TELKOM
NIKA
, Vol. 11, No. 9, September 20
13, pp.
5224
~52
2
8
ISSN: 2302-4
046
5224
Re
cei
v
ed Fe
brua
ry 24, 20
13; Re
vised
June 8, 201
3; Acce
pted Jun
e
20, 2013
Agricultural Knowledge Grid Construction
Tan Cuiping
*
, Zheng Hua
i
guo, Zhang Junfeng, Sun Sufen, Li Guangd
a
Institute of Information o
n
Sci
ence a
nd T
e
chnol
og
y of
Agric
u
ltura
l
, Beiji
ng
Academ
y of Ag
ricultura
l
an
d
Forestr
y
Scie
n
c
es, Beiji
ng, P. R. China
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: tcpspring
@
1
63.com
A
b
st
r
a
ct
In ord
e
r to
eli
m
inate
the
a
m
bi
guity i
n
s
e
mant
ic un
dersta
ndi
n
g
s d
u
rin
g
the
r
e
trieva
l of th
e
users, a
s
w
e
ll as
mi
ni
n
g
the r
e
lati
on
ship
betw
een
the co
nc
e
p
t of agric
ultur
a
l
know
le
dge, t
he ass
o
ci
ation
o
f
know
led
ge a
m
ong
hetero
g
e
n
eous
data
base
s
nee
ds to be
set up, w
h
ich
ena
ble
users t
o
discov
e
r us
e
f
ul
know
led
ge c
l
u
e
s, an
d gr
ad
u
a
lly for
m
sol
u
ti
on for
the
ulti
mate
q
uestio
n
.
Base
d o
n
th
e
char
acteristics
of
agric
ultura
l kn
o
w
ledge
an
d th
e ach
i
ev
e
m
ent
s of know
le
dg
e
grid r
e
se
arch,
w
i
th a co
mb
ina
t
ion of tra
d
itio
n
a
l
agric
ultura
l th
e
s
auri
an
d
onto
l
ogy
techn
o
l
o
g
y
, the a
g
ri
cu
ltural k
now
le
dg
e gri
d
w
a
s c
o
nstructed w
i
th
th
e
resourc
e
lay
e
r, semantic l
a
ye
r and us
er lev
e
l. It has
bee
n
appl
ie
d for se
ma
ntic exte
nsi
on o
n
retrieva
l
,
know
led
ge l
i
n
ks, and know
l
edg
e reas
oni
n
g
dia
g
n
o
si
s. It gains so
me
achiev
e
m
e
n
ts, w
h
ich provi
d
e
technic
a
l su
pp
ort and exp
e
ri
e
n
ce for the de
e
p
ly agr
icultur
a
l
know
led
ge ser
v
ices.
Ke
y
w
ords
: kn
ow
ledg
e gri
d
, agricult
u
re, onto
l
ogy
Copy
right
©
2013 Un
ive
r
sita
s Ah
mad
Dah
l
an
. All rig
h
t
s r
ese
rved
.
1. Introduc
tion
The te
rm G
r
i
d
of the
information te
chn
o
logy field
so
urces from
P
o
we
r G
r
id
[1]. The
Gri
d
techn
o
logy is an increa
sin
g
ly powful techn
o
logy
, an
d it has bee
n su
ccessfull
y
used to gri
d
comp
uting [2
-3], virtual organi
zation
s [4], etc.
The applicaiton for kno
w
le
dge m
anag
ement a
n
d
sha
r
ing p
r
o
d
u
c
ed a n
e
w
co
nce
p
t of Knowled
ge Gri
d
.
Fran
Berman
ea
rlier p
r
op
ose
d
the
con
c
ept
of Kno
w
ledge
G
r
id, K
nowl
edge
G
r
i
d
is a
n
intelligent int
e
rconn
ectio
n
environ
ment that enabl
es
u
s
ers o
r
virtual
roles to effe
ctively capture,
publi
s
h, sh
are and ma
na
ge kn
owl
edg
e re
sou
r
ces,
and othe
r service
s
for t
he users an
d to
provide th
e
requi
re
d knowl
edge
se
rvice
s
, su
pp
ort for
kno
w
led
ge inn
o
v
ation, and
work
together [5].
After that, se
veral auth
o
rs have given
the
definition
of the Kno
w
ledge
Grid,
such
as
Liang
hong
Di
ng [6], Jing
Li [7], and so on. Amon
g that, The d
e
finition by H. Zhuge i
s
m
o
re
clea
r. The Knowl
edge G
r
i
d
is an intelli
gent and
sustainable inte
rcon
ne
ction e
n
vironm
ent that
enabl
es p
e
o
p
le and ma
chine
s
to effectively captu
r
e, publish, share a
nd ma
nage
kno
w
le
dge
resou
r
ces. It also provid
es app
ro
priat
e
on-d
e
ma
n
d
servi
c
e
s
to supp
ort scientific re
sea
r
ch,
techn
o
logi
cal
innovation,
coo
perative teamwo
rk, problem
solvin
g, and d
e
ci
sion ma
king.
It
inco
rpo
r
ate
s
epistem
ology
and
ontolo
g
y
to refle
c
t
human
cogni
tive cha
r
a
c
te
ristics; expl
oi
ts
so
cial, e
c
olo
g
ical
and
economi
c
p
r
in
ci
ples; a
nd
ad
opts te
chni
qu
es a
nd
stan
d
a
rd
s d
e
velop
ed
durin
g wo
rk towa
rd the
future web [8-10].
Along with t
he develo
p
m
ent of agri
c
u
l
tural in
forma
t
ionizatio
n
in
China, a
g
ri
cultural
informatio
n reso
urce
s a
n
d
se
rvice
s
pl
atform
con
s
tru
c
tion h
a
s ma
de rema
rkabl
e achievem
e
n
ts.
A numbe
r of national, p
r
ov
incial a
nd mu
nicip
a
l agri
c
u
l
tural informa
t
ion sha
r
e
d
reso
urce
cent
er
have b
een
e
s
tablish
ed,
su
ch
as the
China Ag
ricu
ltural Sci
ence
Data Cente
r
, Beijing
agri
c
ult
u
re
digital inform
ation re
so
urce
cente
r
, Guan
gdo
ng
agri
c
ultu
ral i
n
formatio
n reso
urce
s sh
aring
platform,
Sichuan agri
c
ult
u
ral sci
en
ce and
te
chn
o
lo
gy literature i
n
formatio
n re
sou
r
ces
sh
ari
ng
platform.
Dat
a
content
cov
e
rs pla
n
tation
, bree
ding,
a
qua
culture,
bi
otech
nol
o
g
y, biosafety, food
safety, reso
u
r
ce
s an
d en
vironme
n
t, quality standa
rd
s, ag
ri
cultu
r
al zo
ning,
microbial
sci
ence,
and so on.
It is worth n
o
ting that the c
oncept of agri
c
ultural knowl
edge h
a
s o
b
vious a
m
biguit
y
, such
as the syn
o
n
y
mous relatio
n
shi
p
(cucum
ber, cu
ke
), n
o
matter the " cucu
mbe
r
" or " cuke" as
sea
r
ch term
s, you want to find the sam
e
cont
e
n
t, but in gene
ral, the
both could n
o
t be combi
n
ed
by the syste
m
. In addition, for the sa
me name,
different pe
ople
have different
comp
reh
e
n
s
i
ons,
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Agricultural K
nowl
edge
Gri
d
Con
s
tru
c
tio
n
(Tan
Cuipi
n
g)
5225
su
ch
as "ca
b
bage",
can
b
e
con
s
ide
r
ed
to be
"
Chi
nese
cab
bag
e " o
r
"
c
ole".
If usin
g
sim
p
ly
keyword
s
fo
r
retrieval, th
e
results are
n
o
t they
want
usu
a
lly. The
relation
ship
of
the
con
c
e
p
t
of
agri
c
ultu
ral
knowl
edge
is
very rich a
n
d
com
p
lex,
like the ralation
ship
of simil
a
rity, parent
-child,
host, feedi
ng,
bre
eding,
co
nstitution, introdu
ction, wi
t
h
refe
ren
c
e
to, etc. It is v
e
ry valua
b
le
for
u
s
er
s
,
w
h
ic
h is
imp
o
r
t
a
n
t k
n
ow
le
dg
e c
l
u
e
s
to
solve the pro
b
lem. But b
e
ca
use of the
hetero
gen
eity of ag
ricultural scie
nce
s
databa
se
, ea
ch scientific databa
se
is an
"info
r
mati
on
islan
d
", whi
c
h ma
ke
s th
e e
s
tablish
m
ent of th
e knowl
edg
e rel
a
tion
shi
p
bet
ween
the
hetero
gen
eo
us data
b
a
s
e
s
very difficult.
In ord
e
r to
so
lve the am
big
u
ity probl
em
on the
ret
r
iev
a
l of the
u
s
ers, an
d to mi
n
e
u
s
eful
kno
w
le
dge
relation
ship, agri
c
ultu
ral kno
w
le
dge
grid wa
s constructe
d based on
t
he
cha
r
a
c
teri
stics of agri
c
ultu
ral kno
w
led
g
e
.
It c
an provi
de kn
owle
dg
e clue
s, and
grad
ually led
the
us
ers
to the ultimate s
o
lution.
2. Rese
arch
Metho
d
In this
pape
r,
we
e
s
tablish
ed three
-
leve
l agri
c
ultu
ral
kno
w
le
dge
grid archite
c
tu
re, and
made
hete
r
o
gene
ou
s re
source
s, sema
ntic web
a
n
d
user envi
r
on
ment to
be
a
n
o
r
ga
nic wh
ole.
Based
on tra
d
itional ag
ricultural the
s
a
u
ru
s, we
u
s
e
d
ontology te
chn
o
logy to build con
c
ept
ual
relation
shi
p
diction
a
ry, a
nd imp
r
ove
d
ne
w
word
d
i
scovery m
e
chani
sm, en
ri
che
d
the
exi
s
ting
vocab
u
lary st
ructu
r
e a
nd relation
s. The
kno
w
led
ge g
r
id was a
ppli
ed for seman
t
ic extensio
n on
retrieval,
kno
w
led
ge lin
ks,
and
kno
w
led
ge re
asoni
n
g
diagno
si
s. In the pro
c
e
s
s
of building, t
he
job of indexin
g is so he
avy. To that effe
ct,
Visual Studio 2005 a
nd
Protégé
com
puter ap
plication
s
o
ftware was
us
ed.
2.1. Agricultural Kno
w
l
e
dge Grid Architecture
The architect
u
re
of agricultural
knowledge gr
id i
s
illustrated in Fi
gure
1. The
resource
layer i
s
used
to a
c
hi
eve t
he p
h
ysi
c
al
conne
ctio
n
bet
wee
n
the
a
g
ricultural
re
so
urces d
a
taba
se.
Heteroge
neo
us databa
se
integratio
n
e
ngine
i
s
ch
arge with
the
sha
r
ing
an
d
manag
eme
n
t of
hetero
gen
eo
us
datab
ase. Throug
h it, the differen
c
e
of info
rmatio
n re
so
urce
service
platfo
rm is
shiel
ded.
Figure 1. Agri
cultural Kno
w
ledge G
r
id Archite
c
ture
In the sem
a
ntic layer,
several di
ctio
nary an
d kn
owle
dge d
a
taba
se i
s
fou
nd with a
combi
nation
of
tradition
al agri
c
ultu
ral
th
esa
u
ri and
ontology tec
h
nology, which
is
us
ed for text
mining, retrie
val, and inte
lligent re
aso
n
ing. Based
on vocabul
ary and ve
ct
or spa
c
e m
o
del
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NIKA
Vol. 11, No
. 9, September 201
3: 522
4 – 5228
5226
(VSM), enco
d
ing identification, segm
e
n
tation, PO
S tagging, syn
t
ax analysis for farme
r
s’ local
verna
c
ula
r
ch
ara
c
ter i
s
rea
lized
by natu
r
al lan
gua
ge
retrieval
engi
ne. Co
ncept
ual rel
a
tion
sh
ip
diction
a
ry
rev
eals the
ba
si
c
relation
shi
p
of u
s
ing
of, o
n
be
half of, b
e
long,
divide
d with,
refe
re
nce
with.
Me
anwhile,
the oth
e
r relatio
n
shi
p
s of
bei
ng
suitabl
e to,
damagi
ng
on
or ho
sting
i
n
,
descri
p
tion, p
i
cture
s
a
nd video are cre
a
ted. Ag
ricultural p
r
ofe
ssi
onal kno
w
led
ge datab
ase for
rea
s
oni
ng
diagno
si
s is
made
of the
expert
s
’ ex
perie
nce a
n
d
kno
w
le
dge.
Knowl
edg
e
links
databa
se i
s
u
s
ed to reveal the referen
c
e
relation
ship
betwe
en hete
r
oge
neo
us d
a
t
abases.
In the user-l
evel, text mining en
gine
is
use
d
to mine the text content, and the
relation
shi
p
betwe
en texts, kno
w
le
dg
e eleme
n
ts,
and knowl
edge lin
ks of hetero
gen
eou
s
databa
se
s, which p
r
ovide
s
users the de
pt
h of agricultural kno
w
led
ge se
rvice
s
.
2.2. Ne
w
Wo
rd Discov
e
r
y
Method
s for
the Agric
u
ltural Kno
w
l
e
dge Grid
Becau
s
e
of the lon
g
prod
uction
cycl
es,
agr
i
c
ultu
ral concepts and relation
shi
p
s betwe
en
con
c
e
p
ts a
r
e
lagging
behi
nd in tra
d
itional ag
ricult
ural thesa
u
ri. I
n
ord
e
r to
so
lve it, based
on
natural l
ang
uage p
r
o
c
e
s
sing a
nd int
e
lligent ag
gregation te
ch
nology, this
pape
r extra
c
ted
keyword
s
fro
m
use
r
s’ qu
e
s
tion
s, sea
r
ch term
s, an
d
labels, foun
d netwo
rk h
o
t
word
s thro
u
g
h
automatically agg
reg
a
ting,
and
ad
d th
e ne
w
wo
rd
s and
the
co
nce
p
tual
rela
tionshi
ps to t
he
diction
a
ry.
Figure. 2 Ne
w Wo
rd
s Di
scovery Method
s
3. Results a
nd Analy
s
is
As a
re
sult, the
conte
n
ts
of this a
g
ri
cul
t
ural
kno
w
le
d
ge g
r
id
have
cove
red
veg
e
table
s
,
fruit tree
s, crops, livesto
ck, poultry, an
d aqui
cult
u
r
e
,
and it has
reveale
d
18
mainly sem
a
ntic
relation
shi
p
s.
200 agri
c
ul
tural hete
r
og
eneo
us data
base re
sou
r
ce
s are a
ssociate
d
in the
sema
ntic l
a
yer. As
an im
portant
com
p
onent of
th
e
agri
c
ultu
ral
kno
w
le
dge
service
platform,
agri
c
ultu
ral knowl
edge
gri
d
has b
een
u
s
ed to three
se
ctors. Firstly for t
he extensio
n of sea
r
ch
terms, the recall rate a
nd pre
c
i
s
ion
rate has b
een imp
r
ove
d
, Secondly
for the detailed
informatio
n di
splay, the kn
owle
dge lin
ks were co
nne
cted to one ne
t, which
can
recom
m
en
d the
kno
w
le
dge
cl
ues,
and
lea
d
the
user to
form the
en
d
solutio
n
. Thi
r
dly for the
ve
getable
di
sea
s
e
and pe
st diag
nosti
c syste
m
, the reasonin
g
ability was
signifi
cantly improve
d
.
3.1. Search
Term Cen
t
ra
l Extensions
Search te
rm
cent
ral ext
ensi
o
n
s
can
effe
ctively eliminate th
e
retrieval
a
m
biguity
probl
em
s, su
ch a
s
the u
s
e
r
input the
ke
y word
s of "g
reen fo
od", the system
will
recomme
nd t
he
simila
r word
s "natural", "o
rgani
c fo
od", the s
ubdivision word
s
of "Grad
e
A g
r
een
food "a
n
d
“Grade AA green food", an
d the asso
cia
t
ed words li
ke "banne
d pe
sticid
es", "limited pesti
cide
s".
It
not
only e
x
tends nea
r-synonym
s
an
d syn
onym
s
relation
shi
p
o
f
sea
r
ch te
rms, b
u
t al
so
will
guide the u
s
e
r
to further ex
plicit retri
e
val
need
s,
con
s
t
r
uct a mo
re rational search strate
gy, and
enha
nce user retrieval efficiency.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Agricultural K
nowl
edge
Gri
d
Con
s
tru
c
tio
n
(Tan
Cuipi
n
g)
5227
Figure 3. Search T
e
rm
Cen
t
ral Extension
s
3.2. Kno
w
l
e
d
g
e Links be
tw
e
e
n
Hetero
geneou
s Da
tabase
s
Knowle
dge a
s
soci
ation be
tween hete
r
o
gene
ou
s dat
aba
se
s is re
alize
d
by Agricultural
knowledge grid, whi
c
h can effect
ively i
m
prove the
user's knowle
dge
di
scovery capabilities.
For
example, "b
reedin
g
u
n
its"
of the ve
g
e
table va
rieti
e
s datab
ase
is rel
a
ted with
a
g
ri
cult
ural
institution
s
d
a
taba
se, and
"experts of the field"
of agricultural institutions data
base is relat
ed
with ag
ricultu
r
al expe
rts
d
a
taba
se. "the
appli
c
at
ion
of pesti
cide
s" of the veg
e
table vari
eties
databa
se
is related
with
registe
r
ed
pe
sticid
e d
a
tab
a
se, "m
anufa
c
ture
r"
of re
g
i
stere
d
p
e
sti
c
ide
databa
se i
s
related
with a
g
ricultural ent
erp
r
ises
data
base, and
so
on. Every kn
owle
dge
poin
t
is
not isolate
d
, and throug
h any kno
w
le
d
ge nod
e, the
use
r
s
ca
n en
ter into the whole ag
ri
cultu
r
al
kno
w
le
dge n
e
twork to find
the solution t
hey really ne
eds.
3.3. Vegetabl
e Disea
se an
d Pest Diag
n
o
stic
Sy
stem based on
the Agricultural Kno
w
l
e
dge
Grid
Based
on p
r
ofession
al kn
owle
dge d
a
ta
base and
co
nce
p
tual rela
tions di
ctiona
ry, it is
reverse rea
s
oning throug
h the different damage
ch
a
r
a
c
teri
stics of the five parts of the
vegetable
s
,
root, stem, l
e
af, flowe
r
, an
d fruit. Fi
nall
y
system
giv
e
s th
e
diagn
osi
s
by
judg
e
and
the value of possibility, and the user can
c
onfirm it by
diseases
and pests’ pi
ctures.
Figure 4. To
mato Disea
s
e
Diagn
osti
c System
Figure 5. To
mato Disea
s
e
Diagn
osti
c
Mech
ani
sm
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NIKA
Vol. 11, No
. 9, September 201
3: 522
4 – 5228
5228
4. Conclusio
n
Agricultural
knowl
edge
gri
d
is
one
exa
m
ple
of the
appli
c
ation
of grid te
ch
nol
ogy in
information processi
ng a
nd knowl
edge acquisition. It
is st
ill in the prelim
inary stage
of
developm
ent
due
to th
e
lack of in
d
e
xing. But
with the
d
e
velopme
n
t of
the
ag
ri
cu
ltural
informatio
nization in Chi
n
a, agri
c
ultura
l kno
w
led
ge
grid research
will be mo
re deeply. In the
future, it may be u
s
ed i
n
clou
d comput
ing, pe
st
and
dise
ase mo
nitoring, e
a
rly
wa
rning
of the
biologi
cal inv
a
sio
n
inform
a
t
ion monitori
n
g
, and so o
n
.
Referen
ces
[1] I
Foster.
T
he Grid: A New
Infrastructure for 21
st Century Scie
nce.
Ph
ysics T
oda
y. 20
02; (2)
:
42-47.
[2]
Chett
y
M, Bu
yya
R. W
eavi
n
g comp
utatio
n
a
l gri
d
s:
ho
w
ana
log
ous
are
the
y
w
i
th
ele
c
trical gri
d
s?
Co
mp
uting i
n
Scienc
e an
d Engi
neer
in
g.
20
02; (4): 61–
71.
[3]
Se
yedE
l
y
ar
Ha
shemser
e
sht,
Ali Asg
har
Po
urha
ji K
a
zem.
RDVBT
: Reso
urce D
i
stanc
e
Vector Bi
nar
y
T
r
ee Algorithm
for Res
ource
Discover
y
in
Grid.
IAES International Jo
ur
nal of
Artifici
al Intelligence.
201
2; 1(2): 45-
53.
[4]
Den
g
Gua
ng,
Liu Qi
ng
w
a
n
g
, Li Z
e
n
g
y
u
an,
Z
hang
Xu, Hu
ang Z
h
enc
hu
n. A Kin
d
of L
i
d
a
r App
licati
o
n
Grid Bas
e
d
on
eSci
ence
’
s V
i
e
w
.
T
E
LKOMN
I
KA Indo
nesi
a
n Jo
urna
l
of El
ectrical
Eng
i
n
e
e
rin
g
.
20
12;
10(5): 11
47-
11
50 .
[5]
Fran Berman.
From
T
e
raGrid to Kno
w
ledge Grid.
Commun
i
cations of the
ACM.
2001; 4
4
(
11): 27-2
8
.
[6]
Lia
nho
ng Di
ng and Xia
ng Li a
nd
Y
unp
en
g Xi
ng.
Pus
h
in
g
Sc
ientific Doc
u
ments
b
y
Discov
e
rin
g
Intere
s
t
in Informati
on
F
l
o
w
w
i
th
in E-
Scienc
e Kn
o
w
l
edg
e Grid.
Gri
d
an
d Co
op
era
t
ive Co
mputin
g
.
2005; 4
98-
510.
[7] Jing
Li.
A Me
mory
Base
d M
ode
l for Kn
ow
ledg
e Orga
ni
z
a
tion a
nd S
har
i
ng i
n
Kn
ow
led
ge Grid.
IFI
P
T
C
8 Publicatio
ns. 2007; 1
295
-129
9.
[8]
H Z
huge. Chi
n
a’s E-Scienc
e Kno
w
l
e
d
ge Gri
d
Enviro
nment.
IEEE Intellige
n
t System
s.
20
04; 19(1): 13-
17.
[9]
H Z
hug
e. Com
m
uniti
es an
d E
m
ergin
g
Sem
a
ntics in S
e
ma
n
t
ic Link
Net
w
or
k: Discover
y
a
nd L
ear
nin
g
.
IEEE Transactions on Kn
ow
le
dge a
nd D
a
ta Engi
neer
in
g.
2009; 21(
6): 785
-799.
[10]
H Z
h
u
ge. S
e
mantic
Link
in
g
throu
g
h
Spac
es fo
r
C
y
ber-
p
h
y
s
i
cal-s
o
ci
o I
n
telli
ge
nce: A
Method
olo
g
y
.
Artificial Intelligence.
20
11; 17
5: 988-1
0
1
9
.
Evaluation Warning : The document was created with Spire.PDF for Python.