Internati
o
nal
Journal of Ele
c
trical
and Computer
Engineering
(IJE
CE)
V
o
l.
5, N
o
. 2
,
A
p
r
il
201
5, p
p
.
34
0
~
34
3
I
S
SN
: 208
8-8
7
0
8
3
40
Jo
urn
a
l
h
o
me
pa
ge
: h
ttp
://iaesjo
u
r
na
l.com/
o
n
lin
e/ind
e
x.ph
p
/
IJECE
Big Data: Challenges, Opportuniti
es and Cloud Based Solutions
Hami
d
B
a
ghe
r
i
*,
Ab
dus
al
a
m
Ab
dul
l
a
h
S
h
al
t
o
oki
*
*
*
University
of Kurdistan,
I
r
an
** University
of
Human Develop
m
ent, Ir
aq
Article Info
A
B
STRAC
T
Article histo
r
y:
Received Nov 20, 2014
Rev
i
sed
D
ec 31
, 20
14
Accepte
d
Ja
n 22, 2015
W
e
are
liv
ing
in
an
era
of
infor
m
ation exp
l
os
io
n. Th
ere
ar
e
cha
lleng
es
with
large and
complex amount of data gene
r
a
ted ever
y
day
b
y
social networks,
wikis, blogs, emails, tr
affic s
y
s
t
em, bri
dges, airplanes and
engine, satellites
and weather s
e
n
s
ors
.
90% of current dat
a
in the world has
been creat
ed in the
las
t
two
ye
ars
.
Our s
m
art planet
beco
mes more and more intelligent. B
e
sides
the ch
allenges p
o
sed b
y
such
vast amount
of data including stor
age, search
,
sharing,
anal
y
s
is, and
visualizati
on, th
ere ar
e also m
u
ch opportu
nities
for t
h
e
world as it beco
mes more and more digi
t
a
li
zed
.
This stud
y
pr
ese
n
ts Big Data
and highlights its key
con
cep
ts
a
nd s
t
at
e-of-the-
a
rt im
plem
ent
a
tio
n as
wel
l
as
research
challen
g
es and suggests res
earch d
i
rections for futur
e
. IT log
analy
t
ics, Fraud
detection pattern,
social media pattern and modeling an
d
management p
a
tterns ar
e some o
f
oppor
tunities.
Hadoop is a clou
d based
and
open source solu
tion for
Big Data Analy
t
ics which has been written b
y
java.
Hadoop solution
is currently
still immature
. In this paper, thr
e
e topics are
suggested for research dir
ection: Security
issues in
Big Data, co
ntext-
aware
inform
ation r
e
tr
i
e
val
,
and in
te
gr
ating onto
l
og
y
with Big
Data.
Keyword:
Big
Data an
alytics
C
l
ou
d c
o
m
put
i
n
g
H
a
doo
p
M
a
pR
ed
uce
Copyright ©
201
5 Institut
e
o
f
Ad
vanced
Engin
eer
ing and S
c
i
e
nce.
All rights re
se
rve
d
.
Co
rresp
ond
i
ng
Autho
r
:
Ham
i
d B
a
gheri
,
In
fo
rm
ation Te
chn
o
lo
gy
,
Kurd
istan
Un
i
v
ersity,
Ira
n, Kurdistan,
Sa
nanda
j
, Pas
d
ara
n
st
reet.
Em
a
il: h
.
b
a
gh
eri@uok
.ac.i
r
1.
INTRODUCTION
We are living in an era of inform
ation expl
os
ion which large scale am
ount of dat
a
is getting
i
n
creasi
n
gl
y
l
a
rge
r
beca
use
o
f
vi
rt
ual
wo
rl
d
s
, wi
ki
s,
bl
o
g
s
,
e-m
a
il
, onl
i
n
e gam
e
s, VoI
P
t
e
l
e
ph
one
, d
i
gi
t
a
l
p
h
o
t
os, i
n
stan
t
m
e
ssag
e
s (IM), tweets, traffic system
, b
r
i
d
g
e
s, airp
lan
e
s and
en
g
i
ne, satellites, weath
e
r
sens
ors
.
9
0
%
of
cu
rre
nt
dat
a
i
n
t
h
e
wo
rl
d
h
a
s bee
n
c
r
eat
ed
in th
e last two
years [1
]. T
h
ere are c
h
allenges
of
m
a
nagi
n
g
s
u
c
h
vast
am
ou
nt
o
f
het
e
r
oge
n
e
ou
s dat
a
fo
r exam
pl
e dat
a
vari
et
y
and v
o
l
u
m
e
and anal
y
t
i
cal
co
m
p
lex
ity. Bi
g
d
a
ta an
alytics h
a
s b
e
en
grown in
t
h
e last y
ears
[2-3
].
FaceBook acc
um
ulates huge
am
ounts of data with
about
800 m
i
llion users a
n
d billions
of
page
views every
day which cause m
a
ny
challenges to storing
and
processi
ng all these data. FaceBook needs
anal
y
t
i
c
s t
ool
t
o
m
i
ne and
m
a
ni
pul
at
e l
a
r
g
e am
ount
of
dat
a
(a
bo
ut
1
5
t
e
ra
by
t
e
s) e
v
ery
day
i
n
d
i
ffere
nt
l
a
ng
uage
s,
di
f
f
e
rent
t
i
m
es, fr
o
m
di
fferent
l
o
c
a
t
i
ons a
n
d f
r
o
m
di
fferent
pl
a
t
form
s.
In th
is section
b
i
g
d
a
ta ch
aracteristic, four t
y
p
e
s
o
f
an
alytics and
op
po
rt
un
ities to
create bu
sin
e
ss
v
a
lue will
be di
scus
sed
.
1.
1.
B
i
g D
a
t
a
c
h
ar
acteri
s
t
i
c
M
o
st
defi
ni
t
i
ons o
f
bi
g
dat
a
foc
u
s o
n
t
h
e
si
ze of dat
a
i
n
st
ora
g
e b
u
t
t
h
ere are ot
he
r im
port
a
n
t
at
t
r
i
but
es of bi
g dat
a
:
dat
a
va
ri
et
y
and dat
a
vel
o
ci
t
y
[2]. T
h
ese three
Vs
of
big data
(Volum
e, Variety and
Velo
city) are sh
own
in fi
g
u
re
1
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Big
Da
t
a
: C
hallen
g
e
s, Opp
o
r
tu
n
ities and
Cl
o
u
d
Ba
sed
So
lu
tio
ns
(
H
a
m
i
d
Ba
gheri
)
34
1
Fi
gu
re
1.
C
h
a
r
act
eri
s
t
i
c
s of
B
i
g Dat
a
[
1
]
There
i
s
va
ri
et
y
of s
o
urces i
n
bi
g
dat
a
f
o
r
exam
ple web sources i
n
cluding s
o
cial m
e
dia and logs
whi
c
h m
a
ke i
t
com
p
l
e
x. U
n
s
t
ruct
u
r
ed
dat
a
(f
or e
x
am
pl
e audi
o,
vi
de
o, T
e
xt
) a
nd
sem
i
-
s
t
r
uct
u
re
d dat
a
(f
o
r
in
stan
ce
XML, RSS feed
) is no
w
jo
in
ed
with stru
ctured
dat
a
. Acc
o
r
d
i
n
g t
o
M
i
cros
o
f
t
O
v
er
85
pe
rcent
of
dat
a
capt
u
red i
s
uns
t
r
uct
u
red
[
4
]
.
Vel
o
ci
t
y
or s
p
e
e
d f
o
r e
x
am
pl
e vi
de
o cam
era
scan
ni
n
g
i
n
a c
r
o
w
d
f
o
r r
eco
g
n
i
z
i
n
g
specific
face is
anothe
r c
h
arac
teristic of
big data.
1.
2.
Four
Type
s
of An
alytics
New a
n
al
y
t
i
c
s appl
i
cat
i
o
n f
o
r
exam
pl
e Vi
de
o a
nd a
u
di
o a
p
pl
i
cat
i
on a
r
e n
eeded
t
o
pr
oce
ss st
ream
i
n
g
b
i
g
d
a
ta. So
und
m
o
n
ito
r to
pred
ict eart
h
qu
ak
es an
d
sa
tellite i
m
ag
es to
reco
gn
ize cloud
p
a
ttern
s are ano
t
h
e
r
exam
ple whic
h ha
ve to be
ana
l
yzed.
Th
e term
“an
alytics” h
a
s fo
ur typ
e
s
[3
]: Qu
an
titativ
e Research and
Develo
p
m
en
t, Dat
a
Scien
tists,
Op
eration
a
l An
alytics, an
d
Bu
sin
e
ss In
tellig
en
ce and
Disco
v
e
ry. By p
u
t
tin
g
b
i
g
d
a
ta an
d
an
alytics to
g
e
th
er
we will d
i
sco
v
er m
o
st sig
n
i
fican
t resu
lts in
b
u
s
i
n
ess
v
a
lu
e.
In
t
h
e n
e
x
t
sectio
n
b
i
g
d
a
ta i
s
sho
w
n
as a sp
ecial
asset th
at pro
d
u
ces
b
i
g opp
ortu
n
ities fo
r bu
sin
e
ss.
1.
3.
Sma
r
ter
a
n
d Intellig
ent Planet
:
Big
Da
ta
Oppo
rtunities
B
e
si
des t
h
e chal
l
e
nge
s p
o
s
e
d by
suc
h
v
a
st
am
ount
o
f
dat
a
(B
i
g
D
a
t
a
), t
h
ere a
r
e
al
so
m
u
ch
o
ppo
rt
u
n
ities fo
r th
e
world
as it b
eco
m
e
s
mo
re an
d
m
o
re
d
i
g
italized
. For ex
am
p
l
e, in
fo
rm
atio
n
d
e
ri
ved
fro
m
digital records can m
a
ke doc
tors' j
o
b easier in accurately diagnosing an
d treating illne
sses, and bring down
h
ealth
care costs fo
r
p
a
tien
t
s, an
d th
e
ov
erall
q
u
a
lity an
d
effi
cien
cy of
h
ealth
care will b
e
im
p
r
o
v
e
d
[5-7
].
Our sm
arter p
l
an
et h
a
s b
e
come
m
o
re an
d
m
o
re in
tell
ig
en
t. Th
ere are so
m
e
b
i
g
o
p
portu
n
ities th
at
deri
ve fr
om
big dat
a
[
6
]
:
IT Lo
g anal
y
t
i
c
s, Fra
ud
det
ect
i
on pat
t
e
r
n
, soci
al
m
e
di
a pat
t
e
r
n
an
d M
o
del
l
i
ng a
n
d
m
a
nagem
e
nt
p
a
t
t
e
rns.
2.
CU
R
R
ENT
S
T
ATE A
N
D
R
ELEVANT
TOPIC
S
Big Data technol
ogies a
r
e not a re
placem
ent for current technologies
; they are a com
p
lem
e
nt [6].
Big
Data m
u
st b
e
in
teg
r
ated
with
th
e
rest
of en
terpri
se infrastru
c
tu
re. Besid
e
s th
e c
u
rre
nt
sol
u
t
i
o
n
s
f
o
r bi
g
data analysis, there a
r
e som
e
new c
h
allenge
s
, for in
sta
n
ce
need
for robus
t statistical
me
thods and m
a
naging
m
i
ssi
ng dat
a
[
8
-1
0]
. As m
e
nt
i
one
d i
n
sect
i
o
n
1 we
ha
ve
un
stru
ctured
and
stru
ctured
d
a
ta; in
teg
r
ation
b
e
tween
th
em
is an
o
t
her ch
alleng
e
[1
1
]
.
In th
is sectio
n
b
i
g
d
a
ta an
alytics so
lutio
n
,
t
o
o
l
s and tech
n
i
q
u
e
s will b
e
revie
w
ed
.
2.
1.
Clou
d Based
Big Data
Solu
tion
Clo
u
d
co
m
p
u
tin
g
prov
id
es
n
e
w cap
ab
ilities fo
r p
e
rfo
r
min
g
an
alysis acro
ss all
data in
an
or
ga
ni
zat
i
on.
I
t
uses
new
t
echni
cal
a
p
p
r
oac
h
es t
o
st
o
r
e, se
arch
, m
i
ne and
di
st
ri
b
u
t
e
m
a
ssi
v
e am
ount
s
o
f
dat
a
[6].
Problem
s suc
h
as la
rge-scale im
age
processi
ng,
s
e
ns
or
dat
a
c
o
rrel
a
t
i
o
n
,
s
o
ci
al
net
w
or
k a
n
al
y
s
i
s
,
encry
p
t
i
on/
dec
r
y
p
t
i
o
n,
dat
a
m
i
ni
ng,
si
m
u
lat
i
ons,
an
d
p
a
t
t
e
rn
reco
g
n
i
t
i
on ca
n
be
s
o
l
v
e
d
i
n
t
h
e
cl
ou
d
com
put
i
ng d
o
m
ai
n.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
IJEC
E V
o
l
.
5, No
. 2, A
p
ri
l
20
15
:
34
0 – 3
4
3
34
2
To cope with
problem
m
e
ntione
d above about F
aceB
ook,
Cloud allows Facebook to levera
ge m
o
re
than
8,500 Central Proce
ssing Unit (CPU) c
o
res a
n
d peta
byt
es of dis
k
space to create ri
ch data a
n
alytics on
a
wi
de ra
nge
o
f
busi
n
ess cha
r
a
c
t
e
ri
st
i
c
s.
N
e
w
cloud
com
p
u
tin
g
tech
no
log
i
es su
ch
as H
a
doo
p,
MapReduce and
BigTable are
driving analytic
trans
f
orm
a
tion in the
way orga
nizations
store
,
acce
ss a
n
d process m
a
ssive am
ount
s of disparate data
via
m
a
ssi
vel
y
para
l
l
e
l
and
di
st
ri
b
u
t
e
d
IT
sy
st
em
s. C
l
o
u
d
ap
pl
i
cat
i
on a
r
chi
t
ect
ures
are
base
d
on
t
w
o
pri
n
ci
pl
es:
Elasticity
: o
n
l
y u
s
e co
m
p
u
ting reso
urces
wh
en
n
eed
ed
.
Scalab
ility: h
i
g
h
l
y elastic in
frastru
c
ture t
o
resp
on
se ch
ang
i
ng
co
nd
itio
n su
ch
as d
a
ta
vo
lumes.
Researches
are
dri
v
ing across
the cloud
ecos
y
ste
m
fo
r s
o
m
e
reas
ons
. Fi
rst
of al
l
,
cl
ou
d
p
r
o
d
u
ces t
h
e
n
e
w an
alytic cap
a
b
ilities of
big
d
a
ta. Second
, it
p
r
ov
id
es
massiv
e
ly scalab
le an
alytics an
d th
ird
reason
is all
facilities
listed
ab
ov
e are com
b
in
ed
with
th
e security
an
d
fin
a
n
c
ial ad
v
a
n
t
ag
es of switch
i
ng
to
a clo
u
d
com
put
i
ng en
v
i
ro
nm
ent
.
Th
er
e ar
e ch
allen
g
e
s
w
h
en
dealin
g
w
ith
b
i
g
d
a
ta
o
f
v
o
l
umes g
r
eater
than
10
ter
a
b
y
tes. A
ltho
ugh
rel
a
t
i
onal
dat
a
base m
odel
s
a
r
e capa
b
l
e
of
r
u
nni
ng
i
n
a
Dat
a
C
l
ou
d,
m
a
ny cu
rre
nt
rel
a
t
i
o
nal
d
a
t
a
base
sy
st
em
s
fail in
th
e Data Clo
u
d
i
n
two
i
m
p
o
r
tan
t
ways:
M
a
ny
rel
a
t
i
o
na
l
dat
a
base
sy
st
em
s cann
o
t
sca
l
e t
o
s
u
p
p
o
rt
p
e
t
a
by
t
e
s or
g
r
e
a
t
e
r am
ount
s
o
f
dat
a
st
o
r
age
.
Whe
n
com
p
l
e
x
dat
a
i
s
no
rm
al
i
zed i
n
t
o
a
r
e
l
a
t
i
onal
t
a
bl
e
f
o
rm
at
im
pedance m
i
sm
at
ch ha
ppe
ns
.
Wh
en
d
a
ta is co
llected
,
often th
e
first step
is to tran
sfo
r
m
th
e
d
a
t
a
,
n
o
rm
al
ize th
e d
a
ta, and in
sert a
row in
to
a
rel
a
t
i
onal
dat
a
base.
Next
,
us
ers q
u
ery
dat
a
base
d o
n
key
w
or
ds o
r
p
r
e-l
o
a
d
ed sea
r
c
h
q
u
e
r
i
e
s and
wai
t
f
o
r
th
e resu
lts to retu
rn
.
On
ce ret
u
rn
ed
, u
s
ers
sift
th
ro
ugh
resu
lts.
2.
2.
Hadoop: Ope
n
Source
Heart
of Bi
g
Data
and Cl
oud Or
iented
Approach
Hadoop is a
Top level Apache project
ope
n s
o
urces
oftware
fram
e
wo
rk
th
at’s
wri
tten
in
j
a
v
a
pr
o
g
ram
m
i
ng l
a
ng
ua
ge [
9
]
.
I
t
enabl
e
s ap
pl
i
cat
i
ons t
o
wo
r
k
wi
t
h
t
h
o
u
sa
nds
o
f
com
put
at
i
onal
i
nde
pe
nde
n
t
com
put
ers an
d
pet
a
by
t
e
s of dat
a
. Ha
do
o
p
was de
ri
ve
d fr
om
Googl
e'
s M
a
pR
ed
uce an
d G
o
o
g
l
e
Fi
l
e
Sy
st
em
(
G
FS).
H
a
doop
h
a
s t
w
o p
a
r
t
s: a f
ile system (H
adoo
p D
i
strib
u
t
ed File Syste
m
o
r
H
D
FS)
an
d a
p
r
og
r
a
m
m
in
g
para
di
gm
(M
apR
e
d
u
ce)
. Tas
k
s s
u
c
h
as s
o
r
t
i
ng,
dat
a
m
i
ning
, i
m
age
m
a
ni
p
u
l
a
t
i
on,
s
o
c
i
al
net
w
o
r
k a
n
al
y
s
i
s
,
i
nve
rt
ed i
nde
x
con
s
t
r
uct
i
o
n
an
d m
achi
n
e l
ear
ni
n
g
a
r
e
pri
m
e jo
bs
f
o
r
M
a
pR
educe
.
HDFS is a d
i
stribu
ted
,
scalable, an
d
po
rtab
l
e
file syste
m
w
r
itten
in
Jav
a
fo
r th
e Hado
op
fram
e
wo
rk
.
HDFS st
o
r
es
larg
e
files acro
ss m
u
ltip
le mach
in
es.
It ach
i
ev
es
reliab
i
lity b
y
rep
licatin
g
th
e d
a
ta
acro
s
s
m
u
lt
i
p
l
e
host
s
.
HD
FS
was
des
i
gne
d t
o
ha
n
d
l
e
ve
ry
l
a
rge
fi
l
e
s.
Forrester re
ga
rds
Ha
doop a
s
the m
o
st si
gni
ficant
p
a
rt
of th
e
n
e
x
t
-g
en
eration
Enterp
rise Data
Ware
h
ousi
ng
(
E
D
W) i
n
t
h
e
cl
ou
d [
7
]
.
Ha
d
o
o
p
i
m
pl
em
en
t
s
the core fea
t
ures that
are
at the heart of
m
o
st
m
odern E
D
Ws
: cloud-facing
architect
u
r
es, i
n
-
d
at
abase
ana
l
y
t
i
c
s,
m
i
xed
wo
rkl
o
a
d
m
a
nagem
e
nt
and a
hy
b
r
i
d
stora
g
e layer.
3.
PROP
OSE
D
RESEA
R
C
H DIRE
CTIO
N
In th
is section
t
h
ree research directio
n
s
are
pro
p
o
s
ed
:
3.
1.
Sugges
t
ed T
o
pic 1:
Conte
x
t-Awar
e
In
fo
r
m
ati
o
n Re
trieval (I
R)
Searc
h
an
d R
e
t
r
i
e
val
wi
t
h
a h
uge am
ou
nt
of
st
ruct
u
r
e
d
an
d
unst
r
uct
u
re
d d
a
t
a
are affect
ed
by
C
ont
ex
t
i
n
m
a
ny
way
s
. Fo
r e
x
am
pl
e i
n
f
o
rm
at
i
on i
n
bi
g dat
a
c
o
nsi
d
e
r
ed a
s
a
som
e
t
h
i
ng dy
n
a
m
i
c over t
i
m
e and
changing ci
rcum
s
t
ances. For
sup
p
o
rt
i
n
g t
h
es
e dy
nam
i
c si
t
u
at
i
on c
ont
e
x
t
m
u
st
be appl
i
e
d t
o
searc
h
a
n
d
IR
a
n
d
new
f
r
am
ewor
k s
h
o
u
l
d
be a
p
pl
i
e
d.
3.
2.
Sugges
t
ed T
o
pic 2: Big
Data Sec
u
rity
Ch
allenges
To
day
,
dat
a
ba
se m
a
nagem
e
n
t
sy
st
em
s onl
y
su
p
p
o
r
t
sec
u
ri
t
y
pol
i
c
i
e
s at
f
i
ne
grai
n
l
e
vel
[
12]
fr
om
inappropriate access; while due to the le
ss structured a
nd i
n
form
al nature
of
big data curre
nt soft
ware
has no
suc
h
sa
feguards.
The future
of big
data will be in the cloud
but thes
e sol
u
tions als
o
com
e
with som
e
challenges s
u
ch
as security. In
Big Data Anal
ysis on cloud,
som
e
re
searches about Acce
ss Contro
l, encryp
tion
for tack
lin
g
securi
t
y
pr
o
b
l
e
m
and en
fo
rci
n
g secu
ri
t
y
pol
i
c
i
e
s
m
u
st
be d
one
. F
o
r d
e
fi
ni
ng
ne
w m
odel
s
and m
e
t
hods
we ca
n
follow “
D
ata Security as a
Service (DaS) “a
pproac
h.
3.
3.
Sug
ges
t
ed T
o
pi
c 3
:
In
te
gra
t
i
n
g O
n
t
o
l
o
g
y
w
i
th B
i
g
Da
t
a
An
al
yti
c
s
W
i
t
h
a h
u
g
e a
m
ount
of
dat
a
col
l
ect
ed by
Web 2
.
0
,
t
h
ere i
s
anot
her
fi
el
d f
o
r
researc
h
.
O
n
t
o
l
ogy
i
s
t
h
e
st
ruct
u
r
al
fram
e
wo
r
k
f
o
r
or
g
a
ni
zi
ng
i
n
fo
rm
at
i
on.
To
day
,
bi
g
dat
a
i
s
n
o
t
j
u
st
a
b
o
u
t
si
z
e
o
f
dat
a
. T
h
e
m
o
st
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Big
Da
t
a
: C
hallen
g
e
s, Opp
o
r
tu
n
ities and
Cl
o
u
d
Ba
sed
So
lu
tio
ns
(
H
a
m
i
d
Ba
gheri
)
34
3
i
m
p
o
r
tan
t
in
terest is d
i
gg
ing
and
an
alyzin
g
un
stru
ctured
d
a
ta.
For tak
i
ng
adv
a
n
t
ag
es of
o
ppo
rt
u
n
ities
m
e
ntioned i
n
pre
v
ious secti
o
n, Big Data
m
i
ght
bene
fit from
ontology technol
ogy and Ontology-base
d
analysis.
4.
CO
NCL
USI
O
N
There
are c
h
al
l
e
nges
wi
t
h
l
a
rge a
n
d com
p
l
e
x am
ount
of
dat
a
ge
nerat
e
d
every
day
by
so m
a
ny
di
ffe
re
nt
so
u
r
c
e
s an
d
fr
om
di
ffe
rent
pl
at
f
o
r
m
s. Accor
d
i
n
g
t
o
[1]
a
b
out
9
0
%
of
w
o
rl
d'
s dat
a
h
a
s
been
creat
ed
in
th
e last two
years. Ou
r sm
art p
l
an
et b
e
comes
m
o
re an
d
m
o
re in
tell
ig
ent. Besid
e
s th
e ch
allen
g
es po
sed
b
y
suc
h
vast
am
ount
of d
a
t
a
, t
h
ere are al
so m
u
ch
o
p
p
o
rt
uni
t
i
es for t
h
e w
o
rl
d as i
t
beco
m
e
s
m
o
re and
m
o
re
di
gi
t
a
l
i
zed. Th
i
s
st
udy
prese
n
t
s
B
i
g Dat
a
and hi
g
h
l
i
g
ht
s i
t
s
key
concept
s
and c
u
r
r
ent
ap
pr
oac
h
es as w
e
l
l
as
researc
h
c
h
allenge
s and
suggests three
re
search direc
tion
s
for fu
ture. IT l
o
g
an
aly
tics, Frau
d d
e
tectio
n
p
a
ttern, so
cial med
i
a p
a
ttern
an
d
m
o
d
e
lin
g
an
d
m
a
n
a
g
e
m
e
n
t
p
a
ttern
s are
so
m
e
o
f
o
p
portu
n
ities. Hadoo
p
is a
clo
u
d
b
a
sed
an
d
o
p
e
n
so
urce so
lu
tion
fo
r
Big
Data An
al
ytics wh
ich
is still
i
mmatu
re. In
th
is p
a
p
e
r, th
ree
to
p
i
cs ar
e suggested
f
o
r
r
e
sear
ch
d
i
r
ectio
n
:
Secur
ity issu
es in
Big
D
a
ta, co
n
t
ex
t-
aw
ar
e in
fo
r
m
atio
n
r
e
triev
a
l,
an
d in
tegrating on
to
log
y
with
Big
Data.
REFERE
NC
ES
[1]
Big Data, for b
e
tter or worse: 90
% of world'
s d
a
ta gener
a
ted over
last two
y
e
ars. S
c
ie
n
ceDaily
. R
e
trieved August 2
3
,
2013, from http://www.sc
iencedaily
.com- /rel
eas
es/2013/05/13052
2085217.htm
[2]
T Sutikno, D Stiawan
,
IMI Su
broto. Fortif
y
i
n
g
Big Data
infr
astructur
e
s to Face Secur
i
t
y
an
d Privac
y
Issue
s
.
TELKOMNIKA (Telecommunica
tion Computing
Electronics and
Control)
. 2014;
12(4): 751-752
.
[3]
Ne
i
l
Ra
d
e
n,
B
i
g
Da
ta Ana
l
yti
c
s
Archite
cture
,
20
12, Hired
Brains, Inc
[4]
Microsoft Big Data,
solution brief www.
microsoft.
com
[5]
Michael
Farb
er, MikeCameron, Christ
opher
Ellis , Massive Data
Analy
t
ics and
cloud, Booz Allen
Inc, 2011
[6]
C.Zikopou
lis, C
h
.Eaton, D.d
e
Ro
os, Unde
rstanding Big Data: An
aly
t
ics for En
ter
p
rise Class Ha
doop and Streaming
Data, Th
e McGr
aw-Hill Com
p
an
ies, 2012
.
[7]
The Forrester W
a
ve™: Enterpri
s
e
Hadoop
Solutions, Q1 2012
[8]
H. Dem
i
rkan, D. Delen
,
Lev
e
rag
i
ng the capab
ilit
ies of se
rvice-or
ient
ed decision support s
y
stem
s:
Putting anal
y
t
i
c
s
and big
data
in
cloud, Decis. Support
S
y
st. (2012
)
,
doi:10.1016
/j.d
ss.2012.05.048
[9]
http://hadoop
.ap
ache.org/
[10]
Jianqing Fan
,
H
a
n Liu
,
Sta
tist
i
c
a
l Ana
l
y
s
is of
B
i
g
Data on Phar
macogenomics,
Advanc
ed Dr
ug
Deliv
er
y R
e
v
i
ew
s
(2013), doi: 10
.1
016/j.addr.2013
.04.008
[11]
Managing Data
in Motion: Data Integr
ation
Bes
t
Pract
ice Techn
i
ques and
Techn
o
logies, First Ed
ition (2013)
125
-
128. doi:10.1016
/B978-0-12-397167-8.00018-2
[12]
Big Dat
a
: What
It Is
and Wh
y
You Should Car
e
. R
i
ch
ard
L
.
V
illars
. Car
l
W.
Olofson. Matth
ew Eastwood. June
2011
BIOGRAP
HI
ES OF
AUTH
ORS
Ham
i
d Bagheri rece
ived the B
.
S
.
and M.S. degrees in Software
engineering f
r
om the Shahid
Beheshti Univer
sity
,
Tehr
an, in 2
011.
Since 2009
, he has been
working in Informat
ion Technolog
y in Kurdistan
University
. His
res
earch in
ter
e
s
t
s
include S
e
rvi
ce Orient
ed Arhcit
ectur
eand
,
Big Data and Ult
r
a large S
c
ale
S
y
ste
m
s.
Abdusalam Abdullah Shaltooki r
eceived h
i
s M.S. degree in
software eng
i
n
eering from th
e
Sulaimniah Univ
ersity
, in 2011
.
His research
inte
rests include software Eng
i
neerin
g and Big
data
Anal
y
s
is
. He
is
working as
a l
e
c
t
urer in
th
e Univ
ersity
of Human
Development, Ir
aq.
Evaluation Warning : The document was created with Spire.PDF for Python.