TELKOM
NIKA
, Vol.12, No
.4, Dece
mbe
r
2014, pp. 10
96~110
4
ISSN: 1693-6
930,
accredited
A
by DIKTI, De
cree No: 58/DIK
T
I/Kep/2013
DOI
:
10.12928/TELKOMNIKA.v12i4.413
1096
Re
cei
v
ed Se
ptem
ber 11, 2014; Revi
se
d No
vem
ber
16, 2014; Accepted Novem
ber 30, 20
14
Intelligent Interface for a Knowledge-based System
N
y
oman Karna*
1
, Iping Supriana
2
, Ulfa Maulidev
i
3
Sekol
ahT
eknik
Elektro dan Inf
o
rmatika, In
stitut T
e
knolo
g
i
B
and
un
g, Indon
esia
Ganesh
a
10, B
and
un
g, +
62-2
2
-25
009
35
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: bogi
@stude
n
t
s.itb.ac.id
1
, iping@stei.itb.ac.id
2
, ulfa@stei.it
b.ac.id
3
A
b
st
r
a
ct
Every knowledge-based system
has its
own
know
ledge for
m
alis
m
dependi
ng
on the pr
oblem
that
nee
ds so
lvin
g, goa
l to b
e
ac
hiev
ed, a
nd
propos
ed s
o
lu
ti
o
n
. T
h
is
mea
n
s
the kn
ow
ledg
e conta
i
n
ed i
n
th
e
system w
ill differ from on
e system to
anoth
e
r. This also me
ans that th
i
s
know
ledg
e cann
ot be used
b
y
anoth
e
r syste
m
, w
h
ich
in
tur
n
me
ans
every
system
must s
t
art w
i
th a l
ear
nin
g
p
has
e ri
g
h
t at the
b
egi
n
n
in
g.
One of the sol
u
tions to ov
erc
o
min
g
this pro
b
le
m is
pr
ovi
d
i
ng a u
n
ifi
ed
mode
l that
can
accept a
ll type
s of
know
led
ge, w
h
ich gu
arante
e
s
automat
ic inte
raction b
e
tw
een the know
led
ge-b
a
sed syst
ems. Interacti
o
n in
this p
aper
is
d
e
fine
d as
kno
w
ledge
shar
in
g, inte
gr
atio
n,
and
transfer fr
om on
e syste
m
to
an
other.
This
researc
h
prov
i
des a
mo
del
and c
ond
uctsa
test on inte
r
a
ction ca
pa
bil
i
t
y. It w
ill help
to acceler
a
tethe
establ
ish
m
e
n
t of a new
know
ledg
e-bas
ed sy
stem b
e
c
aus
e i
t
does not ne
e
d
know
led
ge i
n
itiali
z
a
t
i
o
n
.
Ke
y
w
ords
: u
n
ifie
d kn
ow
le
dge r
epr
esent
ation, s
hari
n
g
k
now
le
dge, know
led
ge int
egrati
on, tran
sfer
kn
o
w
l
e
dg
e
1. Introduc
tion
Knowle
dge i
s
a prop
ositio
n
that contain
s
facts
an
d is
definitive, whi
l
e rep
r
e
s
entat
ion is a
relation b
e
tween two do
mains, the symbol and
what the sy
mbol rep
r
e
s
ents. Knowl
e
dge
rep
r
e
s
entatio
n is a field of study
that
explore
s
sym
bol formali
s
m
,
which is u
s
ed to represent
kno
w
le
dge [1
]-[2]. Knowle
dge itself ha
s many inte
rpretation
s a
n
d
so fa
r there have be
en
no
agre
ed d
e
finitions fo
r kno
w
ledge. Althou
gh there ar
e
no definitive
definition
s
for kno
w
led
ge,
we
can lo
okat se
veral a
s
pe
cts, machin
e lea
r
ning,
expe
rt system
s, and
kno
w
led
ge
manag
eme
n
t. In
machi
ne lea
r
ning, inform
ation is acq
u
ired a
nd
re
tained(sto
r
ed
) for future
recall to obtain
kno
w
le
dge f
r
om the
existi
ng info
rmatio
n. In an
ex
pe
rt syste
m
, inf
o
rmatio
n i
s
a
c
qui
red
from
an
expert'
s
kno
w
ledge
and
ret
a
ined fo
r futu
re
re
call to
ob
tain the exp
e
rt's kno
w
led
g
e
.
In kno
w
le
dg
e
manag
eme
n
t, a large
amo
unt of kno
w
l
edge i
s
also
basi
c
ally
stored in i
n
form
a
t
ion fact type
s,
whi
c
h will be
easily re
call
e
d
to expand the user'
s
kno
w
led
ge.
Every knowl
e
dge re
pre
s
e
n
t
ation need
s a spe
c
ific lan
guag
e to pro
v
ide an optimal way
torep
r
e
s
ent a
symbol. Th
e
symbol it
self
is ne
ede
d to
rep
r
e
s
ent th
e kn
owl
edg
e
in su
ch
a way
that it can be
easily un
de
rstood
by a si
mple ma
chin
e, that is, a p
r
og
ram. La
ng
uage
will p
r
o
v
ide
an effective way to repre
s
e
n
t kno
w
led
g
e
usi
ng three a
s
pe
cts [1], wh
ich are as foll
ows.
1.
Syntax, to d
e
fine ho
w th
e formali
s
m
of kn
o
w
led
g
e
representati
on form
s a senten
ce with
clea
r an
d sta
ndard st
ru
ctu
r
e by buil
d
in
g it with
logi
cal symbol
s (pun
ctuatio
n,
con
n
e
c
tives,
variable
s
) an
d non-l
ogi
cal
symbol
s (fun
ction an
d pre
d
icate
s
)
2.
Semantic, to
define
ho
w th
e form
alism
o
f
kn
o
w
le
dge
repre
s
e
n
tation
form
s a
sent
ence
with a
stru
cture that can b
e
und
erstood th
roug
h
:
a. interpretation
b. denotatio
n
c.
sat
i
sf
a
c
t
i
on
3.
Pragmati
c
, to
define h
o
w
the formali
s
m of kn
owl
e
dge
rep
r
e
s
en
tation form
s
a mea
n
ingful
sente
n
ce.
Lang
uage i
s
use
d
to re
prese
n
t kn
owle
dge that is
d
e
cla
r
ative an
d ha
s a spe
c
ific a
n
d
definitive me
aning. A
de
cl
arative
s
ente
n
ce
that fulfill
s all
three
aspect
s
(syntax
,
sem
antic, a
nd
prag
matic) can be
used t
o
esta
blish knowl
edge
re
pre
s
entatio
n
usin
g logi
c-b
a
se
d form
ali
s
m,
su
ch a
s
first
-
order lo
gic.
Beside
s logi
c-ba
sed
fo
rma
lism, we
can
also choo
se
to use othe
r
approa
che
s
,
like frame
-
based a
nd
also
rul
e
-
b
a
s
ed
app
ro
aches. Ea
ch
of the p
r
eviously
mentione
d types of forma
lism ha
s its ow
n be
nefits and disa
dvantage
s, dep
endin
g
on the
conte
n
t and
what the kno
w
led
ge will b
e
use
d
for [3].
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Intelligent Interface for a K
nowl
edge
-b
a
s
ed System
(Nyom
an Bogi
Aditya Karna
)
1097
Knowle
dge repre
s
e
n
tation
is a bran
ch
of artifi
cial intelligen
ce, whi
c
h is a field
of study
that explores ho
w to
rep
r
ese
n
t inform
ation that
is
a
c
qu
ir
ed
from a
n
y
wh
er
e in
a fo
r
m
a
t
th
a
t
can
be
und
erstood by
a
pro
g
ram. A
p
r
og
ram
he
re
refe
rs to a
kno
w
l
edge
-ba
s
e
d
System, such
a
s
a machi
ne le
arnin
g
syste
m
or an expe
rt system
. Th
e kno
w
le
dge
rep
r
e
s
entatio
n itself can al
so
be de
scrib
e
d
as a
kn
owle
dge mo
del, b
e
ca
use it
can
explain the
model of the
kno
w
le
dge, t
h
e
syntax an
d t
he
sema
ntics of th
e info
rmation.
Wh
en
we
ch
oo
se th
e m
ode
l of kno
w
led
geit
depe
nd
s on three a
s
p
e
ct
s.
1.
Problem, wha
t
is the proble
m
to be addressed
?
2.
Goal, wh
at sh
ould the kno
w
led
ge fulfil?
3.
Propo
se
d sol
u
tion, how d
o
e
s the kno
w
le
dge solve the
proble
m
?
From th
e pro
b
lem a
s
pe
ct
point of view,
the mod
e
l of
kno
w
le
dge
can be
se
para
t
ed into
deci
s
io
n ma
king, the
re
co
mmend
er,
an
d the
hum
an
life en
han
ce
r. From th
e
deci
s
io
n ma
king
perspe
c
tive, the kn
owle
dg
e model is si
g
n
ificantly
relat
ed to data mining. From th
e recomme
nd
er
perspe
c
tive, the kno
w
led
g
e
model i
s
si
gnifica
ntly
rel
a
ted to the e
x
pert sy
stem. In human
life
enha
ncement
, the kno
w
le
dge m
odel
is sig
n
ifica
n
tly related
to
kn
owle
dge
man
ageme
n
t. Fro
m
the pro
p
o
s
ed
solutio
n
poi
nt of view, the model
of knowl
edge
ca
n also be
ap
proa
ch
ed u
s
i
n
g
machi
ne lea
r
ning an
d the
expert syste
m
along
wi
th
an app
rop
r
i
a
te method o
f
reasoning.
The
method of reasonin
g
itself is comp
ri
sed of
a de
cisi
on tree,
Bayesian, ru
le-ba
s
e
d
, ba
ck
prop
agatio
n, a su
ppo
rt vector ma
chine,
and an
asso
ciation rul
e
[4]-[5]. Each of these propo
se
d
method
s hav
e different be
nefits, depe
n
d
ing on the p
r
opo
sed
soluti
on offered.
The contextu
al differen
c
e
s
between t
hese
thre
e asp
e
ct
s re
su
lt in every machi
n
e
learni
ng o
r
e
x
pert syste
m
usin
g differe
n
t
kno
w
led
ge repre
s
e
n
tation
, depen
ding
on the p
r
obl
e
m
,
goal, and pro
posed sol
u
tio
n
involved. For example,
to solve a problem rel
a
te
d to knowl
e
d
g
e
from an
exp
e
rt, the kn
owledge m
odel
will be e
s
t
abl
ishe
d u
s
ing
a de
cisi
on tree that provide
s
attributes to d
e
termin
e whi
c
h p
a
th to ta
ke until a
sol
u
tion o
r
re
co
m
m
endatio
n ha
s be
en
rea
c
h
ed.
On the other
hand, to solv
e a proble
m
related
to que
stion and a
n
sweri
ng, the knowl
edge m
o
del
cho
s
e
n
is a rule-b
ased tab
l
e that provid
es attrib
utes t
o
determi
neth
e
nea
re
st answer th
at can
be
given. The
s
e
two example
s
dem
on
strate
that t
here is
no gen
eri
c
or unified kn
owledge mo
del to
be used in m
u
ltiple ca
se
s (a multi-propo
sed
solutio
n
).
2. Related Works
A kno
w
le
dge
model
or
kn
owle
dge
rep
r
ese
n
tation i
s
how
we
can
define a
form
ula that
has the
ability to describe t
he
knowledge withi
n
. Th
e
formula must
consi
s
t of a tuple that cont
ain
at least three
c
omp
one
nts.
1. Knowle
dge
at
om
2. Rule
3
.
R
e
la
tion
A kno
w
led
ge
atom, as the
first co
mpo
n
e
n
t of a kno
w
l
edge m
odel
o
f
formalism, e
x
plains
the knowl
edge entity itself
in t
he
simplest form. T
h
is
will ensure
that a program
will be abl
e
to
unde
rsta
nd the kn
owl
edg
e and ru
n a comp
utation
on it. Every
kno
w
le
dge at
om also h
a
s an
optional ad
di
tional attribut
e to explain a sp
e
c
ific behavio
r of the kno
w
led
ge to help the
comp
utation.
Rule
s, a
s
the
se
co
nd
com
pone
nt, provi
de a
list of i
n
stru
ction
s
tha
t
can
be
used
to
con
d
u
c
t inferences from th
e inform
ation
store
d
in
the
tuple. A rule i
t
self ca
n only
be u
s
ed a
s
an
inferen
c
e ex
clusively withi
n
the tuple. The re
sult
of
this infere
nce is a link th
at conn
ect
s
one
tuple with a
nother. A rel
a
tion, on th
e other
han
d, as the thi
r
d compo
n
e
n
t, describ
es an
interconn
ecti
on b
e
twe
en t
uple
s
. It provides the
list
of all th
e
con
n
e
c
tion
s from
a
particula
r tupl
e.
The lin
k itself only de
scrib
e
s a
line th
at dra
w
s an im
a
g
inary inte
rco
nne
ction b
e
twee
n two tu
p
l
es.
The result of all the interco
nnected tupleswill be a mesh net
wo
rk called a semantic network.
Up
till
n
o
w, rese
arch on knowl
edge
-b
a
s
ed
sy
st
em h
a
s fo
cu
se
d o
n
the a
ppli
c
at
ion level,
whi
c
h is ho
w to establish
a kno
w
led
g
e
-
based sy
st
e
m
and utilize
it for a speci
f
ic purp
o
se. This
kno
w
le
dge
-b
ase
d
system
will an
swer a
spe
c
ific p
r
obl
em in a spe
c
i
f
ic domain of
kno
w
le
dge.
1.
Do
cume
nt cl
assificatio
n
, whe
r
e
a kn
o
w
led
ge-
ba
se
d sy
stem i
s
assign
ed
a t
a
sk to
ret
r
ieve
document
s from a
spe
c
ifi
c
sou
r
ce, co
ndu
ct in
form
ation extra
c
ti
on fro
m
the
document to
obtain the me
ta data, and create a
classi
fication u
s
ing
the meta data
[6]-[12]
2.
Real time i
n
formatio
n, wh
ere a
kn
owl
e
dge-ba
sed
system acqui
res
kno
w
le
dg
e from a
real
time sou
r
ce with sp
atial in
formation [13]
-[20]
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 12, No. 4, Dece
mb
er 201
4: 109
6 – 1104
1098
3.
Knowle
dge
m
anag
ement
u
t
ilization,
whe
r
e
a
kno
w
le
d
ge-b
a
sed
sy
stem is built
a
nd m
ana
ged
fo
r
s
p
ec
ia
l pur
p
o
s
e
[2
1
]
-
[
26
]
4.
Multilingual i
n
formatio
n, whe
r
e a
kn
owle
dge
-ba
s
ed
sy
stem acq
u
ire
s
kn
owle
dge
fro
m
multilingual source,
incl
udi
ng
tacit kn
owl
edge [27]-[2
9
]
.
The
req
u
ire
m
ents
and
ide
a
s
fo
r b
u
ilding
a unifie
d
m
o
d
e
l to a
c
comm
odate th
e a
b
il
ity for
kno
w
le
dge t
r
ansfe
r, sha
r
in
g, and
integ
r
ation b
e
twee
n two
sy
stem
s h
a
ve al
rea
d
y provo
k
ed
lo
ng
discu
ssi
on
s [30]. Ho
weve
r, there a
r
e
sti
ll a few
re
cen
t
studie
s
that
hav
e cond
uct
ed research
on
the co
re of a
kno
w
le
dge
-b
ase
d
syste
m
, which is
the
kno
w
le
dge re
pre
s
entatio
n itself. Most of the
recent pap
ershave involved
rese
arch o
n
the appli
c
atio
n level,as de
scrib
ed ab
ove.
To accom
m
o
date kn
owl
e
dge tra
n
sfe
r
, sha
r
ing,
an
d integration
ability, this resea
r
ch
focu
se
s on
semantic net
work form
alism
.
There
a
r
e t
w
o type
s of
semantic net
work ap
pro
a
ch
es
[31]-[37].
1.
Static kn
owl
e
dge
rep
r
e
s
en
tation, wh
ere
the net
wo
rk i
s
static
(pred
e
fined) an
d b
u
ilt to solve
a
spe
c
ific p
r
obl
em
2.
Dynami
c
kno
w
led
ge rep
r
e
s
entatio
n, wh
ere the
network is dyn
a
mi
c an
d built to
solve multi
-
probl
em
s.
Static kno
w
l
edge
re
presentati
on
prov
ides ea
se
of
use a
n
d
ea
se
of buil
d
in
g. Static
kno
w
le
dge
repre
s
e
n
tation
is
ea
sy to
build
be
cau
s
e it u
s
e
s
two-ph
ase m
e
thod, training
an
d
testing. In the
training
pha
se, a sem
antic netwo
rk
i
s
b
u
ilt usin
g all t
he availabl
e
node
s (tu
p
le
). In
the testing p
hase, the se
mantic
net
wo
rk p
r
op
oses
a probl
em an
d it must pro
v
ide a solutio
n
or
recomme
ndat
ion a
c
cordin
gly. One
exa
m
ple of
stati
c
kno
w
le
dge
re
pre
s
e
n
tation i
s
kno
w
le
dge
ontology. Dy
namic kno
w
l
edge
re
pre
s
e
n
tation, on th
e
othe
r h
and,
provid
es flexibility in sol
u
tion
finding, b
e
ca
use
the
se
m
antic fa
cto
r
i
s
b
u
ilt w
hen
a p
r
oble
m
i
s
propo
sed
to
the n
e
two
r
k. In
dynamic
kno
w
led
ge
rep
r
e
s
entatio
n, the
network i
s
re
built when
a
new no
de
(tu
p
le) is integ
r
a
t
ed
into the net
work.
One
exa
m
ple of dyna
mic
kno
w
led
ge re
present
ation is
COK
B
(Co
m
putati
onal
Obje
ct Knowl
edge Ba
se
).
Both app
roa
c
hes
above,
static and
dyn
a
mic
kn
owle
dge
rep
r
e
s
en
tation use th
e sa
me
tuple formul
ation, as de
scri
bed in form
ul
a 1.
KM =
{KA, R, Rule}
(1)
Whe
r
e:
-
KM mean
s
the knowl
e
d
ge mo
del o
r
kno
w
led
g
e
rep
r
e
s
entati
on for a
sp
ecific dom
ai
n
kno
w
le
dge
-
KA means the kno
w
led
ge atom, which descri
b
e
s
th
e
simple
st and
smalle
st form of knowle
d
ge,
su
ch a
s
axio
m or a co
nce
p
t
-
R mean
sthe
relation bet
we
en KMs (int
e
r
-KM) o
r
between KAs (int
ra-KM)
-
Rule means a rule that explicitly
store
s
the formula fo
r infere
nce.
2.1
Static Kn
o
w
l
e
dge Re
presen
ta
tion
Static kno
w
l
edge
rep
r
e
s
entation de
scrib
e
s
a se
mantic n
e
twork
com
p
ri
se
d of node
s
(KM) an
d lin
ks that
con
n
e
ct the no
de
s. One ex
am
ple of static
kno
w
le
dge repre
s
e
n
tation
is
kno
w
le
dge
o
n
tology. In the kn
owl
edg
e
ontology
a
pproach, kn
owl
e
dge i
s
represented a
c
cordi
ng
to the domai
n cha
r
a
c
teri
st
ics
of the kn
owle
dge
itsel
f. A study on cre
a
ting the
relation
between
kno
w
le
dge m
odel an
d kno
w
led
ge ontol
ogy [22] is de
sc
ribe
d in Fig
u
re 1. Fig
u
re
1 explain
s
th
at,
from a kno
w
l
edge mo
del point of view, both the domain and task kno
w
le
dge
model
sa
re rel
a
ted
to kno
w
ledg
e atom and
its relation,
while the reasonin
g
kn
owle
dge mo
del is rel
a
te
d to
rule
sfrom
a
kn
owle
dge ont
o
l
ogy point of view.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Intelligent Interface for a K
nowl
edge
-b
a
s
ed System
(Nyom
an Bogi
Aditya Karna
)
1099
Figure 1. Rel
a
tion betwee
n
Knowle
dge
Model an
d Knowl
edge
On
tology
2.1 D
y
namic
Kno
w
l
e
dge
Repr
esen
ta
tion
In dynami
c
kno
w
le
dge
repre
s
e
n
tation
, the com
p
o
s
ition in th
e
kno
w
le
dge
model i
s
slightly different from
that
with
static
kn
owle
dge
re
prese
n
tation. T
he
kno
w
le
dg
e atom
is call
ed
a
comp
utationa
l object (com
-obje
c
t) an
d is the
simpl
e
st and sm
all
e
st kn
owle
dg
e entity[31].
It
comp
ri
se
s four co
mpo
nent
s, attribute
s
,
functio
n
s, fact
s, and rules
(Figure 2).
Figure 2. Co
mputational
Obje
ct as the
Knowled
ge
Atom
Whe
r
e:
-
Attributes me
an a list of attribute
s
co
rrespondi
ng to the obje
c
t
-
Functio
n
s m
e
an com
putati
onal interrelat
ions b
e
twe
e
n
attributes
-
Fact
s mean
a gro
up of propertie
s
o
r
e
v
ents rel
a
ti
ng
to the object
;
this includ
e
s
11 types
of
facts, su
ch a
s
object type, obje
c
t definiti
on, obje
c
t si
milarity, object depend
en
cy, etc.
-
Rules
mean rules
for inferenc
e from fac
t
s
From
the
com-o
b
je
ct (concept)
entity, a
kn
owl
e
dge
rep
r
e
s
e
n
tation i
s
b
u
ilt usi
n
g
Figure 3.
Figure 3. Knowled
ge Represe
n
ta
tion for Computatio
n
a
l Obje
ct
The co
ncept contai
ns
a cl
ass
of
co
m-o
b
ject
s a
nd it
s rel
a
tion i
s
drawn
u
s
ing
hi
era
r
chy
and
relatio
n
,
whe
r
e thi
s
rel
a
tion in
clud
es the o
per
ator and
fu
nctio
n
. Rule
s co
ntain
instructio
ns or
guidelin
es th
at can
be u
s
ed fo
r inferences fr
om t
he con
c
ept. With the dyn
a
mic
kno
w
le
dg
e
rep
r
e
s
entatio
n approa
ch, usin
g a com
p
utational obje
c
t mean
s that the network i
s
built usin
g an
inferen
c
e
rule
to create
a h
i
era
r
chical rel
a
tion from th
e con
c
e
p
t. The network can then b
e
u
s
ed
to answer a
problem li
ke an expert
would. Th
e
links that interconn
ect
con
c
e
p
ts in the
comp
utationa
l obje
c
t co
nta
i
n the di
re
ction; for ex
ampl
e a "IS-A" explain
s
that on
e co
ncept is
a
membe
r
of
anothe
r
con
c
ept. Thi
s
di
rection
explai
ns th
e
relati
on b
e
twe
en
one
co
ncept
and
anothe
r. Fig
u
r
e
4 d
e
mon
s
t
r
ates fou
r
co
mputational
o
b
ject
s that
bu
ild the
network hi
erarchy a
n
d
the link directi
on.
Knowledge
Model
DomainK
nowled
ge
Model
Task
Knowled
g
e
Model
Reasonin
g
Knowled
g
e
Model
Knowledge
Atom
Relation
Rule
Knowledge
Ontolo
gy
Attributes
Functions
Facts
Rules
Co
m
-
Object
Concept
Hierar
chy
Relations
Operators
Functions
Rules
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 12, No. 4, Dece
mb
er 201
4: 109
6 – 1104
1100
Figure 4. Knowled
ge Mod
e
l
Relation in
Comp
utation
a
l Obje
ct
Beside the
computation
a
l obje
c
t, there i
s
also anoth
e
r
example fo
r dynamic
kno
w
led
ge
rep
r
e
s
entatio
n, whi
c
h is t
he rel
a
tion
model o
r
Re
la-Mo
del [35]
. The Rel
a
-Model b
u
ild
s the
sema
ntic net
work by assi
milating the human m
e
th
od of storin
g
information
and ho
w hu
mans
retrieve the
stored info
rma
t
ion. Several
definition
s
for Rela
-Mod
el
aregive
n u
s
in
g formula
2 to
formula 5. Fo
rmula 2 d
e
fin
e
s the relatio
n
betwe
en co
nce
p
ts u
s
ing
a list of rules.
(C, R, R
u
les
)
(2)
Whe
r
e:
-
C is a g
r
oup
of con
c
ept
s d
e
fined by
C = {
C
1, C2,
…}
(3)
This ha
s
spe
c
ific an attrib
utes-val
ue
s combinatio
n a
c
cordi
ng to the usa
ge of the con
c
e
p
t.
-
R is a bin
a
ry relation b
e
tween con
c
ept
s, defined by
R = {
R
1, R2,
…}
(4)
-
Rule
s mea
n
s
a set of rule
s
for infere
nce purp
o
ses
{f1, f2, …,fp}
{fp+1, fp+
2
, …, fq}
(5)
The
esta
blishment
of th
e kno
w
led
g
e
stru
ctu
r
e
(d
efining every
rel
a
tion R betwe
en
con
c
e
p
ts) i
s
achi
eved u
s
i
ng iterate
d
inferen
c
e fo
r
every po
ssi
bl
e con
c
e
p
t, starting fro
m
the
con
c
e
p
t that has
a rul
e
fo
rthe ne
ar-sol
ution of
the
probl
em give
n. A comp
arative analysi
s
of
these
thre
e
model
s
(kn
o
wle
dge
ont
ology, comp
ut
ational o
b
j
e
ct, an
dRela
-
Mod
e
l)
ca
n
be
explained in
Table 1.
Table 1. Co
m
parative Anal
ysis of the Th
ree Mo
del
s
Method/
Feature
Kno
w
ledge Ontolog
y
Computational Object
Rela-Model
Kno
w
ledge
Atom
Kno
w
ledge atom
as a
referenc
e
Kno
w
ledge atom
as a
foundation to dev
elop a
hierarchical structure
Kno
w
ledge atom
as a
foundation to dev
elop a
hierarchical structure
Relation
There a
r
e fixed
relationsbet
w
een
know
ledge
atom
s
Relation exists between
know
ledge
atom
s in a vector
format
Relation exists between
know
ledge
atom
s in a binar
y
format
Inference
Rule is predefine
d
Rule is adaptive, based on the
problem
Rule is adaptive, based on the
problem
Relatio
n
Ru
le
C
once
p
t
Relatio
n
Ru
le
C
once
p
t
Relatio
n
Ru
le
C
once
p
t
Relatio
n
Ru
le
C
once
p
t
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Intelligent Interface for a K
nowl
edge
-b
a
s
ed System
(Nyom
an Bogi
Aditya Karna
)
1101
3. Rese
arch
Metho
d
This
re
sea
r
ch an
swers a
comm
on p
r
o
b
lem a
bout
knowl
edge
sh
aring
in a
kn
owle
dge
-
based sy
ste
m
, in terms
of how two knowl
edge
-
b
a
s
ed system
s can sha
r
e
th
eir
kno
w
led
g
e
so
that one sy
stem ca
n re
use anot
he
r
system'
s
kno
w
l
edge. Thi
s
will ensu
r
e a
new
kno
w
le
d
ge-
based
syste
m
do
es not
n
eed to
pe
rform kno
w
led
g
e
initialization,
and
in
stead
use
s
kn
owl
e
dge
acq
u
isitio
n from anoth
e
r
system. To find a sol
u
tion
for this p
r
o
b
le
m, this re
sea
r
ch i
s
cond
ucted
by carryingo
u
t
a literature
studyof
text books, pap
ers, and the Inter
net. As a result of this study,
the write
r
propo
se
s a me
thod that incorpo
r
at
e
s
ca
pability into a kno
w
led
g
e
-
based sy
ste
m
,
a
cap
ability to ensu
r
e kno
w
le
dge integ
r
atio
n, kno
w
led
g
e
shari
ng, and
kno
w
le
dge transfe
r.
4. Results a
nd Discu
ssi
on
An idea
ab
o
u
t providi
ng
a unifie
d
kn
owle
dge
re
prese
n
tation to
en
sure
kno
w
led
g
e
transfe
r,
sha
r
ing, and
inte
gration
bet
ween t
w
o
syst
ems
ha
s al
re
ady bee
n p
r
o
posed [30]. T
he
research was conducted
by building
a framework,
CreekL,
whi
c
h has the c
apability to provide a
comp
atible
knowl
edge
rep
r
esentation
fo
r
similar
do
m
a
in kno
w
led
g
e
.
The
r
e are
two aspe
cts
from
this re
sea
r
ch
that can be fu
rther inve
stig
ated.
1.
Cre
e
kL can
only
a
c
comm
odate a com
patible rep
r
e
s
entatio
n
fo
r simila
r
d
o
mai
n
kno
w
led
ge.
This
can be a
d
vanced by condu
cting research on diffe
rent but rel
a
ted domai
n kn
owle
dge
2.
Cre
e
kL
can
only acco
mm
odate
com
p
a
t
ibility for kn
owle
dge t
r
an
sfer f
r
om
on
e sy
stem to
anothe
r. Thi
s
can
be
adva
n
ce
d by
cond
ucting
re
se
arch o
n
h
o
w to
accomm
odat
e kn
owl
edg
e
integratio
n an
d kno
w
le
dge
sha
r
ing.
By p
r
o
v
id
ing
a
u
n
i
fied
k
n
ow
le
dg
e
r
e
pr
es
e
n
t
a
t
ion
,
kno
w
led
g
e
tr
a
n
s
f
e
r
, in
te
g
r
a
t
io
n
,
and
sha
r
ing
can
be po
ssible.
The
kno
w
le
d
ge tra
n
sfe
r
capability en
sure
s all th
e
kno
w
le
dge
within
one
syste
m
can b
e
tran
sfe
rre
d to
a
ne
w syst
em
. Th
e
kn
owl
edg
e
sharin
g
ca
pabi
lity ensu
r
e
s
t
h
e
kno
w
le
dge
wi
thin one
sy
stem can b
e
sh
ared
to
en
ha
nce
th
e kno
w
ledge of
an
other system. T
he
kno
w
le
dge in
tegration
cap
ability ensu
r
e
s
the
kn
owl
e
dge withi
n
o
ne syste
m
can be u
s
ed
by
anothe
r syst
em. This will
ease the d
e
v
elopment
of
a new kno
w
ledge
-ba
s
e
d
system b
e
ca
use
there
will be
no kn
owl
edg
e acq
u
isitio
n; instead, a
n
e
w
system
can lea
r
n from
alrea
d
y existing
and lea
r
nt sy
stem
s. The capabilitie
s aresh
o
wn in Figure 5.
Figure 5. Knowled
ge Tran
sfer, Shari
ng,
and Integratio
n betwe
en 2
system
s
The main focus of this sol
u
tion is demo
n
stratin
g
ho
w to create a n
e
w kn
owl
edg
e-ba
se
d
system
and
i
n
itiate the
kn
owle
dge f
r
o
m
anoth
e
r
sy
stem.
Th
ere
are
t
w
o po
ssible solution
s
to
accomm
odat
e these
cap
a
b
ilities an
d these a
r
e sho
w
n in Figure 6.
1.
Distri
buted knowl
edge
-b
a
s
ed sy
stem,
whe
r
e an i
n
telligent interf
ace i
s
provid
ed to brid
ge
kno
w
le
dge i
n
feren
c
e f
r
om
a correct
kn
o
w
led
ge b
a
se
system. T
he i
n
terface p
r
ovi
des a tabl
e
whe
r
e a pie
c
e of kno
w
led
ge re
side
s in
a knowl
edg
e-ba
se
d syst
em; so wh
en
a problem i
s
prop
osed to the interfa
c
e, the interfa
c
e can di
spatch the infere
nce to the co
rre
ct kno
w
le
dge.
Knowled
g
eRep
re
sentat
i
on
A
Knowled
g
eRep
re
sentat
i
on
B
tr
anspor
table
information
res
o
urce
information
res
o
urce
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 12, No. 4, Dece
mb
er 201
4: 109
6 – 1104
1102
2. Autonomo
u
s
kno
w
le
dge
-b
ase
d
system
with
kno
w
le
dge sh
ari
ng,
where every knowle
dge
-
based
syste
m
provid
es
a
n
intelligent in
terface.
T
h
is i
n
terface can
ensure
anoth
e
r kno
w
led
g
e
-
based sy
ste
m
can infe
rkn
o
wle
dge fro
m
another
syst
em.
Figure 6. Knowled
ge-ba
se
d System to En
su
re Kno
w
ledge Integ
r
at
ion, Sharin
g, and
Tran
sfe
r
5. Conclusio
n
and Futu
r
e
Rese
arch
A kno
w
le
dge
-ba
s
ed
sy
ste
m
ha
s a
nu
mber
of vari
ances
ba
sed
on the
kn
o
w
led
ge
rep
r
e
s
entatio
n use
d
to sto
r
e the info
rm
ation. This
le
ads to difficul
t
ies in re
usi
n
g kno
w
le
dge
that
is al
rea
d
y stored i
n
a
kno
w
le
dge
-b
ase
d
sy
stem
. This
re
sea
r
ch
p
r
opo
se
s a
metho
d
for
inco
rpo
r
ating
sto
r
ed
kno
w
led
ge fo
r
reuse p
u
rp
oses
by an
oth
e
r
system
u
s
ing
kno
w
le
dge
integratio
n,
kno
w
le
dge
sharin
g, an
d
kn
owl
edg
e
tran
sfer.
T
he p
r
op
osed
method
i
s
an
autonom
ou
s kno
w
le
dge
-b
ase
d
system,
an addition
a
l
method to a
n
alrea
d
y existing di
stribut
ed
kno
w
le
dge
-b
ase
d
sy
stem
. An auton
o
m
ous kn
o
w
l
edge
-ba
s
e
d
system en
su
res
that
e
a
c
h
kno
w
le
dge
-b
ase
d
system
has the capa
bility to dr
aw inferen
c
e
s
fro
m
anothe
r system to find an
answe
r to a specifi
c
pro
b
le
m or even en
han
ce its kno
w
led
ge.
Future research will
impl
ement
this
proposed m
e
thod usi
ng a uni
fied platform
that will
accomm
odat
e multiple kn
owle
dge
-ba
s
ed system
s t
hat are to be
implemented
on one platform.
In using thi
s
platform, the
method
will b
e
devel
op
ed, tested, and m
easure
d
to en
sure kno
w
led
ge
integratio
n, sharin
g, and transfe
r ca
n be
accommo
dat
ed.
Referen
ces
[1]
Ron
a
ld
J. Brachma
n
, Hect
or J. Lev
esq
ue.
Kn
ow
led
g
e
Re
pres
entat
ion
an
d R
e
a
s
oni
ng
. San
F
r
ancisco: Mor
gan Ka
ufman
n
. 2004: 2-1
1
.
[2]
John F
.
So
w
a
.
Know
led
ge R
epres
entati
on: Log
ical,
P
h
il
os
oph
ical, a
nd
C
o
mputati
o
n
a
l F
oun
dati
ons
.
Boston: Cours
e
T
e
chnolo
g
y
,
CENG
AGE Le
arni
ng. 20
00: 1
1
-29.
[3]
Emil V
a
ssev,
Mike H
i
nc
he
y.
Kno
w
l
e
dg
e
Repr
esentati
o
n
an
d R
eas
oni
ng for
Intel
lig
ent Soft
w
a
r
e
S
y
stems.
IEEE Computer
. 20
11; 44(8), 9
6
-9
9.
[4]
Jia
w
e
i
Ha
n, Michel
in
e Kamb
er.
Data Mini
n
g
: Conce
p
ts a
nd T
e
chn
i
q
ues
. San F
r
ancis
co. Morgan
Kaufman
n
. 200
6: 291-3
37.
[5]
Qiu Yun, F
a
n
Jingc
hao, Z
h
o
u
Guomi
n
. Res
earch
on th
e K
n
o
w
l
e
d
ge ES
T
ool Based
on
Binar
y T
r
ee.
T
E
LKOMNIKA Indon
esi
a
Jour
nal of Electric
al
Engin
eeri
ng.
2
014; 12(
3): 223
6-22
44.
[6]
Shia
n-Hu
a Li
n, Meng Ch
ang
Che
n
, Jan-Min
g
Ho. AC
IRD: Intelli
ge
nt Internet Docum
ent Organiz
a
ti
o
n
and R
e
trieva
l.
IEEE Transacti
ons on Kn
ow
le
dge a
nd D
a
ta Engi
neer
in
g
. 2002; 14(
3): 599
-614.
[7]
Xi
an
gfen
g Lu
o
,
Jun Z
han
g, F
e
i
y
u
e
Ye, Pen
g
W
ang, Ch
ua
nlia
ng
Cai. Po
w
e
r Seri
es Re
prese
n
tatio
n
Mode
l of
T
e
xt
Kno
w
l
e
d
g
e
Ba
sed
on
Hum
a
n
Co
ncept
Le
ar
nin
g
.
IEEE Transactions
on S
ystem
s, M
an,
and Cy
bern
e
tic
s
: Systems
. 20
14; 44(1): 8
6
-1
02.
[8]
Razva
n Stefan
Bot, Yi-fang
B
r
ook W
u
,
Xi
n
Che
n
, Quanz
hi
Li.
A Hy
brid
C
l
assifier
Ap
pro
a
ch for W
e
b
Retriev
ed Doc
u
ments Class
ificatio
n
. Proc
e
edi
ngs
of th
e
Internatio
na
l C
onfere
n
ce
on
Informatio
n
T
e
chnolog
y: C
odi
ng a
nd Com
putin
g. 200
4; 1: 326-33
0.
[9]
Hao Ch
en, S
h
i Ying, Jin
Liu, W
e
i Wang.
SE4SC:
A Specific Search Engine for Softwar
e
Co
mp
one
nts
. Procee
din
g
s o
f
the F
ourth Internati
o
n
a
l
C
onfere
n
ce o
n
Comp
uter and
Information
T
e
chnolog
y. 2
004; 4: 86
3-86
8.
[10]
Hsi-Ch
en
g
Ch
ang,
C
h
i
un-C
h
ie
h Hsu.
Usi
ng T
o
p
i
c Key
w
ord Clusters
for Auto
mati
c Docu
me
n
t
Clusteri
n
g
. Procee
din
g
s of the T
h
ird Internatio
nal
C
onf
erenc
e on Informatio
n
T
e
chnol
og
y an
d
Appl
icatio
ns. 2
005; 1: 41
9-42
4.
a
KBS 1
Intelligent Interfa
c
e
KBS 2
KBS 3
KBS 4
KBS 5
b
KBS 1
KBS 2
KBS 3
KBS 4
KBS 5
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Intelligent Interface for a K
nowl
edge
-b
a
s
ed System
(Nyom
an Bogi
Aditya Karna
)
1103
[11]
Li-Ch
un Su
ng,
Chin-
H
w
a
K
uo,
Meng Ch
an
g Che
n
, Yeal
i Su
n.
Progressiv
e
Analys
is Sche
me for W
e
b
Document Classification
. Proceedings
of the 20
05 IEEE/WIC/ACM Internat
ional Conference on
Web
Intelli
genc
e. 20
05: 606-
60
9.
[12]
Shuch
ao W
an,
Jun W
e
i, Jing
yu S
ong, H
eqi
ng Gua
n
.
Dev
e
lo
pin
g
a Sel
e
ction Mod
e
l fo
r Interactive
W
eb Services
.
IEEE Internatio
nal C
onfere
n
ce
on Web
Servic
es. Chica
go. 2
006: 23
1-2
38.
[13]
Chie
n
Chi
n
C
h
en, Ya
o-T
s
ung
Ch
en, Me
ng
C
han
g
C
h
e
n
. 20
07. An
Agi
n
g
T
heor
y f
o
r Eve
n
t
Life-C
ycl
e
Mode
lin
g.
IEEE Transactions
on Systems,
Man, and Cyb
e
rnetics – Part
A: Sys
t
ems a
nd Hu
ma
ns
.
200
7; 37(2): 23
7-24
8.
[14]
Jun W
ang, L
i
Z
ou, Hong P
e
ng, Gexian
g Z
han
g. An Exte
nde
d Spik
ing
Neur
al P S
y
st
em for F
u
zz
y
Kno
w
l
e
d
ge Re
prese
n
tatio
n
.
Internati
o
n
a
l J
o
urna
l of In
nova
t
ive Co
mputi
n
g
,
Informati
on
a
nd C
ontro
l,
ICIC International
. 20
11; 7(7(
A)): 3709–
37
2
4
.
[15]
Manis
h
Jos
h
i,
Viren
d
ra B
hav
sar, Haro
ld B
o
l
e
y
.
A K
now
le
d
ge R
epr
esenta
t
ion Mo
de
l for
Matchmaki
n
g
Systems i
n
e
-Marketpl
aces
.
Procee
di
ngs
of the 1
1
th
Inter
natio
na
l
Confer
enc
e
on El
ectron
ic
Commerce. 2
0
09: 362-
36
5.
[16]
R. Mohame
d
, J. W
a
tada.
Evidenc
e T
heory
Based K
now
le
dge R
epr
esent
ation
. Proc
eed
ings of t
h
e
13th Inter
nati
o
nal
Co
nfere
n
c
e
o
n
Informati
o
n
Int
egr
atio
n a
nd W
e
b-bas
ed
App
licati
ons
a
nd S
e
rvices.
201
1: 74-8
1
.
[17]
Jian
gna
n Qiu,
Yunj
ie D
u
, Ya
nzha
ng W
a
ng.
Extraction a
n
d
Re
prese
n
tati
on of F
eat
ure
Events bas
ed
on a Kn
ow
led
g
e
Mode
l
. IEEE/WIC/ACM Internatio
nal
Conf
erenc
e
on Web I
n
telli
ge
nce a
n
d
Intellig
ent
Agent T
e
chnol
og
y. S
y
dne
y. 2
008: 21
9-2
22.
[18]
Craig
Schlenoff,
Sebt
i Foufou, Stephen Balakirsk
y
.
P
e
rfor
ma
nce Ev
alu
a
t
i
on
of Ro
botic
Know
led
g
e
Repr
esentati
o
n
(PERK)
. Proceed
ings
of the W
o
rksh
op
on Perform
a
nce Metrics f
o
r Intelli
ge
nt
S
y
stems. Mar
y
lan
d
. 201
2: 1-8
.
[19]
YAN Lei, W
A
NG Xin
y
i
n
g, DONG Junlei.
A Pow
e
r Grid Know
ledg
e R
epres
entati
on
Using A
gent-
base
d
Kn
ow
le
dge
Re
pres
ent
ation
in
Perv
a
s
ive C
o
mp
utin
g
. T
he 2n
d IE
EE Internati
o
n
a
l C
onfer
enc
e
on Informatio
n
Mana
geme
n
t a
nd Eng
i
n
eer
i
n
g
(ICIME). Chengdu. 20
10: 29
7
-
300.
[20]
Jun
Xu, Yo
ng
h
u
i Y
ao, T
ao P
e
i, C
han
gq
ing
Yao.
Geo
g
ra
p
h
ic K
now
le
dge
Map
an
d Its
Appl
icatio
n
in
Seis
mic K
now
l
edg
e R
epres
e
n
tation
.
17th In
ternatio
nal
Co
n
f
erence
on
Ge
oinform
a
tics. F
a
irfa
x. 20
09:
1-5.
[21]
Z
e
ti Darle
e
n
a
Eri, Rusli A
b
d
u
llah, Ma
rz
an
ah
A. Jabar, Mas
r
ah Azrifa
h, Azmi Murad.
Ont
o
lo
gy-bas
ed
Know
led
ge M
o
del f
o
r Virtu
a
l
Co
mmun
ities
Profile
. Ma
la
ys
ian
Co
nferenc
e in
Soft
w
a
r
e
Engi
neer
in
g
(M
y
SEC). Joh
o
r
Baru. 2011: 5
08-5
11.
[22]
Che
n
ji
an
Ha
o.
Research o
n
Know
led
ge Mo
del for Ontolo
gy-Base
d
Kno
w
ledge Bas
e
. Internationa
l
Confer
ence
on
Business C
o
m
putin
g an
d Glo
bal Informatiz
a
tion. Shan
gh
ai.
2011: 3
97-3
9
9
.
[23]
Li Z
h
ipi
ng a
nd
Sun Yu.
A F
o
rma
l
Mod
e
l of Know
led
ge Bas
e
Systems in I
n
telli
ge
nt T
u
toring Syste
m
s
.
Internatio
na
l
C
onfere
n
ce on Comp
utation
a
l Intelli
genc
e a
n
d
Soft
w
a
re
En
gin
eeri
ng. W
u
h
an. 2
009:
1-
4.
[24]
Jian
hui Sh
i, Xi
aomi
ng C
h
i,
Lihu
a Liu.
D
e
sig
n
of a Dyna
mic Co
ll
ab
orative L
earn
i
ng Oriente
d
Know
led
ge M
o
del
. Inform
atio
n T
e
chnol
og
y
and Artific
i
al
In
tellig
enc
e C
onf
erenc
e (IT
A
IC). Chon
gq
ing.
201
1: 385-
388.
[25]
Chu
an Z
h
a
ng,
Xi
angs
he
ng
Yang, Sh
imin
Du.
A Distri
buted K
now
le
dge Mo
de
l fo
r Know
led
g
e
Mana
ge
me
nt System
. W
i
C
O
M 2008: Internati
ona
l Co
nferenc
e on
W
i
reless Com
m
unic
a
tions,
Net
w
orki
ng a
n
d
Mobi
le Com
p
uting. Da
lia
n. 2
008: 1-4.
[26]
Chin-B
in
W
a
n
g
, Yuh-M
i
n
Ch
en, Yu
h-Z
e
n
Che
n
.
A D
i
stri
buted
Kn
ow
led
ge M
ode
l for
Coll
ab
orative
Engi
neer
in
g K
now
led
ge Ma
n
age
ment in A
l
l
i
ed C
onc
urrent
Engi
neer
ing
. Internatio
na
l
E
ngi
neer
in
g
Mana
geme
n
t Confer
ence (I
E
M
C). 2002: 70
1-70
2.
[27]
Chu
ng-Yi H
u
a
ng an
d Run
g
-
C
hin
g
Ch
en.
Theses Clust
e
r
Based on B
ilin
gu
al an
d Synony
mous
Keyw
ord sets Using Mutua
l
Informatio
n
. Procee
din
g
s o
f
the Eight
Internatio
nal C
o
nferenc
e on
Machi
ne Le
arn
i
ng a
nd C
y
b
e
rn
etics. Baodi
ng.
2009: 2
999-
30
04.
[28]
D.S. Kala
na
Mend
is, Asoka
S. Karu
n
a
n
a
nda, U. S
a
ma
ratung
a, U. R
a
tna
y
ak
e.
Ta
cit Kn
o
w
l
e
dge
Modeling in Intellige
nt Hybr
id system
s
. Internati
o
n
a
l C
onfere
n
ce o
n
Industrial a
n
d
Information
S
y
stems, ICIIS. Penade
ni
ya.
200
7: 279-
284.
[29]
T
ony
Vea
l
e an
d
Yanfe
n
H
ao.
A F
l
uid K
now
l
edg
e Re
pres
e
n
tation for
Un
d
e
rstand
ing
an
d
Generati
ng
Creativ
e Metaph
ors
. COLING '
08: Proceed
ings of
t
he 22n
d Internati
ona
l Co
nferenc
e on
Comp
utation
a
l Lin
guist
ics. 20
08: 945-
95
2.
[30]
Agnar
A
a
modt
.
A Know
ledg
e
Repres
entati
o
n System for I
n
tegrati
on of
Genera
l
an
d
Case-S
pecific
Know
led
g
e
. 6t
h Internati
o
n
a
l
Confer
ence
on
T
ools
w
i
t
h
Arti
ficial Inte
lli
ge
nc
e. Ne
w
Orl
ean
s.1994:
836-
839
[31]
Nho
n
V. Do.
Mode
l for Kn
o
w
l
e
d
ge B
a
ses
of Comp
utatio
nal Ob
jects.
IJCSI International Journal of
Co
mp
uter Scie
nce Issues
. 20
10; 7(3): 11-2
0
.
[32]
Hu-Ch
en L
i
u, Qing-L
i
an L
i
n, Lin
g
-
X
ia
ng Ma
o,
Z
h
i-Ying Z
h
ang. D
y
n
a
mic Adaptiv
e F
u
zzy Petri Nets fo
r
Kno
w
l
e
d
ge R
e
prese
n
tatio
n
a
nd Re
ason
in
g.
IEEE Transactions on Syste
m
s, Man, and
Cyber
netics:
System
s
. 20
13
; 43(6): 139
9-1
410.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 12, No. 4, Dece
mb
er 201
4: 109
6 – 1104
1104
[33]
Hu-Ch
en
Liu,
Lon
g Li
u, Qin
g
-Li
an L
i
n, Na
n
Liu. Kn
o
w
l
e
dge Ac
quis
i
tio
n
an
d Re
pres
entatio
n Usi
n
g
F
u
zz
y
Ev
ide
n
t
i
al
Reas
on
ing
an
d D
y
n
a
mi
c Ada
p
tive F
u
zz
y Petri
Ne
ts.
IEEE Transactions
on
Cyber
netic
s. 2
013; 43(
3): 105
9-10
72.
[34]
Rob
e
rt Harriso
n, Christin
e W
.
Chan.
A T
o
o
l
for Dyna
mic Kn
ow
ledg
e Mod
e
l
i
ng
. 6th IEEE International
Confer
ence
on
Cogn
itive Infor
m
ati
cs (ICCI'
07). Lake T
ahoo. 200
7: 513-
521.
[35]
Nho
n
V. Do, H
i
en
D. Ng
u
y
en.
A Know
le
dg
e
Mode
l a
bout R
e
lati
ons
and
A
pplic
atio
n
. ISSDM 20
12
:
6th Intern
ation
a
l C
onfere
n
ce
on N
e
w
T
r
en
ds in
Inf
o
rmati
on Sci
enc
e, Service Sc
ienc
e
and
Dat
a
Minin
g
. T
a
ipei. 201
2: 707-
710.
[36]
Guoha
i Z
han
g,
Yushe
ng L
i
. Multipl
e
Disc
i
pl
ines
Pro
duct D
e
sig
n
Kno
w
l
e
d
ge Re
pres
enta
t
ion Strateg
y
Based
on On
tolog
y
and S
e
mantic N
e
t
w
ork.
T
E
LKOMNIKA Indon
es
ian J
ourn
a
l o
f
Electrica
l
Engi
neer
in
g.
2013; 11(
10): 60
74-6
079.
[37]
Li Yue-
xin, Ho
ng Z
ong-
xian
g.
Researc
h
of Semantic N
e
t
w
o
r
k Kno
w
l
edg
e Repr
esentati
o
n
and Quer
y
Algorit
hm bas
e
d
on
Rel
a
tio
n
a
l
Mod
e
l.
T
E
LK
OMNIKA Indo
nesi
an Jo
urn
a
l
of Electrica
l
Engi
neer
in
g.
201
3; 11(1
1
): 6591-
659
9.
Evaluation Warning : The document was created with Spire.PDF for Python.