
nd p
e
rh
a
ps
t
he
m
os
t com
plete
se
m
a
ntic
W
S
d
e
pl
oy
ed
te
ch
no
l
ogy.
In
our
a
ppro
ac
h,
we
ch
os
e
t
o
util
iz
e
dif
fer
e
nt
sim
i
la
rity
m
easur
e
s
becaus
e
of
the
diff
e
r
ence
i
n
the
pr
eci
sio
n
a
nd
e
ff
ect
ive
ness
of
each
m
easur
e.
This
le
d
t
o
ap
plyi
ng
th
e
k
-
m
eans
cl
us
te
rin
g
to
filt
er
an
d
cl
assify
the
disc
ov
e
red
ser
vices
a
cc
ordin
g
to
thei
r
si
m
il
arity
scor
es
to
m
ini
m
iz
e
th
e
com
plexity
of
the
sel
ect
io
n
ph
a
se.
On
e
of
the
pr
i
ncipal
di
ff
ic
ulti
es
of
k
-
m
eans
cl
us
te
rin
g
is
to
de
fine
in
adv
a
nce
t
he
num
ber
of
cl
us
t
ers.
A
si
m
ple
strat
egy
to
achieve
t
his
is
the
el
bow
m
et
ho
d,
w
hich
con
s
ist
s
of
vary
ing
K
a
nd
f
ollow
in
g
the
e
vo
luti
on
of
t
he
intra
-
cl
ass
inerti
a.
T
he
idea
is
to
vi
su
al
iz
e
the
"
el
bow
,"
wh
e
re
the
ad
diti
on
of
a
cl
ass
does
not
corres
pond to
anyt
hing in
the
d
at
a str
uctu
rin
g.
The
ob
ta
i
ned
di
scov
e
red
WSs
belo
ng
to
t
he
cl
ass
ha
ving
th
e
cl
os
est
ce
ntr
oid
t
o
1,
wh
ic
h
m
eans
that
the
W
Ss
belo
ngin
g
to
this
cl
ass
hav
e
sim
il
ar
it
y
scor
es
glob
al
ly
c
lose
to
1.
Hen
ce,
they
w
il
l
be
the
discov
ere
d
serv
ic
es
that
a
re
the
m
os
t
sim
il
ar
to
the
se
nt
qu
e
ry.
O
ur
pro
po
se
d
ap
proach
ai
m
s
to
s
olv
e
WSs
disc
ov
e
ry
us
in
g
t
he no
ti
on
of sim
il
arit
y
and the
k
-
m
ea
ns
al
gorithm
. Th
e ste
ps
of this
appr
oac
h
ca
n be s
how
n
as
fo
ll
ow
s:
Step
1:
c
onsist
s
of
cal
culat
i
ng
t
he
six
m
easur
e
s
of
sim
i
la
rity
e
m
plo
yed
in
our
ap
pr
oach
;
E
uclide
an
distance,
Ma
nhat
ta
n
distance
,
Wu
a
nd
Palm
er'
s
si
m
i
la
rity
m
easur
e,
C
os
i
ne
sim
i
la
rity
,
Jacca
rd
I
nd
e
x,
and
Lo
gical
Corres
pondence;
bet
ween
the
c
us
t
om
er
request
a
nd
the
WSs
a
vai
la
ble
in
the
WSs
re
gistry.
T
he
se
si
m
il
arity
scor
es
will
be
saved
in
the
tem
po
rar
y
database
,
wh
ic
h
will
be
util
iz
ed
in
the
seco
nd
ste
p.
Thi
s
ph
a
se
al
lows
us
to
s
ur
m
ount
the
pro
blem
of
ch
oo
si
ng
su
it
a
ble
sim
i
la
rity
m
easur
es
for
t
he
WS
disco
ve
ry
pro
blem
.
An
d
not
f
oc
us
in
g
on
one
m
easu
re
of
sim
il
arity
can
m
ake
us
overl
ook
t
he
be
nef
it
s
of
ot
he
r
m
easur
es.
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
A sema
ntic w
eb
servi
ces
d
isc
overy
appro
ac
h
inte
grati
ng
mu
lt
iple si
mila
rit
y m
eas
ur
es
…
(
Mou
r
ad F
ari
ss
)
1233
Step
2:
aft
er
cr
eat
ing
the
data
bas
e
of
si
m
il
ar
it
y
scor
es
achieved
by
the
six
si
m
il
arit
y
m
ea
su
res
,
we
em
pl
oy
the
k
-
m
eans
al
gorithm
on
t
he
se
sco
res
t
o
i
de
ntify
the
WSs
cl
assifi
ed
i
n
di
ff
ere
nt
cl
ust
er
s
acco
rd
i
ng
to
the
degree
of
sim
i
la
rity
between
the
consum
er'
s
req
uest
an
d
the
W
Ss.
T
he
us
e
o
f
k
-
m
e
ans
cl
us
te
ri
ng
is
pr
ece
de
d by th
e sp
eci
ficat
io
n of t
he K
num
ber
of
cl
us
te
rs
ut
il
iz
ing
the
el
bow
m
et
ho
d.
Step
3:
de
fi
ne
the
gro
up
of
WSs
m
os
t
si
m
il
ar
to
the
cl
ien
t
request
am
on
g
the
cl
us
te
rs
ob
ta
ine
d
f
ro
m
the
pr
e
vious
ste
p.
The
su
it
able
cl
us
te
r
is
the
on
e
with
a
centr
oid
cl
os
e
r
to
1.
This
cl
us
te
r
m
us
t
be
the
se
t
of
WSs
disco
vered
by
the
syst
e
m
and
res
po
nd
m
os
t
eff
ect
ively
to
the
custom
er'
s
req
ue
st
based
on
the
si
m
il
arities calcu
la
te
d by the
diff
e
re
nt sim
i
lar
it
y
m
easur
es.
T
h
e
m
o
d
e
l
p
r
e
s
e
n
t
e
d
i
n
F
i
g
u
r
e
1
g
i
v
e
s
t
h
e
s
t
e
p
s
o
f
o
u
r
p
r
o
p
o
s
e
d
a
p
p
r
o
a
c
h
.
T
h
e
c
l
i
e
n
t
r
e
q
u
e
s
t
i
s
s
e
n
t
t
o
t
h
e
d
i
s
c
o
v
e
r
y
s
y
s
t
e
m
,
a
n
d
t
h
e
s
i
m
i
l
a
r
i
t
y
d
a
t
a
ba
s
e
w
i
l
l
b
e
l
o
a
d
e
d
b
y
t
h
e
s
i
m
i
l
a
r
i
t
y
s
c
o
r
e
s
c
o
m
p
u
t
e
d
f
r
om
t
h
e
s
i
x
s
i
m
i
l
a
r
i
t
y
m
e
a
s
u
r
e
s
e
m
pl
o
y
e
d
.
T
h
e
n
,
w
e
a
p
p
l
y
t
h
e
k
-
m
e
a
n
s
a
l
g
o
r
i
t
hm
t
o
g
r
o
u
p
t
h
e
W
S
s
i
n
t
o
c
l
u
s
t
e
r
s
a
n
d
d
e
f
i
n
e
t
h
e
m
o
s
t
s
i
m
i
l
a
r
c
l
u
s
t
e
r
t
o
t
h
e
c
l
i
e
n
t
r
e
q
u
e
s
t
.
T
h
i
s
a
p
p
r
o
a
c
h
a
i
m
s
t
o
i
n
c
r
e
a
s
e
t
h
e
a
c
c
u
r
a
c
y
o
f
W
S
s
d
i
s
c
o
v
e
r
y
a
n
d
a
v
o
i
d
t
h
e
p
r
o
b
l
e
m
s
e
n
c
o
u
n
t
e
r
e
d
w
i
t
h
s
om
e
s
i
m
i
l
a
r
i
t
y
m
e
a
s
u
r
e
s
u
t
i
l
i
z
i
n
g
k
-
m
e
a
n
s
c
l
u
s
t
e
r
i
n
g
o
n
t
h
e
c
om
p
u
t
e
d
s
i
m
i
l
a
r
i
t
y
s
c
o
r
e
s
.
T
h
e
s
e
a
d
v
a
n
t
a
g
e
s
w
i
l
l
i
n
f
l
u
e
n
c
e
t
h
e
s
e
l
e
c
t
i
on
p
h
a
s
e
a
n
d
t
h
e
c
om
p
o
s
i
t
i
o
n
o
f
W
S
s
.
Figure
1
.
The
pro
po
se
d WS
disco
ver
y a
ppr
oach
5.
RESU
LT
S
A
ND
D
IS
C
USS
ION
5.1.
Dataset
an
d
exper
im
e
nt
al
set
u
p
T
h
e
t
e
c
h
n
o
l
o
g
i
c
a
l
e
n
v
i
r
o
nm
e
nt
e
m
pl
o
y
e
d
t
o
i
m
p
l
e
m
e
n
t
o
u
r
p
r
o
p
o
s
e
d
a
p
p
r
o
a
c
h
a
n
d
o
t
h
e
r
m
e
c
h
a
n
i
s
m
s
a
n
d
i
t
s
e
v
a
l
u
a
t
i
o
n
i
s
t
h
e
J
A
V
A
l
a
n
g
u
a
g
e
.
T
h
e
e
x
p
e
r
i
m
e
n
t
a
l
c
om
p
u
t
a
t
i
o
n
s
r
a
n
o
n
W
i
n
d
o
w
s
1
0
I
n
t
e
l
C
o
r
e
i
5
C
P
U
(
2
.
6
G
H
z
)
a
n
d
8
G
B
o
f
R
A
M
.
O
u
r
a
p
p
r
o
a
c
h
w
a
s
e
v
a
l
u
a
t
e
d
e
x
p
e
r
i
m
e
n
t
a
l
l
y
u
t
i
l
i
z
i
n
g
t
h
e
s
e
r
v
i
c
e
r
e
p
o
s
i
t
o
r
y
o
f
t
h
e
w
e
l
l
-
k
n
o
w
n
O
W
L
S
-
T
C
v
4
.
0
t
e
s
t
c
o
l
l
e
c
t
i
o
n
.
T
h
e
r
e
a
r
e
1
0
8
3
S
W
S
s
w
r
i
t
t
e
n
i
n
O
W
L
-
S
1
.
1
a
n
d
4
2
q
u
e
r
i
e
s
.
T
h
e
r
e
a
r
e
n
i
n
e
s
e
r
v
i
c
e
d
o
m
a
i
n
s
:
e
d
u
c
a
t
i
o
n
,
f
o
o
d
,
m
e
d
i
c
i
n
e
,
t
r
a
v
e
l
,
e
c
o
n
om
y
,
c
o
m
m
u
n
i
c
a
t
i
o
n
,
w
e
a
p
o
n
r
y
,
g
e
o
g
r
a
p
h
y
,
a
n
d
s
i
m
u
l
a
t
i
o
n
.
T
a
b
l
e
3
s
h
o
w
s
t
h
e
d
e
t
a
i
l
s
o
f
O
W
L
S
-
T
C
v
4
.
0
.
S
o
m
e
s
e
r
v
i
c
e
s
s
e
e
m
i
n
m
o
r
e
t
h
a
n
o
n
e
c
a
t
e
g
o
r
y
.
H
e
n
c
e
,
i
f
w
e
c
o
n
s
i
d
e
r
j
u
s
t
t
h
e
f
i
r
s
t
o
c
c
u
r
r
e
n
c
e
o
f
e
a
c
h
s
e
r
v
i
c
e
,
t
h
e
n
um
be
r
o
f
s
e
r
v
i
c
e
s
i
s
1
0
8
3
;
i
f
w
e
c
o
n
s
i
d
e
r
r
e
p
e
t
i
t
i
o
n
s
a
c
r
o
s
s
d
i
f
f
e
r
e
n
t
c
a
t
e
g
o
r
i
e
s
,
t
h
e
n
um
b
e
r
o
f
s
e
r
v
i
c
e
s
i
s
1
1
4
0
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
24
, N
o.
2
,
N
ove
m
ber
2021
:
12
28
-
123
7
1234
Tabl
e
3
. Det
ai
ls o
f O
WLS
-
T
C v
4.0
Do
m
ain
s Rain
f
all
d
ata
Nu
m
.
o
f
services
Nu
m
.
o
f
r
eq
u
ests
Edu
catio
n
286
6
Medical Ca
re
73
1
Fo
o
d
34
1
Tr
av
el
197
6
Co
m
m
u
n
icatio
n
59
2
Econ
o
m
y
395
12
W
eapo
n
40
1
Geo
g
raph
y
60
10
Si
m
u
latio
n
16
3
5.2.
E
va
lu
at
i
on
m
etrics
The
perform
ance
of
t
he
pro
pose
d
ap
proac
h
is
evaluated
ut
il
iz
ing
tree
m
e
tric
s:
pr
eci
sio
n,
recall
,
a
nd
F
-
m
easur
e
[38]
.
These
m
et
ri
cs
are
t
he
m
os
t
com
m
on
ly
us
ed
t
o
e
valuat
e
the
perform
a
nce
of
WS
discov
e
ry
appr
oach
es
.
Pre
ci
sion
is
the
nu
m
ber
of
accurate
res
ults
div
ide
d
by
the
num
ber
of
al
l
resu
lt
s
returne
d,
wh
il
e
recall
is
the
nu
m
ber
of
acc
urat
e
resu
lt
s
div
i
ded
by
the
nu
m
ber
of
re
su
lt
s
to
be
retu
r
ne
d,
as
presente
d
in
(
7)
and (
8)
.
(
)
=
⁄
(7)
(
)
=
ℎ
⁄
(8)
Ther
e
is
a
n
inve
rse
co
rr
el
at
io
n
betw
een
the
two
eq
uatio
ns
,
wh
e
re
it
is
po
s
sible
to
increas
e
on
e
valu
e
at
the
co
st
of
re
du
ci
ng
the
oth
e
r.
I
n
general,
they
a
re
no
t
t
reated
se
par
at
el
y.
T
he
F
-
m
easur
e
c
om
bin
es
pr
eci
sio
n
a
nd
recall
into
a
sing
le
e
qu
at
io
n,
i.e.,
a
com
posit
e
har
m
on
ic
m
ean,
re
du
c
es
the
eff
ect
s
of
la
rge
unusual
values
an
d
am
plifie
s
the
im
pacts
of
sm
a
ll
on
es,
a
s
sh
ow
n
in
(
9).
As
a
weig
hte
d
un
a
nim
ou
s
m
e
asur
e
,
the F
-
m
easur
e
is m
uch
m
or
e crit
ic
al
than
eit
he
r pr
eci
sio
n or
recall
sep
a
ratel
y.
−
=
2
∗
∗
+
(9)
5.3.
Results
’ di
scussion
F
i
r
s
t
,
w
e
p
e
r
f
o
r
m
e
d
t
h
e
e
x
p
e
r
i
m
e
nt
s
w
i
t
h
a
q
u
e
r
y
o
f
t
h
e
"
c
a
r
_
p
r
i
c
e
_
s
e
r
v
i
c
e
"
d
a
t
a
s
e
t
f
r
om
t
h
e
"
M
e
di
c
a
l
C
a
r
e
"
d
om
a
i
n
,
a
n
d
w
e
o
b
t
a
i
n
e
d
t
h
e
r
e
s
u
l
t
s
d
i
s
p
l
a
y
e
d
i
n
T
a
b
l
e
4
.
T
h
e
l
a
t
t
e
r
s
h
o
w
s
t
h
e
s
c
o
r
e
s
c
a
l
c
ul
a
t
e
d
b
y
t
h
e
f
i
r
s
t
s
t
e
p
f
o
r
s
om
e
c
o
n
t
r
o
l
s
e
r
v
i
c
e
s
.
A
l
l
t
h
e
s
i
m
i
l
a
r
i
t
y
m
e
a
s
u
r
e
s
u
s
e
d
h
a
v
e
d
i
f
f
e
r
e
n
t
s
c
o
r
e
s
f
o
r
e
a
c
h
s
e
r
v
i
c
e
,
w
h
i
c
h
e
x
p
l
a
i
n
s
t
h
e
c
h
o
i
c
e
o
f
a
p
e
r
f
e
c
t
s
i
m
i
l
a
r
i
t
y
m
e
a
s
u
r
e
a
m
o
n
g
t
h
e
d
i
f
f
e
r
e
n
t
m
e
a
s
u
r
e
s
e
m
p
l
o
y
e
d
.
A
s
s
h
o
w
n
b
y
t
h
e
"
E
u
c
l
i
d
e
a
n
d
i
s
t
a
n
c
e
"
,
t
h
e
s
i
m
i
l
a
r
s
e
r
vi
c
e
i
s
"
B
o
o
k
N
o
n
M
e
d
i
c
a
l
T
r
a
n
s
p
o
r
t
"
w
i
t
h
a
s
c
or
e
o
f
0
.
9
4
,
b
u
t
f
o
r
t
h
e
"
W
u
a
n
d
P
a
l
m
e
r
"
s
i
m
i
l
a
r
i
t
y
m
e
a
s
u
r
e
,
i
t
i
s
"
B
o
o
k
M
e
d
i
c
a
l
T
r
a
n
s
p
o
r
t
"
w
i
t
h
a
s
c
o
r
e
o
f
0
.
9
9
.
We
use
d
t
he
k
-
m
eans
m
et
ho
d
to
determ
ine
the
ser
vices
w
it
h
the
best
si
m
il
ar
scor
es
to
rem
ov
e
this
var
ia
ti
on,
us
i
ng
the
diff
e
re
nt
sco
res
of
the
si
m
il
arity
m
ea
su
res
.
Ne
ver
th
el
ess,
be
f
or
e
ut
il
iz
ing
the
k
-
m
eans
m
et
ho
d,
we
m
us
t
de
fine
the
nu
m
ber
of
K
cl
asses
util
iz
ing
the
el
bo
w
m
et
ho
d
.
Fig
ure
2
in
dicat
es
that
the
act
ual num
ber
of clusters
is
K
=4.
The
WSs
will
be
cl
assifi
e
d
i
nt
o
f
our
cl
asses.
By
cal
culat
in
g
th
e
ce
ntr
oid
s
of
eac
h
cl
ass
,
we
ac
hieve
the
res
ults
disp
la
ye
d
in
Ta
ble
5,
w
hich
re
presents
the
ce
nt
ro
ids
of
eac
h
cl
ass
an
d
the
nu
m
b
er
of
WSs.
T
he
disco
ver
e
d
W
Ss
are
the
se
rvi
ces
in
the
cl
ass
whose
ce
ntr
oi
d
is
cl
os
est
to
1.
Acc
ordi
ng
to
our
a
ppr
oac
h,
we
can
ass
um
e
that
the
m
os
t
sim
il
ar
cl
ass
of
W
S
s
is
cl
ass
3,
w
hic
h
c
onta
ins
72
WSs
and
gi
ves
a
98.
63
%
pr
eci
sio
n de
pe
nd
i
ng on t
he
st
ud
ie
d
do
m
ai
n.
Table
4
. Det
ai
ls
sim
il
arit
y scor
es
for WSs
dis
cov
e
re
d
f
r
om
t
he
m
edical
care dom
ai
n
Service na
m
e
Euclid
ean
d
istan
ce
Manh
attan
d
istan
ce
W
u
and
Pal
m
e
r’
si
m
ilarit
y
Measu
re
Co
sin
e
si
m
ilarit
y
The
Jaccard
in
d
ex
Log
ical
Co
rr
esp
o
n
d
en
ce
Bo
o
k
No
n
Medical
Tr
an
sp
o
rt
0
.94
0
.86
0
.90
0
.94
0
.79
1
h
ik
in
g
_
d
estin
atio
n
0
.73
0
.54
0
.59
0
.89
0
.58
0
.4
g
etAd
d
ressOf
Locatio
n
0
.30
0
.36
0
.00
0
.01
0
.27
0
Bo
o
k
MedicalTra
n
sp
o
rt
0
.86
0
.84
0
.99
0
.86
0
.87
0
.6
au
to
_
p
ricec
o
lo
r
0
.76
0
.63
0
.58
0
.83
0
.84
0
.6
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
A sema
ntic w
eb
servi
ces
d
isc
overy
appro
ac
h
inte
grati
ng
mu
lt
iple si
mila
rit
y m
eas
ur
es
…
(
Mou
r
ad F
ari
ss
)
1235
Figure
2
.
Th
e
e
lbow m
et
h
od
s
howing t
he op
t
i
m
al
K
Table
5
. T
he
C
entr
oid
s
of Cl
asses O
btaine
d
Clas
s
Variance int
ra
-
cl
ass
Nu
m
b
e
r
o
f
service
s
Cen
tral
serv
ice
1
0
.43
3
278
title_
co
m
ed
y
f
il
m
2
0
.57
4
489
au
to
_
tech
n
o
l
o
g
y
3
0
.75
6
72
ho
spita
l_
bio
psy
4
0
.29
3
246
g
o
v
ern
m
en
t
m
iss
ile
weapo
n
_
fun
d
i
n
g
I
t
w
a
s
t
e
s
t
e
d
f
o
r
s
e
v
e
r
a
l
q
u
e
r
i
e
s
a
n
d
W
S
s
t
o
e
v
a
l
u
a
t
e
o
u
r
a
p
p
r
o
a
c
h
,
a
n
d
i
t
s
r
e
s
u
l
t
s
w
e
r
e
c
om
p
a
r
e
d
t
o
t
h
e
o
t
h
e
r
s
i
m
i
l
a
r
i
t
y
m
e
a
s
u
r
e
s
.
E
x
p
e
r
i
m
e
n
t
s
p
r
e
s
e
n
t
e
d
i
n
F
i
g
u
r
e
3
s
h
o
w
t
h
a
t
o
u
r
a
p
p
r
o
a
c
h
i
s
m
o
r
e
p
e
r
f
o
r
m
e
d
t
h
a
n
o
t
h
e
r
m
e
t
h
o
d
s
f
o
r
W
S
d
i
s
c
o
v
e
r
y
.
T
o
a
n
a
l
y
z
e
t
h
e
s
e
r
e
s
u
l
t
s
,
w
e
c
a
n
s
a
y
t
h
a
t
t
h
e
o
t
he
r
s
i
m
i
l
a
r
i
t
y
m
e
a
s
u
r
e
s
c
a
l
c
u
l
a
t
e
t
h
e
s
i
m
i
l
a
r
i
t
y
b
e
t
w
e
e
n
t
h
e
r
e
q
u
e
s
t
a
n
d
t
h
e
W
S
s
.
H
o
w
e
v
e
r
,
o
u
r
a
p
p
r
o
a
c
h
b
e
n
e
f
i
t
s
f
r
om
t
h
e
i
r
r
e
s
u
l
t
s
t
o
b
u
i
l
d
a
m
o
r
e
r
e
a
l
s
e
r
v
i
c
e
s
i
m
i
l
a
r
i
t
y
v
a
l
u
e
.
M
o
s
t
w
o
r
k
s
i
n
W
S
d
i
s
c
o
v
e
r
y
u
s
e
d
t
h
e
s
i
m
i
l
a
r
i
t
y
m
e
a
s
u
r
e
s
a
l
r
e
a
dy
m
e
nt
i
o
n
e
d
t
o
s
h
o
w
t
h
e
e
f
f
i
c
i
e
n
c
y
a
n
d
e
f
f
e
c
t
i
v
e
n
e
s
s
o
f
o
u
r
a
p
p
r
o
a
c
h
.
W
e
c
om
p
a
r
e
d
t
h
e
r
e
s
u
l
t
s
o
b
t
a
i
n
e
d
w
i
t
h
t
he
o
t
h
e
r
s
i
m
i
l
a
r
i
t
y
m
e
a
s
u
r
e
s
s
t
u
d
i
e
d
;
t
h
e
r
e
s
ul
t
s
s
h
o
w
t
h
e
s
u
c
c
e
s
s
o
f
o
u
r
a
p
p
r
o
a
c
h
b
y
i
n
c
r
e
a
s
i
n
g
t
h
e
p
r
e
c
i
s
i
on
(98%
)
a
n
d
r
e
c
a
l
l
(
9
5
%
)
,
a
s
w
e
l
l
a
s
t
h
e
F
-
M
e
a
s
u
r
e
(
9
6
%
)
o
f
t
h
e
W
S
s
d
i
s
c
o
v
e
r
e
d
,
c
om
p
a
r
e
d
t
o
t
h
e
a
p
p
r
o
a
c
h
e
s
t
h
a
t
u
s
e
t
h
e
d
i
f
f
e
r
e
n
t
s
i
m
i
l
a
r
i
t
y
m
e
a
s
u
r
e
s
.
T
h
i
s
s
u
c
c
e
s
s
i
s
d
u
e
t
o
u
s
i
n
g
t
h
e
k
-
m
e
a
n
s
a
l
g
o
r
i
t
hm
o
n
t
h
e
s
i
m
i
l
a
r
i
t
y
s
c
o
r
e
s
o
b
t
a
i
n
e
d
b
y
m
a
n
y
s
i
m
i
l
a
r
i
t
y
m
e
a
s
u
r
e
s
,
t
h
e
n
t
h
e
e
l
i
m
in
a
t
i
o
n
o
f
p
o
s
s
i
b
l
e
d
i
s
a
d
v
a
n
t
a
g
e
s
f
o
r
a
n
y
m
e
a
s
u
r
e
.
A
l
t
h
o
u
g
h
,
d
e
c
r
e
a
s
e
t
h
e
n
um
b
e
r
o
f
W
S
s
d
i
r
e
c
t
e
d
a
t
t
h
e
W
S
s
e
l
e
c
t
i
o
n
p
h
a
s
e
.
(a)
(b)
(c)
Figure
3.
Disc
ov
e
ry
re
su
lt
s
of six
sim
i
la
rity
m
easur
es
with
our
a
ppr
oach
;
(
a) F
-
m
easur
e,
(
b) r
ec
al
l,
(c)
preci
sion
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
24
, N
o.
2
,
N
ove
m
ber
2021
:
12
28
-
123
7
1236
On
the o
ther
ha
nd,
to
validat
e
ou
r
r
esults
ag
ai
ns
t
oth
er
a
ppro
ac
hes
in
the
li
te
ratur
e. W
e
i
m
ple
m
ented
two
m
et
ho
ds
previ
ou
sly
m
ent
ion
e
d
in
the
r
el
at
ed
w
orks
sec
ti
on
[
29
]
,
[30]
.
Fo
r
t
he
ap
proa
ch
pro
pose
d
in
[29]
,
we
util
iz
ed
the
best
interval
judge
d
by
the
auth
or
s
.
Fig
ure
4
sh
ows
a
com
par
ison
of
pr
eci
sio
n,
reca
ll
,
and
F
-
m
easur
e
f
or
the
m
entione
d
m
et
ho
ds.
T
he
obta
ined
out
com
es
con
fi
rm
the
ef
fecti
ve
ne
ss
of
our
a
pp
ro
ac
h
com
par
ed
t
o
th
e stu
died
a
ppr
oa
ches.
Figure
4.
Eval
uation
m
et
rics for our a
ppr
oa
ch
6.
CONCL
US
I
O
N
To
im
pr
ov
e t
he
d
isc
overy p
r
ocess of
S
WSs, we
use
d
a
n
O
WL
-
S and s
ix
si
m
il
arity
m
easur
es
, n
am
e
ly
Eucli
dea
n
dist
ance,
Ma
nhat
ta
n
distance
,
Wu
a
nd
Palm
er,
Cosi
ne
sim
il
arity,
Jacca
rd
in
de
x,
an
d
l
og
ic
al
m
at
ching
,
bet
ween
t
he
WSs
store
d
in
t
he
reg
ist
ry
a
nd
t
he
consum
er'
s
req
ue
st
.
Th
en
,
we
ap
ply
the
k
-
m
ean
s
cl
us
te
rin
g
al
gorithm
in
order
to
ob
ta
i
n
the
m
os
t
si
m
il
ar
serv
ic
es
t
o
t
he
consum
er'
s
request,
not
base
d
on
a
sing
le
sim
il
arity
m
easur
e
(the
us
ual
case
of
the
Eucli
dea
n
distance)
bu
t
ba
sed
on
the
six
si
m
il
arit
y
m
ea
su
re
s
us
e
d.
T
he
car
ri
ed
out
e
xp
e
rim
ents
hav
e
show
n
the
hi
gh
perform
ance
of
this
a
ppro
ac
h
in
te
rm
s
of
pr
eci
sion
(98%
)
a
nd
re
cal
l
(95%),
as
well
as
th
e
F
-
Me
asu
re
(96%).
We
ass
um
e
that
a
pra
ct
ic
al
W
S
s
disco
ver
y
appr
oach
can
help
cust
om
ers
m
ake
their
distribu
te
d
a
ppli
cat
ion
s
m
or
e
e
ff
ic
ie
nt.
Mo
re
ov
e
r,
ot
her
si
m
il
arity
m
easur
es ca
n
e
nr
ic
h
the
prop
ose
d
a
ppr
oach d
ependin
g o
n
th
e cust
om
ers’
nee
ds
or the e
xp
ect
ed
ef
fici
enc
y.
REFERE
NCE
S
[1]
N.
Bal
aji,
S.
R.
Murugai
y
an,
M.
S.
Sale
embasha
,
R.
Baskar
an
,
a
nd
P.
Dhava
che
l
van,
“
Enha
nc
ements
for
UD
DI
using
Us
er
Prefe
ren
t
ia
l
W
eb
Service
Selecti
on
Model
base
d
on
SL
A,”
Int.
J.
Eng
.
Technol.
,
vol
.
5,
no.
5
,
pp.
4057
-
4067
,
2013.
[2]
C.
Dhasarathan
,
V.
Thi
rum
al,
a
nd
D.
Ponnuran
gam,
“
Data
priv
acy
br
ea
ch
pre
v
ent
ion
fr
amework
for
the
cl
oud
servic
e
,
”
Se
cur.
Comm
un.
Net
wo
rks
,
vol
.
8
,
pp
.
9
82
-
1005
,
2015
,
doi:
10
.
1002/sec.105
4
.
[3]
P.
Khutade
and
R.
Phalni
kar
,
“
QoS
base
d
web
servic
e
discover
y
using
oo
con
ce
pts,
”
Int.
J
.
A
dv.
Technol.
En
g.
Re
s.
,
vol
.
2
,
no
.
6,
pp
.
81
-
86
,
20
12.
[4]
M.
Fariss,
N.
E
l
Alla
l
i,
H.
As
a
i
di,
and
M.
B
ell
ouki,
“
Revi
ew
of
Ontolog
y
B
a
sed
Approac
hes
for
W
eb
Service
Discove
r
y
,
”
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
A
d
v
a
n
c
e
d
I
n
f
o
r
m
a
t
i
o
n
T
e
c
h
n
o
l
o
g
y
,
S
e
r
v
i
c
e
s
a
n
d
S
y
s
t
e
m
s
,
pp.
78
-
87,
2019,
doi
:
10
.
10
07/978
-
3
-
030
-
11
914
-
0_8
.
[5]
L.
Li
and
I.
Horrocks,
“
A
Software
Fram
ework
for
Matc
hm
aki
ng
Based
on
S
emanti
c
W
eb
T
ec
hnolog
y
,
”
Int
.
J.
El
e
ct
ron.
Comm
er.
,
vol. 8, no. 4, pp. 39
-
60,
2004
,
doi
:
10
.
1080/1
0864415.
2004.
1
1044307.
[6]
S.
Y.
Li
n
,
C.
H.
La
i
,
C.
H.
W
u,
and
C.
C
.
Lo
,
“
A
trustworth
y
Q
oS
-
base
d
col
l
ab
ora
ti
v
e
filter
ing
appr
oac
h
for
we
b
servic
e
discov
er
y
,
”
J. Sy
st
.
So
ft
w
.
,
vol. 93, pp. 21
7
-
228,
2014
,
doi
:
10.
1016
/j.jss.2
014.
01.
036
.
[7]
P.
Pleba
ni
and
B
.
Pernici,
“
URBE:
W
eb
Service
Ret
ri
eva
l
B
ase
d
on
Sim
il
ari
t
y
Ev
al
ua
ti
on,
”
IEEE
Tr
ansacti
ons
on
Knowle
dge
and
Data
Engi
n
ee
rin
g
,
vol
.
21
,
no
.
11
,
pp
.
1629
-
1642
,
Nov.
2009
,
do
i:
10.
1109/T
KD
E.2009.35
.
[8]
G.
Cassar,
P.
Barna
ghi
and
K.
Moess
ner
,
“
Probabil
isti
c
Matc
hm
ak
ing
Methods
for
Autom
at
ed
Servic
e
Discove
r
y
,
”
IEE
E
Tr
ansacti
o
ns on
Serv
i
ce
s C
omputing
,
vo
l. 7, no. 4, pp. 654
-
6
66,
Oct
.
-
De
c. 20
14,
doi
:
10
.
1109
/T
SC
.
2013.
28
.
[9]
M.
Klusch,
B.
Fries,
and
K.
Sy
c
ara,
“
OW
LS
-
MX
:
A
hy
brid
Sem
ant
ic
W
eb
servic
e
m
at
chm
ake
r
for
OW
L
-
S
servic
es,
”
J. W
e
b
Semant.
,
vo
l. 7
,
no
.
2
,
pp
.
121
–
133,
2009
,
doi
:
1
0.
1016/j.webse
m
.
2008.
10.
001.
[10]
M
.
K
l
u
s
c
h
a
n
d
P
.
K
a
p
a
h
n
k
e
,
“
O
W
L
S
-
M
X
3
:
A
n
a
d
a
p
t
i
v
e
h
y
b
r
i
d
s
e
m
a
n
t
i
c
s
e
r
v
i
c
e
m
a
t
c
h
m
a
k
e
r
f
o
r
O
W
L
-
S
,
”
i
n
T
h
i
r
d
I
n
t
e
r
n
a
t
i
o
n
a
l
W
o
r
k
s
h
o
p
o
n
S
e
r
v
i
c
e
M
a
t
c
h
m
a
k
i
n
g
a
n
d
R
e
s
o
u
r
c
e
R
e
t
r
i
e
v
a
l
i
n
t
h
e
S
e
m
a
n
t
i
c
W
e
b
(
S
M
R
2
)
,
v
o
l
.
5
2
5
,
2
0
0
9
.
[11]
W
.
Lu,
Y.
Ca
i,
X.
Che,
and
Y.
Lu,
“
Joint
sem
ant
i
c
sim
il
a
rity
assess
m
ent
with
raw
cor
pus
and
struct
ure
d
on
tolo
g
y
for
sem
ant
ic
-
ori
ent
ed
servi
ce
dis
cove
r
y
,
”
Pe
rs
on
al
and
Ubiquit
ous
Computing
,
vol.
20,
no.
3,
pp
.
311
-
323,
2016,
doi:
10
.
1007/s00
779
-
016
-
09
21
-
0.
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
A sema
ntic w
eb
servi
ces
d
isc
overy
appro
ac
h
inte
grati
ng
mu
lt
iple si
mila
rit
y m
eas
ur
es
…
(
Mou
r
ad F
ari
ss
)
1237
[12]
M.
Fang,
D.
Wa
ng,
Z.
Mi
,
and
M.
S.
Obaida
t,
“
W
eb
servic
e
d
iscove
r
y
utilizin
g
logi
cal
rea
son
ing
and
sem
ant
i
c
sim
il
ari
t
y
,
”
Int.
J.
Comm
un.
Syst
.
,
vol. 31, no. 10, pp. 1
-
13,
2018,
doi:
10
.
1002/dac
.
3561.
[13]
R.
A.
H.
M
.
Ru
pasingha
,
I.
Pa
i
k
,
and
B
.
T
.
G.
S.
Kum
ara
,
“
C
al
cu
la
t
ing
web
s
erv
ice
sim
il
ar
ity
using
ontol
og
y
le
arn
ing
with
m
ac
hine
learni
n
g,
”
2015
IEE
E
Inte
rnational
Confe
renc
e
on
Computati
onal
Inte
ll
ig
ence
and
Computing
R
ese
arch
(
ICCIC)
,
2015,
pp
.
1
-
8
,
doi
:
10.
1109
/ICCIC
.
2015.
7435686
.
[14]
H.
Fr
igui
,
“
Clus
te
ring:
Algori
th
m
s
and
Applic
ations,”
2008
Fi
rs
t
Workshops
on
Image
Proce
ss
ing
Theory,
Tools
and
Applications
,
2008
,
pp
.
1
-
11
,
doi: 10.
1109
/IPTA.
2008.
47437
93
.
[15]
N.
Arunac
ha
la
m
,
A.
Am
utha
n,
C.
Kav
y
a
,
M.
S
har
m
il
la,
K.
Us
h
ana
ndhin
i
and
M.
Shanm
u
ghapr
i
y
a,
“
A
surve
y
on
web
servic
e
cl
u
steri
ng,
”
2017
I
nte
rnational
Co
nfe
renc
e
on
Computati
on
of
P
ower,
Ene
rgy
I
nformation
and
Comm
uinc
ati
on
(
ICCPE
IC)
,
2017,
pp
.
247
-
252
,
doi:
10
.
1109/IC
CP
EIC.
2017.
82
90371
.
[16]
S.
Chandra
seka
r
an,
V.
B.
Sriniv
asa
n,
and
L.
Pa
r
thi
ban
,
“
Eff
ic
i
en
t
W
eb
Servic
e
Discove
r
y
and
Selecti
on
Model
,
”
Int.
J. F
u
tur. Revolut.
Comput
.
S
ci
.
Comm
un.
En
g.
,
vol. 4, no. 2,
2018.
[17]
K.
Venka
t
ac
ha
lam
and
N.
K.
Ka
rthi
ke
y
a
n
,
“
Eff
e
ct
iv
e
Feat
ur
e
Se
t
Selecti
on
and
Cent
roid
Cl
assif
ie
r
Algori
thm
for
W
eb
Servic
es
Discove
r
y
,
”
Ind
onesian
Journal
of
El
e
ct
rica
l
E
ngine
ering
and
Computer
Sci
en
ce
,
vo
l.
5,
no
.
2,
pp.
441
-
450
,
Fe
b.
2017
,
doi
:
10
.
11591/i
jeec
s.v5
.
i2.
pp441
-
450.
[18]
Z.
W
u
and
M.
Palmer,
“
Verb
Sem
ant
ic
s
and
Le
x
ic
a
l
,
”
32nd
Annu.
Me
et
.
Assoc.
Comput.
Li
nguist.
,
199
4,
pp.
133
-
138
,
doi
:
10.
3115
/98173
2.
981751
.
[19]
J.
W
u
and
Z
.
W
u,
“
Sim
il
ari
t
y
-
b
a
sed
W
eb
service
m
at
chmaking,”
2005
IEE
E
Int
ernati
onal
Con
fe
r
enc
e
on
Serv
ices
Computing
(
SCC'05)
Vol
-
1
,
vo
l. 1
,
2005,
p
p.
287
-
294,
doi
:
10
.
11
09/SCC.2005.
93
.
[20]
F.
Chen,
C.
Lu,
H.
W
u,
and
M.
Li
,
“
A
sem
ant
ic
sim
il
ari
t
y
m
ea
sure
int
egr
at
ing
m
ult
iple
concept
u
al
relati
onships
for
web
service
d
isc
over
y
,
”
E
xpe
rt
S
yst.
Appl.
,
vol. 6
7,
pp
.
19
-
31
,
20
17,
doi
:
10
.
1016
/j
.
eswa
.
2016.
09
.
028.
[21]
C.
Corley
and
R
.
Mihal
c
ea,
“
Mea
suring
the
sem
ant
i
c
sim
il
ari
t
y
of
te
xts,”
in
Proce
ed
ings
of
the
ACL
workshop
on
empiric
al
modeling of
semant
ic equivalence and e
ntai
lment
,
2005
,
pp.
13
-
18
,
doi
:
1
0.
3115/1631862.1631865.
[22]
H.
Fetha
l
la
h,
M
.
Moham
m
ed,
and
B.
Am
ine
,
“
Sem
ant
ic
W
eb
servic
e
Discov
e
r
y
Based
on
Fu
zzy
Dom
ina
te
d
Score
s,”
in
Pro
ce
ed
ings
of
the
Inte
rnational
Confe
renc
e
on
Inte
lligen
t
Infor
mation
Proce
ss
i
ng,
Sec
uri
ty
an
d
Adv
anc
ed
Comm
unic
ati
on
,
2015
,
pp.
1
-
6
,
doi
:
10
.
1145/2816839.
2
8168
81.
[23]
N.
Kokash,
“
A
Com
par
ison
of
W
eb
Servic
e
Int
erf
ace
Sim
il
ari
t
y
Mea
sures
,”
Proce
ed
ings
of
the
2006
conf
ere
nc
e
on
STAIRS
2006:
P
roce
edi
ngs o
f th
e
Thir
d
Start
ing AI
R
ese
arche
rs
'
Symposium
,
200
6,
pp
.
220
-
231
.
[24]
M
.
M
o
n
t
a
g
u
e
a
n
d
J
.
A
.
A
s
l
a
m
,
“
C
o
n
d
o
r
c
e
t
f
u
s
i
o
n
f
o
r
i
m
p
r
o
v
e
d
r
e
t
r
i
e
v
a
l
,
”
P
r
o
c
e
e
d
i
n
g
s
o
f
t
h
e
e
l
e
v
e
n
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
I
n
f
o
r
m
a
t
i
o
n
a
n
d
k
n
o
w
l
e
d
g
e
m
a
n
a
g
e
m
e
n
t
,
2
0
0
2
,
p
p
.
5
3
8
–
5
4
8
,
d
o
i
:
1
0
.
1
1
4
5
/
5
8
4
8
7
9
.
5
8
4
8
8
1
.
[25]
M.
Fara
h
and
D.
Vande
rpoote
n
,
“
An
outra
nking
appr
oac
h
for
inf
orm
at
ion
ret
ri
ev
al
,
”
Information
Re
tri
ev
al
,
vol.
1
1,
no.
4
,
pp
.
315
-
3
34,
2008
,
doi
:
10
.
1007/s10791
-
00
8
-
9046
-
z.
[26]
H.
Fetha
llah,
C.
Am
ine
,
and
B.
Am
ine
,
“
Automa
te
d
D
iscove
r
y
of
W
eb
Servic
es :
an
Inte
rfa
ce
Matc
h
ing
Approac
h
Based
on
Sim
il
ari
t
y
Me
asure
,
”
i
n
Proce
edi
ngs
o
f
the
1st
Inte
rna
ti
onal
Confe
ren
ce
on
Inte
l
li
g
ent
Semanti
c
We
b
-
Serv
ices and Ap
pli
cations
,
2010
,
pp.
1
-
4,
doi: 10.1145/1874590.1
874603.
[27]
J.
W
u,
L.
Ch
en,
Z.
Zh
eng,
M
.
R.
L
y
u
,
and
Z.
W
u,
“
Cluste
ring
W
e
b
servic
es
to
fa
cilit
a
te
serv
ice
dis
cove
r
y
,
”
Knowl
.
Inf.
S
yst.
,
vo
l. 38
,
no
.
1
,
pp
.
207
-
229,
2014
,
doi
:
1
0.
1007/s10115
-
0
13
-
0623
-
0.
[28]
A
.
D
e
R
e
n
z
i
s
,
M
.
G
a
r
r
i
g
a
,
A
.
F
l
o
r
e
s
,
A
.
C
e
c
h
i
c
h
,
a
n
d
A
.
Z
u
n
i
n
o
,
“
C
a
s
e
-
b
a
s
e
d
R
e
a
s
o
n
i
n
g
f
o
r
W
e
b
S
e
r
v
i
c
e
D
i
s
c
o
v
e
r
y
a
n
d
S
e
l
e
c
t
i
o
n
,
”
E
l
e
c
t
r
o
n
i
c
N
o
t
e
s
i
n
T
h
e
o
r
e
t
i
c
a
l
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
3
2
1
,
2
0
1
6
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
e
n
t
c
s
.
2
0
1
6
.
0
2
.
0
0
6
.
[29]
H
.
F
e
t
h
a
l
l
a
h
,
S
.
M
.
I
s
m
a
i
l
,
M
.
M
o
h
a
m
e
d
a
n
d
T
.
Z
e
y
n
e
b
,
“
A
n
o
u
t
r
a
n
k
i
n
g
m
o
d
e
l
f
o
r
w
e
b
s
e
r
v
i
c
e
d
i
s
c
o
v
e
r
y
,
”
2
0
1
7
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
M
a
t
h
e
m
a
t
i
c
s
a
n
d
I
n
f
o
r
m
a
t
i
o
n
T
e
c
h
n
o
l
o
g
y
(
I
C
M
I
T
)
,
2
0
1
7
,
p
p
.
1
6
2
-
1
6
7
,
d
o
i
:
1
0
.
1
1
0
9
/
M
A
T
H
I
T
.
2
0
1
7
.
8
2
5
9
7
1
1
.
[30]
A.
Fell
ah,
M.
Malki
,
and
A.
El
ci,
“
A
sim
il
ari
t
y
m
e
asure
acros
s
ontol
ogie
s
for
W
eb
servic
es
discove
r
y
,
”
Inte
rnational
Jo
urnal
of
In
formation
Te
chnol
ogy
and
We
b
Engi
n
e
ering
,
vo
l. 11, n
o.
1
,
pp
.
22
-
43
,
2016
.
[31]
J.
Han
,
M.
Kam
ber
,
J
.
Pei
,
Data
mining
con
ce
pts
and
technique
s
.
San
Franc
isco
,
US
A:
Moraga
Kaufm
an,
20
1
1.
[32]
A.
H.
Nasut
ion,
Y.
Murak
ami,
and
T
.
Ishid
a,
“
Gene
rating
sim
ilarity
cl
ust
er
of
I
ndonesia
n
la
ngu
age
s
with
sem
i
-
supervise
d
c
lust
eri
ng,
”
In
te
rnati
onal
Journal
of
El
e
ct
rica
l
and
C
omputer
Engi
n
e
ering
,
vo
l.
9,
no
.
1,
pp.
531
-
538
,
2019,
doi
:
10
.
11
591/i
jece
.
v9i1
.
p
p531
-
538.
[33]
P.
M.
Prihat
ini,
I.
K.
G.
D.
P
utra
,
I
.
A.
D.
Giria
nt
ari
,
and
M.
Sudarm
a,
“Complet
e
aggl
o
m
era
ti
ve
hi
erar
c
h
y
document’s
cl
us
te
ring
b
ase
d
on
fuz
z
y
Luhn’s
gi
bbs
la
te
n
t
dir
ic
h
le
t
al
lo
cation,”
I
nte
rnational
Jou
rnal
of
E
le
c
tric
a
l
and
Computer
E
ngine
ering
,
vol
.
9,
no
.
3
,
pp
.
210
3
-
2111,
2019
,
d
oi:
10
.
11591/ije
c
e.
v9i3
.
pp2103
-
2
111.
[34]
D.
S.
Ma
y
l
awa
t
i
,
T
.
Pri
at
na
,
H.
Sug
il
ar,
and
M
.
A.
Ramdhani
,
“
Data
sci
enc
e
f
or
digi
t
al
cultur
e
improvem
ent
i
n
highe
r
edu
cation
using
K
-
m
ea
ns
cl
uster
ing
and
text
anal
y
tics,”
In
te
rnational
Jour
nal
of
E
lectric
a
l
and
Computer
Engi
ne
ering
,
vol
.
10
,
no
.
5
,
pp
.
4
569
-
4580,
2020
,
doi: 10.
11591
/
ijece
.
v10i5
.
pp456
9
-
458
0.
[35]
L
.
M
u
f
l
i
k
h
a
h
a
n
d
B
.
B
a
h
a
r
u
d
i
n
,
“
D
o
c
u
m
e
n
t
C
l
u
s
t
e
r
i
n
g
U
s
i
n
g
C
o
n
c
e
p
t
S
p
a
c
e
a
n
d
C
o
s
i
n
e
S
i
m
i
l
a
r
i
t
y
M
e
a
s
u
r
e
m
e
n
t
,
”
2
0
0
9
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
C
o
m
p
u
t
e
r
T
e
c
h
n
o
l
o
g
y
a
n
d
D
e
v
e
l
o
p
m
e
n
t
,
2
0
0
9
,
p
p
.
5
8
-
6
2
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
C
T
D
.
2
0
0
9
.
2
0
6
.
[36]
W.
B.
Zu
lfi
k
ar,
M.
Irfa
n
,
M.
Gh
ufron,
Jum
adi
,
a
nd
E.
Firm
ans
y
a
h,
“
Marke
tplac
e
aff
iliates
pote
n
tial
ana
l
y
sis
usin
g
cosine
sim
il
arit
y
and
vision
-
base
d
page
segm
ent
a
ti
on,
”
Bul
l
et
in
o
f
El
e
ct
rica
l
Enginee
ring
and
Info
rm
ati
cs
,
vol.
9,
no.
6
,
pp
.
2492
-
2498,
2020
,
d
oi
: 10.11591/
e
ei.v9i
6.
2018.
[37]
P.
D.
Ibnugra
ha,
L.
E.
Nugroho
,
and
P.
I.
Santos
a,
“
An
appr
oac
h
for
risk
esti
m
at
i
on
in
informati
o
n
sec
urity
using
te
xt
m
ini
ng
and
ja
ccard
m
et
hod,
”
Bulletin
of
Elec
tri
cal
Engi
n
eer
ing
and
Inform
ati
cs
,
vol
.
7,
no.
3,
p
p.
393
-
3
99,
2018,
doi
:
10
.
11
591/e
e
i.
v7i3
.
847
.
[38]
A.
Altha
f
Ali
an
d
R.
M.
Shafi
,
“
Te
st
-
re
tri
ev
al
fr
a
m
ework:
Perform
anc
e
profiling
and
te
sting
w
eb
sea
rch
eng
ine
on
non
factoi
d
qu
eri
es,
”
Indon
esi
an
Journal
o
f
El
e
ct
rica
l
Eng
i
nee
ring
and
C
omputer
Scienc
e
,
vo
l
.
14,
no
.
3,
pp.
1373
-
1381
,
2019,
doi
:
10
.
11
591/i
jeec
s.v14
.
i
3.
pp1373
-
1381.
Evaluation Warning : The document was created with Spire.PDF for Python.