TELKOM
NIKA Indonesia
n
Journal of
Electrical En
gineering
Vol. 14, No. 2, May 2015, pp. 343 ~ 35
2
DOI: 10.115
9
1
/telkomni
ka.
v
14i2.750
5
343
Re
cei
v
ed Fe
brua
ry 11, 20
15; Re
vised
April 21, 201
5; Acce
pted
May 1, 201
5
Data Exchange Design with SDMX Format for
Interoperability Statistical Data
Jaka Sembir
ing*
1
, Ana Ul
u
w
iy
ah
2
1
School of Elec
trical Eng
i
ne
eri
ng an
d In
forma
tics, Institut
T
e
knol
ogi Ba
nd
u
ng,
Jl. Ganesha
10
Bandu
ng 4
0
1
3
2
, Indon
esia, P
h
./F
ax:+
62-2
2
-
250
22
60/+
62-2
2
-25
342
22
2
Educatio
n T
r
ainin
g
Cent
er, Statistics Indon
e
s
ia,
Jl. Jagakars
a
Ra
ya 7
0
, Jakar
t
a Selatan, Ind
ones
ia, Ph./F
ax: +
62-2
1
-78
7
3
782-
83/+
62-
21
-787
549
7
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: jaka@it
b
.ac.i
d
A
b
st
r
a
ct
T
oday
’
s
c
onc
e
p
t of Open
Govern
ment D
a
ta (OGD) for o
pen
ness, trans
pare
n
cy a
nd
e
a
se o
f
access of
data
ow
ned
by g
o
v
ern
m
e
n
t ag
en
cies b
e
co
mes
incre
a
sin
g
ly important.
T
h
is i
n
itiativ
e
e
m
erg
e
s
from the de
mand of data u
s
ersf
orthe dat
a bel
ongs to
the gover
n
m
e
n
t agenc
ies. T
he data serv
ice
s
provi
d
in
g an
easy acc
e
ss, chea
p, fast, and i
n
tero
p
e
ra
b
ility are
ne
ed
ed by th
e us
ers an
d bec
o
m
e
s
importa
nt indi
cator perfor
m
ance for resp
ective gov
er
n
m
e
n
t age
ncie
s. Statistical Data an
d Metada
t
a
Exchan
ge
(SD
M
X) is a
n
e
w
stand
ard f
o
r
m
at
in
the
da
ta diss
e
m
in
ati
on
activities
p
a
rticul
arly i
n
t
h
e
excha
n
g
e
of st
atistical
data
a
nd
metadat
a vi
a Intern
et.
In th
is resp
ect SDM
X
sup
port th
e i
m
p
l
e
m
e
n
tatio
n
o
f
OGD project.
T
h
is pa
per
is
on th
e tech
n
i
cal
desi
gn, d
e
vel
o
p
m
e
n
t a
nd i
m
ple
m
ent
ation
of dat
a
and
meta
data exc
h
ang
e service o
f
statisti
cal data usin
g SDMX
format to su
p
port intero
pera
b
ility d
a
ta thro
ug
h
w
eb services.
T
h
ree res
u
lts
are pr
opos
ed
: (i) framew
or
k for standar
d
i
z
a
t
i
o
n
of st
ru
cture of statistical
pub
licati
ons
da
ta mode
l w
i
th
SDMX; (ii) d
e
s
i
gn
archit
ect
u
re of dat
a shar
i
ng
mo
del;
and
(iii) w
eb s
e
rvi
c
e
imple
m
entati
o
n
of data and meta
dat
a exc
h
ang
e service
usin
g Service
Oriented An
al
ysis and Des
i
gn
(SOAD) meth
o
d
. Impl
e
m
entat
ion at St
atistic
s
Indon
esia (B
PS) is chos
e
n
as a case study to prov
e the
desi
gn c
once
p
t. It is show
n throug
h q
u
a
n
t
itative a
ssess
me
nt an
d b
l
a
ck box
testi
n
g
that the
desi
g
n
achi
eves its ob
jective.
Ke
y
w
ords
: op
en gov
ern
m
ent
data, statistical dat
a an
d met
adata exc
h
a
n
g
e
, w
eb service
Copy
right
©
2015 In
stitu
t
e o
f
Ad
van
ced
En
g
i
n
eerin
g and
Scien
ce. All
rig
h
t
s reser
ve
d
.
1. Introduc
tion
Initiative on Op
en
Data
Gove
rnme
nt (O
GD) is i
n
crea
singly
getting
sup
portfrom
manyco
untri
e
s
. Open
dat
a provide
s
an easy
accessto
the data so
thatvariou
su
serssu
ch asi
n
stitution
s
,
scienti
s
ts
o
r
oth
e
rcommu
nitie
s
are allo
wed to retri
e
ve
datawith
outa
licen
seo
r
p
a
t
entrest
rictio
n
s
[1]. Pr
oviding fre
e
ly availabled
ata
t
o everyone
via
theWe
b
isea
si
er
whe
n
the
data i
s
i
n
te
rco
nne
ct
ed
e
a
ch
othe
r. O
penn
essa
nd
conne
cted
ne
ssof
datais the
go
alof thep
ara
d
i
g
m oflin
ked
o
pend
ata [2
].
Open
data
be
come
s
an i
m
portant
aspe
ct in
a pu
blic in
stitution who
s
e
functio
n
i
s
to se
rve th
e
co
mmunitie
s
. With
op
e
n
data
a
pu
blic
institution
ca
n fulfill its ob
ligat
ion of
creating a
p
rim
e
se
rvicea
s
one of th
eir
Key Perform
ance
Indicators for organi
zational pe
rform
a
nce [3]. Statist
i
cs Indonesi
a (BPS) as
a
publi
c
institution
also h
a
s
a strategic
obje
c
tive of delivering
a
prim
e se
rvice to
the publi
c
in providi
ng
and
dissemin
ating
high
qu
ality
data a
n
d
info
rmation.
BPS
is en
co
ura
g
ed to
chan
ge
the m
e
thod
o
f
dissemin
ating
data and i
n
fo
rmation to
be
in line with
t
he OG
D initia
tives. One of
thecruci
al pa
rts
is on how toexchange
data and metadat
a that
ca
n provideinteroperabilit
yservices fo
r openness
and con
nect
e
dne
ss.Statisti
cal Data
an
d Metadata
E
xchang
e (S
DM
X), whi
c
h i
s
initiated by wo
rld
seven statisti
cal
o
r
ga
nizations, can be u
s
ed as a
fra
m
ewo
r
k in
stan
dardi
zin
g
stat
istical
data a
nd
metadata
excha
nge.
With SDMX it i
s
po
ssible
to sim
p
lify and st
reamli
n
e
the p
r
o
c
e
s
s of
excha
nging
d
a
ta and m
e
ta
data u
s
ingthe
same
dat
a
structu
r
e a
nd
con
c
e
p
t,in order to im
prov
e
the timeliness of data
services,
accessibility, interpretability and
coherence [4-6].In this paper
we will use SDMX for standardi
zing the data
and metadata where for clarity we will use BPS
data as
our speci
a
l
case.The
S
D
MX
standard
will be im
plemented i
n
S
e
rvice Orient
ed
Architecture (SOA) schem
e
descri
bed i
n
[7], where i
n
parti
cular we
will use the
Service
Oriented
Analysis a
n
d
Desi
gn (S
OAD) a
s
a
desig
n app
roa
c
h [8]. We p
r
ovide
the prototype of
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 14, No. 2, May 2015 : 343 – 352
344
impleme
n
tation of th
e d
e
si
gn u
s
in
g
web
se
rvice
s
. T
h
i
s
i
s
a
natu
r
al
choi
ce
si
nce t
he
web
servi
c
e
is de
sign
ed to sup
port inte
rope
ra
bility and intera
ction
s
between
system on a n
e
t
work [9]. In this
pape
r we prop
ose
th
ree
mai
n
result
s:
(i
)
stand
ardi
zati
on fra
m
e
w
o
r
k
of st
atistical
publi
c
ation/a
ggre
gate dat
a
mo
del stru
cture,
(ii
)
d
e
s
ign
a
r
chite
c
ture
of data
sh
ari
ng
se
rvice
model, a
nd
(i
ii) imple
m
ent
ation of d
e
si
g
n
archit
e
c
ture of data
exchang
eu
sing
web
se
rvice.
We
will al
so p
r
e
s
ent the valida
t
ionre
s
ult of the pr
opo
se
d
desi
gn
with b
o
th user/expe
r
t judgm
ent a
nd
black box test
ing.
2. Rese
arch
Metho
d
2.1. Standar
d
ization F
r
a
m
e
w
o
r
k o
f
St
atis
tical Da
ta Struc
t
ure
w
i
th SDMX
This
section
will explain our first result on
the standardi
zation fram
ework of
statistical
publi
c
ation
d
a
ta st
ru
cture.
For
ca
se ill
ustratio
n
we
will
use th
e
bu
sine
ss p
r
oce
s
s of
Da
ta
Dissemi
nation Directorate
(DDS)
of BPS. The busi
n
ess process
of
the mentioned di
rectorat
eis a
pro
c
e
ss flo
w
of activities to dissemi
n
a
te the re
sul
t
of statistical information in the form o
f
servi
c
e
s
to th
e publi
c
u
s
e
r
s. The
obje
c
tive of
our
pro
posed
schem
e is to im
pro
v
e the value
of
publi
c
organi
zation, in this case BPS, through
increasing the openess of
data, ease
of access,
and d
a
ta inte
rope
ra
bility.
Open
ess of t
he data
can
be mea
s
u
r
ed
by availability of both data
stru
cture an
d
conte
n
t that can
be a
c
cessed
and
u
s
ed, reu
s
ed
and redi
strib
u
ted by all d
a
ta
use
r
s.
Ease
of a
c
ce
ss
can
be
me
a
s
ured
by
d
a
ta se
rvicesm
edium
thatca
n be
acce
ssed
easilythroug
h
thedeviceg
a
dgets, mobil
e
phone
s, PC
andla
p
top, etc. Mean
while
interope
rabil
i
ty
can be measured by
capa
bility servi
c
e to be operat
ed on diffe
rent
platform
s.In
order to achi
eve
theseval
ue
s and to sho
w
the con
c
e
p
tprop
o
sed in
this pap
er,
we ad
opt
Gene
ral Stat
istic
Business Process Model (GSBPM)
for busi
n
ess process flow [
10]. By adopting GSBPM we can
derive th
e b
u
sin
e
ss
pro
c
ess a
s
sh
own in Fi
gu
re
1 wherea
s t
he b
u
si
ne
ss
pro
c
e
s
s of d
a
ta
dissemin
ation
c
an
be el
abo
rated fu
rthe
r i
n
Figu
re 2. In
this bu
sin
e
ss p
r
o
c
e
ss m
o
del, the data
to
be published will be taken from
the di
ssemination
data
warehouse, not directly from
subject
matters
. Based on the
GSBPM, we
propos
e
the new bus
i
ness
proc
es
s
arc
h
itec
ture for data
dissemin
ation
as sh
own in
Figure 3. In this ar
chite
c
tu
re, the publi
s
hed data is ta
ken fro
m
a data
wareho
use after being
validated an
d analyze
d
by Analysis and Devel
opment Statistics
Dire
cto
r
ate.
Before p
ubli
c
ation, the
dat
a will
be
validated a
nd ma
pped
to the mo
del
stru
ctures/d
ata sche
me p
r
e
determi
ned i
n
the form
of DSD/MSD. Thi
s
p
r
o
c
e
ss
i
s
condu
cted as a
methodto sta
ndardize format and mo
del stru
ct
ure
of the publishe
d data, and to avoid the
inco
nsi
s
ten
c
y
of data. The
data is read
y to be
publi
s
he
d/rele
ase
d
to the web
servi
c
e p
r
ovi
der
only after it is validated. We focu
s o
n
th
e dissem
i
nati
on of stati
s
tics information
throug
h cre
a
ting
servi
c
e
s
of d
a
ta and met
adata excha
nge whe
r
e fo
r this p
u
rp
ose we
cre
a
tea
new
servi
c
e
for
data and met
adata exchan
ge. This
servi
c
e pr
ovide
s
e
a
sy acce
ss for the users.
Figure 1. Pro
posed bu
sin
e
ss p
r
o
c
e
ss a
c
tivities at
Data Dissemi
nati
on Directo
r
at
e of Statistics
Indonesia aft
e
r adopting the GSBPM.
Note that difference from exi
s
ting condition, the data to be
publi
s
hed
will
be taken fro
m
the dissem
ination data
wareho
use, not directly fro
m
subje
c
t ma
tters
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Data Excha
n
ge De
sig
n
wit
h
SDMX Form
at for In
teropera
b
ility Statistical
Data (Jaka Sem
b
irin
g)
345
Figure 2. Det
a
il bus
ines
s
process
of dat
a diss
em
ination derived from GSBPM. Note that this
decompo
sitio
n
provide
s
u
s
with can
d
idat
e
servi
c
e
s
su
ch a
s
dissemi
nation service
Figure 3. Pro
posed bu
sin
e
ss p
r
o
c
e
ss a
r
chite
c
ture
of
statistics data
disseminatio
n. This
architec
ture has
been aligned with
the future c
o
ntextual plan of BPS
The fra
m
e
w
ork of data
and m
e
tad
a
ta model
structu
r
e i
s
d
e
sig
ned
with
SDMX-
Information M
odel (S
DMX-I
M
) method
wh
ich in
cud
e
s
several comp
o
nents: DS
D, MSD, Data set,
Metadata
set
and SDMX-ML messa
ge.
Based on th
e analysi
s
on
the data to be publi
s
he
d a
nd
SDMX-IM co
mpone
nts, we can d
e
fine the Level 1 st
anda
rdi
z
ation
stru
cture mo
del as in Fig
u
re 4
whe
r
e we ha
ve three main
activiti
es in this process a
s
follows:
a)
delivery of da
ta to be publi
s
he
d to the Data Dissemi
n
ation Re
po
sitory,
b)
mappin
g
lo
ca
l data with Data Structu
r
e
Definition (DSD) an
d Me
tadata Stru
cture
Definition (M
SD), and
c)
publi
s
h/rel
e
a
s
e of data pu
blicatio
n.
Figure 4. Standardization
pro
c
e
ss d
a
ta stru
cture Lev
el 1
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 14, No. 2, May 2015 : 343 – 352
346
We proceed t
o
the Level 2 standa
rdi
z
ati
on of the model stru
ctu
r
e for detail pro
c
e
ss a
s
in Figure 5.
Figure 5. Standardization
pro
c
e
ss d
a
ta stru
cture Lev
el 2
The gen
eral pro
c
e
ss in L
e
v
el 2 can be
explained in
detail as follo
ws.
1)
Determine th
e Con
c
e
p
t and Co
ncept Schem
e. Co
nce
p
t has a
very importa
nt role in the
SDMX-IM b
e
c
au
se it i
s
u
s
ed to de
scrib
e
the
multi-di
mensi
onal ta
ble structu
r
e
or meta
data
stru
cturere
p
o
rt. Co
ncept
stru
ct
ure
con
s
ist
s
of
Dime
nsio
ns, M
e
a
s
ure
s
, an
d Attribute
s
. Th
e
sampl
e
re
sult
of concept a
nd co
ncept schem
e in this paper
can b
e
seen in Ta
bl
e 3.
2)
Cre
a
ting
Cod
e
List. Co
de
list in the SDMX
-IM i
s
a
list contai
ni
ng co
de
s th
at represent
con
c
e
p
ts
(di
m
ensi
o
n
or a
ttribute) i
n
either DS
D o
r
MSD [11]. Sa
mple
of code
list cre
a
ted i
n
this pap
er is
sho
w
n in Ta
b
l
e 3.
3)
Data Str
u
ctu
r
e Definitio
n
(
D
SD
). DS
D i
s
ba
si
cally a
descri
p
tion fo
r con
c
ept
s th
at have b
een
identified an
d
establi
s
he
d. It describe
s
wheth
e
r
the
con
c
e
p
t is a dimen
s
ion o
r
an attribute.
DSD
ca
n b
e
used i
n
Tim
e
Seri
es (TS
)
, Cro
s
s Se
ctional
(CS
)
a
nd multidi
m
e
n
sio
nal
data
tables. In o
u
r ca
se to cre
a
t
e DSD, first we
n
eed to d
i
fferentiate b
e
twee
n TS d
a
ta and
CS
data. Then we identify the con
c
ept, determin
e
the measure a
nd make a
n
associatio
n with
the Cod
e
List
s. The sampl
e
result for ou
r ca
se
can be
seen in Ta
bl
e 3.
4) Metadata
Structure
Definiti
on
(MS
D
). A
s
in DS
D, MS
D is a
stru
ctu
r
e for metad
a
t
a. Metadata
is used a
s
a
descri
p
tion
or refe
re
nce
for the obje
c
t to be exchang
ed. The
purp
o
se of
metadata
structure i
s
to fa
cilitate ind
e
xing, se
archin
g
,
pro
c
e
ssi
ng
and
repo
rting
of statisti
cal
activities. BPS devide met
adata into three types
: descriptive metadata,
structural metadata
and a
d
mini
strative metad
a
ta. The
s
e
metadata
are
asso
ciated
with refere
nce metad
a
ta i
n
SDMX.
5)
Mappin
g
Dat
a
. The
mai
n
probl
em
of d
a
ta exchan
g
e
on
the
dat
a sha
r
ing
sy
stem
s i
s
that
there a
r
e various d
a
taba
se stru
ctu
r
e i
n
tegrate
d
in
one sy
stem,
so that itis n
e
ce
ssary to
perfo
rm data
mappin
g
am
ong
st relate
d
databa
se
to improve
the conve
r
ge
nce [12].
In
our
ca
se
we
emp
l
oy five step
s for m
appin
g
betwe
en lo
ca
l datab
ases
a
nd the
predef
ined
DSD
c
o
nc
ept as
follows
.
a) List
thecon
ce
ptsthat
existin
thelocal d
a
ta
.
b)
Pairingalllocal dataconcept
withthe DS
D concept.
c) Pairingl
ocal
code
swith
D
SDcod
e
s.
d) Tran
slatin
gSDMXQu
e
ry.
e) Tran
slate
into
dataset.
6)
SDMX-ML
m
e
ssag
e. SDM
X
-ML i
s
cre
a
ted ba
se
d o
n
XML; so th
at SDMX
comp
lieswiththe
con
c
e
p
t an
d t
he
rule
s
of X
M
L.The
mo
st impo
rtant
co
mpone
nt in
th
e impl
ement
SDMX XML
are an XML n
a
mespa
c
e an
d XML Sche
ma (XSD).
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Data Excha
n
ge De
sig
n
wit
h
SDMX Form
at for In
teropera
b
ility Statistical
Data (Jaka Sem
b
irin
g)
347
2.2. Architec
ture o
f
Da
ta
Sharing Mod
e
l
The gen
eral
modelof da
taandmeta
dat
aexch
ang
e processp
ro
po
sedin this pa
per isa
simple mod
e
l
wh
ere d
a
taprovid
erpu
blish/di
ssemi
natethe d
a
ta
tothe co
nsu
m
er u
s
in
g
web
s
e
rvic
es
for data s
h
aring im
pleme
n
tatio
n
. The syste
m
diagram ca
n be se
en in
Figure 6.
Figure 6. Dat
a
sha
r
ing m
o
del usi
ng web
service
In our ca
se,
two model
s
of datash
a
rin
g
are involve
d
, i.e.: (i) datasha
ring fro
m
data
provide
r
to the publi
s
he
r system, and (ii
)
data s
hari
n
g from the p
ublisher
syst
em to con
s
u
m
er.
Data
provid
e
r
in
cludi
ng
subje
c
t matters a
nd
gov
ernment in
stitu
t
ions;
sen
d
s
the data
to t
h
e
system u
s
in
g
push m
e
tho
d
, in other
word
s the dat
a
provide
r
pu
sh/sen
d the d
a
ta to the Da
ta
Dissemi
natio
n pu
blicatio
n
repo
sito
ry. Where
a
s t
hed
a
t
a tran
smi
ssi
on fro
m
syste
m
to con
s
um
ers
is
u
s
ing
pull
method wh
ere co
nsume
r
s end
s
a
d
a
ta
re
que
st messag
e
a
n
d
will pull/ret
r
ieve
dataset
from
web servi
c
e system.
The architectu
re
desi
gn of dat
a dissemin
ation that sup
p
o
rts
the above me
ntioned d
a
ta sha
r
ing
can b
e
see
n
in Fig
u
re 7.
Figure 7. Det
a
il of data sh
aring
system
desi
g
n
The disse
m
in
ation data wa
reho
use save
s all
the local
data publi
c
at
ion com
e
s fro
m
th
e
s
ubjec
t matters and othe
r data s
o
urces
.
Based on GSBPM,
this
warehous
e will
be managed
by
the Statistical
s
Info
rmation
Systems
Division
an
d Sta
t
istical
Di
sse
m
ination
Divi
sion
so
that t
h
e
data ca
n be
validated be
fore publi
c
ati
on.In additi
o
n
to dissemi
nation re
po
si
tory, we have a
mappin
g
store to
save th
e
data
co
ncep
t. Statisti
cal
Dissemi
natio
n Divi
sion
wh
o al
so
perfo
rms
the data ma
pping m
a
intai
n
s the m
appi
ng sto
r
e. Wit
h
this a
r
chite
c
ture, all
dat
a will have t
h
e
same
con
c
e
p
t. Both data
storage
are
desi
gne
d
to
sup
p
o
r
t dat
a exch
ang
e
servi
c
e
s
. User i
s
allowed to qu
ery or re
que
st the data via
the w
eb cli
e
n
t, and this we
b client will send a messa
g
e
to the web
service. T
h
is
service will
ret
r
ieve
the dat
a
from disseminat
ion data warehouse
and
mapping store located in the BPS servers.
Figure 8. Dat
a
dissemin
ati
on we
b se
rvice architectu
re
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 14, No. 2, May 2015 : 343 – 352
348
The we
b se
rvice archite
c
ture for data
excha
nge services is
sho
w
n in Figure 8
.
Users
who
send request data in SDMX
Query messages,
which will be
translated
by SDMX Query
Parser, initiat
e
the
pro
c
e
ss of data
di
sse
m
ination
i
n
th
e web
se
rvice
.
The SQ
LQu
e
ry retrieve th
e
data structu
r
e
s
and
data
s
et
s from the
sto
r
e Map
p
ing
a
nd data di
sse
m
ination repo
sitory, and th
e
system
will generate data i
n
SDMX-ML f
o
rmat
which will be sent to the web
c
lient
.
2.2. Design
of Da
ta Exc
h
ange Serv
ic
e
w
i
th SOAD
In this se
ctio
n we will di
scu
ss the d
e
sign pro
c
e
s
s
of data and
metadata ex
cha
nge
servi
c
e
s
. We
use
SOAD method
olog
y as a
n
ap
p
r
oa
ch
of SO
A [8]. Following th
e SO
AD
methodol
ogy
for a
nalysi
s
a
nd d
e
si
gn, first
step
i
s
th
e
Co
nceptual
View (CV
)
where
we obta
i
n
the bu
sine
ss pro
c
e
s
s a
n
d
su
b bu
sin
e
ss p
r
o
c
e
s
s for data a
nd m
e
tadata
se
rvice
as
sh
own
in
Figure 9.
Figure 9. Business process and
su
b-b
u
si
ness process analysi
s
Based th
e p
r
opo
sed
syste
m
depi
cted i
n
Figu
re 3, t
here
are two
system
swe
need to
cre
a
te, i.e.(i) Web Service Provider,
whi
c
h is
u
s
e
d
by data publish
e
r to provide, repo
rt and
dissemin
ate
statistical dat
a, and (ii
)
Web Appli
c
at
io
n Client,
w
hich is u
s
ed
bythe usersto inv
o
ke
data.The ge
n
e
ral fun
c
tion
of the system
is described i
n
Table 1.
Table 1. Gen
e
ral fun
c
tion
of the propo
sed syste
m
Integr
ate
d
Dat
a
Publica
t
io
n an
d
Data Sh
aring
M
a
nage
men
t
Sy
st
e
m
S
e
r
v
ic
e
Descrip
tion
Web Service Provider
Providing publica
t
ion data and
metadata
F
a
cilitatethe
dataproviderto
pr
epareandt
ransmi
t
data/metadata
fo
r the users
Reporting an
d dissemination
publicationdata
F
a
cilitatetheinstr
umentfor
p
ublicati
on and
reportingth
e
data for
publicationsafterbeingvalidated
Web Application
Client
Requesting publication data
Facilitatetheuserstorequestthe da
tabased on
the published categor
yof data
Display
i
ng pu
blication data
Facilitate the
users by
d
i
splay
i
ngdata
fromthe use
r
que
r
y
according
to t
he provided
format
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Data Excha
n
ge De
sig
n
wit
h
SDMX Form
at for In
teropera
b
ility Statistical
Data (Jaka Sem
b
irin
g)
349
Secon
d
ste
p
, the Logi
cal
View (L
G)
co
nsi
s
ts of thre
e layers: bu
siness laye
r, servi
c
e
layer and
co
mpone
nts lay
e
r whi
c
h i
s
d
e
rived from
CV. The re
sult of analysi
s
ca
n be see
n
in
Table 2.
Table 2. Prop
ose
d
se
rvice
s
at Logical Vi
ew
S
y
s
t
em Fu
ncti
o
n
alDo
mai
n
Co
v
e
ra
geof
Fu
n
c
tion
al
sy
st
em
Descrip
tion
Providing publica
t
ion data
and metadata
Registing data pr
ovider
Services
for registeringthe publication data
Sending data pu
blications
Servicestouploadthe publication datatothe
database of dissemination/publication
Mapping local data to DSD
and MSD SDMX
Services used by datapublishertoverif
y
,
mapandvalidatethe publicationdataw
ithth
e
pre
defined DSDand
MSD
Publishing the publications
data
Servicesused by
datapr
ovidersto approvethe
publication datato be published to
the public
Reporting an
d dissemination
publication data
Reporting an
d dissemination
publication data
Servicefor the users to view
th
e p
ublished data
Requesting publication data
Selecting data categor
y
Servicesfor the usersto selectappropriatedata
categories
Selecting data flow
Services
for theusersto choosedataflow
Sending request
Servicesforthe userto choose and
invokethe
datarequest
Display
i
ng pu
blication data
Display
i
ng dat
a
Servicefordisplaying data
w
ithSDM
X-ML,
PDFandXLS da
t
a
format,an
d
for
visualizationw
ithtabulationandgra
phs
Do
w
l
oading data
Service for do
w
n
l
oading data
w
ith
specificformat
Finally, the Physical View
(PV) con
s
ist
s
of
four layers web servi
c
e layer, presentation
layer, ap
ppli
c
ation laye
r a
n
d
data
mo
del
layer.
We
b
service
laye
r i
s
divid
ed i
n
to
two
pa
rts:
we
b
servi
c
e de
sig
n
and service seq
uen
ce
diagram. Pr
e
s
entatio
n layer is an inte
rface de
sig
n
for
Integrated
Data Publi
c
ati
on a
n
d
Dat
a
Shari
n
g
Ma
nagem
ent Sy
stem
wh
ere
i
n
this p
ape
r
we
provide
seve
n interfa
c
e
s
i
n
clu
d
ing: Ho
me Page, Data Cate
gory,
Data a
nd Met
adata Set Upl
oad,
Data Map
p
in
g, Login/Re
g
i
ster, We
b Service
Dat
a
Publish
ed,
and Data
set Prese
n
tation.
Applicatio
n Service l
a
yer i
s
the interfa
c
es for
busi
n
e
ss lo
gic im
pl
ementation
consti
st of eig
h
t
interfaces in
cluding: Id
ata
C
ateg
ory Ap
pS,
IdataMet
adataSetUplo
adAppS, Idat
aMappi
ngAp
p
S,
IloginAppS,
Imembe
rRegi
sterA
p
p
S
, IdataF
lowAppS, IdataPre
sent
ationAppS,
and
IDSDMSDA
p
p
S. Data mo
del layer
co
n
s
ist
s
of two
Data T
r
an
sfe
r
Obje
ct (DT
O
). The fi
rst i
s
the
Data
Dissem
ination DTO
con
s
i
s
ts of L
ogQu
e
ry, Da
taSet, MetadataSet, Log
Do
wnlo
ad an
d
Publish
e
r. T
h
is
comp
one
ntis u
s
ed
to
sto
r
e lo
ca
l
publi
c
ation
s
d
a
ta i
.
e. dataset a
nd meta
data
s
et.
The
se
co
ndis Map
p
ing
Store
co
nsi
s
ts of
DS
D, MS
D,
Mappin
g
, Ma
ster Data Pro
v
ider,
Code
Li
st,
Categ
o
ry Dat
a
and Data
Flow. Thi
s
compon
ent
is
use
d
to store data che
ck, validation and
mappin
g
so t
hat the data is rea
d
yto be dissemin
ated
.
3. Results a
nd Analy
s
is
In this
section we w
ill
show the implementation
result
of the
design involvi
ng public
domain
data
on po
pulatio
n
/
demog
rap
h
y of Statistics
Indon
esi
a
. Ag
greg
ate d
a
ta
for sample
ca
se
is(i
) the popu
lation of Indonesi
a
by province
19
71, 1
980, 1990, 1
995, 2000, a
nd 2010, an
d
(ii)
total and percenta
ge of poverty and p
o
verty li
ne by province 2
007, 2008, 2
009, 2010, 2
011,
2012. Both d
a
ta can b
e
viewed in two categ
o
ri
e
s
,
which are Time Serie
s
(TS) an
d Cross
Sectional
(CS). Usin
g the method d
e
s
cribe
d
in t
he previou
s
section, first we dete
r
min
e
the
con
c
e
p
t wh
ere in this
simp
le ca
se i
s
d
e
noted
with POP_CONCE
P
T sin
c
e
we
deal
with dat
a on
popul
ation/de
mographi
c. Base
d on the
con
c
e
p
t, we
can d
e
rive th
e cod
e
list su
ch a
s
in colu
mn
five of Table
3. We
can
creat
e the DS
D/MSD by first
sep
a
ratin
g
th
e TS and
CS
, then usin
g the
method
de
scribed i
n
the
previous sectio
n we
ca
n
obt
ain the
DS
D/MSD
whe
r
e t
he
sampl
e
re
sult
can b
e
se
en i
n
Table 3.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 14, No. 2, May 2015 : 343 – 352
350
Table 3. Fo
rmat of Messa
ge Impleme
n
tation Guid
e (MIG) DS
D
DIMENSIONS
Position in
key
Concept Scheme
Codelist
ID
Name
ID Scheme
ID CodeList
1 FREQ
Freque
nc
y
POP_C
O
NCEP
T
S
CL_FRE
Q
2
PROVINCE
Population
countr
y
POP_C
O
NCEP
T
S
CL_PR
O
VINCE
3 INDIC_P
O
P
Population
Indicator
POP_C
O
NCEP
T
S
CL_POP_I
NDIC
A
T
4 PO
OR_PO
P
PO
ORNESS
Me
asure
PO
P_CO
NCEP
T
S
CL_PO
P
_P
OO
R
TIME TIME_PERI
O
D
Time
Period
PO
P_CO
NCEP
T
S
MEASURES
TY
PE
Concept Scheme
Codelist
ID
Name
ID Scheme
ID CodeList
Primary
O
BS_VALUE
O
b
servation
value
PO
P_CO
NCEP
T
S
CS ID-AC
Aceh
CL_PR
O
VINCE
ID-BA
Bali
ID-BB
Bangka-Belitung
ID-BE
Bengkulu
ATTRIBU
T
ES
Attachment
Level
Concept Scheme
Codelist
ID
Name
ID Scheme
ID CodeList
O
b
servation
O
BS_STATUS
Status of t
he obs
ervation
POP_C
O
NCEP
T
S
CL_OBS_S
T
AT
US
Series UNIT
Unit
POP_C
O
NCEP
T
S
CL_UNI
T
Series TIME_FO
R
MAT
Time
Form
at
PO
P_CO
NCEP
T
S
CL_TIME_
FORMAT
We continu
e
pre
s
entin
g the re
sult with the
implem
ent
ation of web
s
ervice
provid
er which
can b
e
se
en i
n
Figure 10.
Figure 10. Impleme
n
tatio
n
example of web
servi
c
e p
r
ovide
r
We ve
rify the
stan
dardization a
nd
archit
ecture
d
e
si
gn
qua
ntitatively usin
g
simpl
e
Li
kert
scale by pre
s
enting the re
sult to the user/expe
r
t, wh
i
c
h in this case is the relev
ant person
nel
a
t
Directorate
of Statistical Di
ssemination, Statis
tics Indonesia (BPS).Forthe web serv
ice
impleme
n
tation we pe
rform both qu
an
titative
evaluation thro
ugh
Like
rt scale
and bla
c
k b
o
x
testing u
s
ing
soa
p
UI software. We obtai
ned that fo
r q
uantitative asse
ssm
ent th
e
result is 2
2
.3
in
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Data Excha
n
ge De
sig
n
wit
h
SDMX Form
at for In
teropera
b
ility Statistical
Data (Jaka Sem
b
irin
g)
351
Like
rt scal
e, whi
c
hme
a
n
s
very go
od
. Me
anwhile fo
r bl
ack box te
sti
ng resultusi
n
g so
ap
UI can
be
see
n
in Figu
re 11. We con
c
lud
e
that all of the desig
n
ed se
rvice p
e
r
form a
s
inten
ded.
Figure 11. Re
sult of black
b
o
x testing usi
ng so
ap
UI
4. Conclusio
n
In this p
ape
r we
re
port
o
n
the d
e
si
gn
of data
and
meta d
a
ta e
x
chan
ge u
s
in
g SDMX
format for sta
t
istical data a
t
Statistics Indone
si
a (BP
S
). We imple
m
ent the de
sign co
ncept wit
h
web services
.
Implementation at Statis
tics
Indone
s
i
a (BPS) is
c
h
osen as
a
c
a
s
e
s
t
udy to prove
the de
sig
n
concept. We
descri
b
e
thre
e mai
n
result
s. Th
e first
result i
s
on
th
e fram
ework
for
stand
ardi
zati
on of
stru
cture of
stati
s
tica
l publi
c
ation
s
data mo
del
with SDMX.
The
simple
case
we
pre
s
e
n
ted
ha
s b
een
abl
e to d
e
mo
nstrate the
adv
an
tageou
s
of th
e effort fo
r i
n
terop
e
rability
of
data a
nd m
e
tadata. Seco
nd result is
on the
de
si
g
n
archite
c
tu
re of data
sh
aring
mod
e
l. The
desi
gn is ba
sed on the user nee
ds whe
r
e we abl
e
to align the design with the long term ma
ster
plan of BPS.
Finally, the t
h
ird
res
u
lt is
onthe
implementation of
data
and metadata exchange
servi
c
e
u
s
ing
Service
Orie
nted An
alysis and
Desi
gn
(SOAD) meth
od. We
pre
s
e
n
t the
prototy
p
e
s
y
s
t
em implementation
with web s
e
rvices
.It is
shown through quantitativ
e Likert ass
e
ss
ment
and bla
ck b
o
x
testing that the
overall standard, arch
itecture
d
e
si
g
n
and imple
m
entation ha
ve
achi
evedthei
rrespe
ctive ob
jectives.
Referen
ces
[1]
Ho
xh
a J, Bra
haj A.
Op
en
Goverment D
a
ta o
n
the W
eb: A Se
mant
ic Appr
oac
h
. Internationa
l
Confer
ence
on
Emergin
g
Intel
lige
n
t Data a
n
d
W
eb
T
e
chnol
o
g
ies. T
i
rana, Alban
ia. 20
11: 1
07-1
13.
[2]
Bizer C, H
e
a
l
th T
,
Berners L
ee T
.
Linke
d
D
a
ta-T
he Stor
y
So F
a
r.
Intern
ation
a
l J
ourn
a
l
on S
e
mant
i
c
W
eb and Infor
m
ati
on Syste
m
s (IJSW
I
S)
. 2009; 5(3): 1-22.
[3]
Z
a
karia Z
.
Ke
y Performanc
e
Indicators (K
PIs) in th
e Pu
bli
c
Sector: A St
ud
y i
n
Ma
la
ysi
a
.
C
a
na
dian
Center of Scie
nce an
d Educ
a
t
ion
. 201
1; 7: 102-1
07.
[4]
Statistical Data
and Metad
a
ta
eX
c
h
a
n
g
e
. 201
2. sdmx.or
g
/pa
ge_
id=
10.
[5]
SDMX Sta
ndar
ds Section 1 -F
rame
w
o
rk. 2
0
1
3
.
ww
w
.
s
d
m
x
.
o
r
g
.
[6]
SDMX Sta
ndar
ds Section 1 -S
DMX-Inform
ati
on Mod
e
l. 20
1
3
.
ww
w
.
s
d
m
x
.
o
r
g
.
[7]
Erl T
.
Service
Oriented
Archit
ecture: C
once
p
t,
T
ehnolo
g
y
and
Des
i
gn.
N
e
w
Y
o
rk: Prent
ice H
a
ll
PT
R.
200
5.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 14, No. 2, May 2015 : 343 – 352
352
[8]
Sarno R, Her
d
iy
anti A. A Service Por
tfoli
o for an Enterpris
e
Reso
urce Pla
n
n
in
g.
Internatio
nal Jo
urna
l
of Computer S
c
ienc
e an
d Net
w
ork Security
. 201
0: 144-
156.
[9]
W
eb Services
Architecture. 2
004. http://
w
w
w
.
w
3
.
o
rg/T
R/
w
s
-arch/.
[10]
Vale S. GSBPM. http://
w
ww
1
.
unece.or
g.
[11]
Cross-Dom
a
in
Cod
e
List. 200
9.
w
w
w
.
s
d
m
x
.o
rg.
[12]
Ning
L, Bi
n
X,
Xin Z
.
Data
b
a
se C
onv
ersio
n
Base
d o
n
R
e
lati
onsh
i
p Sc
he
ma M
a
p
p
in
g
. International
Confer
ence
on
Internet T
e
chnolo
g
y
an
d A
ppl
icatio
ns (iT
AP). W
uhan. 20
11: 16-1
8
.
Evaluation Warning : The document was created with Spire.PDF for Python.