Indonesi
an
Journa
l
of El
ect
ri
cal Engineer
ing
an
d
C
omputer
Scie
nce
Vo
l.
12
,
No.
3
,
Decem
ber
201
8
, p
p.
1265
~
1272
IS
S
N: 25
02
-
4752, DO
I: 10
.11
591/ijeecs
.v1
2
.i
3
.pp
1265
-
1272
1265
Journ
al h
om
e
page
:
http:
//
ia
es
core.c
om/j
ourn
als/i
ndex.
ph
p/ij
eecs
Effort E
stimati
on
in Tradi
tional and
Agile M
obile
Appli
cation
Develop
ment &
Testin
g
An
uree
t
Kaur
1
,
Kul
w
ant
Kaur
2
1
I.
K.G Punja
b
T
ec
hni
ca
l
Univer
s
ity
Kapur
tha
l
a, I
ndia
2
School
of
I
T,
A
pee
j
a
y
Insti
tute
of
Mana
gemen
t Te
chn
ic
a
l
C
ampus
Jala
ndhar
,
Ind
ia
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
A
pr
30
, 201
8
Re
vised
Ju
l
14
,
201
8
Accepte
d
Aug
2
1
, 201
8
Sm
art
phones/mo
bil
e
d
evi
c
es
are
endur
ing
a
ll
th
e
aspe
ct
s
of
hum
a
n
li
fe
.
W
it
h
the
significan
t
inc
re
ase
in
demand
for
appl
i
ca
t
ions
running
on
sm
art
phones/mo
bil
e
d
evi
c
es,
dev
el
oper
s
and
te
ste
rs
are
antici
p
ate
d
to
del
iv
e
r
high
qual
i
t
y
,
o
n
ti
m
e
and
withi
n
budget
applications.
Th
e
esti
m
at
ion
o
f
deve
lopment
an
d
te
st
ing
provid
es
a
b
ase
l
ine
an
d
act
as
a
tra
cking
gea
r
for
stake
hold
ers
and
deve
lope
rs.
Th
e
re
are
var
ious
a
pproa
che
s
for
esti
m
at
ion
of
tra
ditiona
l
software
deve
lopm
ent
.
But
m
obil
e
appl
icati
ons
ar
e
conside
re
d
diffe
ren
t
from
tra
dit
ion
al
softwa
re
such
as
those
running
on
desktop,
l
apt
o
p
or
on
th
e
web
.
Man
y
tra
d
it
ion
al
esti
m
at
ion
t
ec
h
nique
s
used
for
t
his
software
are
ad
apted
to
m
obil
e
dom
ai
n.
W
it
h
agi
l
e
software
dev
el
opm
ent
(AS
D)
m
et
hodolog
y
,
t
he
sce
n
ari
o
of
deve
lopment
a
nd
esti
m
at
ion
has
cha
ng
ed
dra
stically
and
so
as
m
obil
e
ap
p
deve
lopmen
t
and
esti
m
a
ti
on.
Thi
s
paper
provide
s
a
S
y
st
emati
c
Literatur
e
Review
(SLR
)
on
tra
d
it
ion
al
esti
m
at
ion
te
chn
ique
s
an
d
agi
l
e
est
i
m
at
ion
techni
ques
appl
i
ed
in
m
obil
e
software
/a
pp
li
c
a
ti
on.
Also
,
eff
o
rt
attributes
an
d
accura
c
y
par
amete
rs
fo
r
esti
m
at
ion
in
m
obil
e
a
pps
ar
e
p
rese
nte
d
.
How
eve
r,
to
d
at
e
,
there
are
v
e
r
y
fewe
r
studi
es
do
ne
on
the
m
obile
app
li
c
at
ion
est
imati
on
dom
ai
n
using
agi
l
e
m
et
hodolog
y
.
Ke
yw
or
d
s
:
Ag
il
e
dev
el
op
m
ent
Effor
t e
stim
at
i
on
Mob
il
e s
of
t
ware/
app
li
cat
ion
So
ft
war
e
engin
eerin
g
Copyright
©
201
8
Instit
ut
e
o
f Ad
vanc
ed
Engi
n
ee
r
ing
and
S
cienc
e
.
Al
l
rights re
serv
ed
.
Corres
pond
in
g
Aut
h
or
:
Anureet
K
a
ur,
I.
K
.G P
unj
a
b Tec
hnic
al
Univers
it
y Ka
purt
hala,
I
ndia
.
Em
a
il
:
anu
m
ah
al
@g
m
ai
l.co
m
1.
INTROD
U
CTION
Ba
ckgrou
nd
:
-
Be
cause
of
popula
rity
an
d
ease
of
us
e
of
m
ob
il
e
de
vices
has
beco
m
e
the
m
os
t
ind
is
pen
sa
ble e
xp
e
dient
for
hu
m
an
essenti
al
s f
r
om
p
ast
f
ew
ye
ars
[1
]
. M
obil
e so
ftware
d
e
velo
per
s
’
are
dri
ve
n
to
release
software
on
ti
m
e
a
nd
within
budg
et
.
Fo
r
this
p
ur
po
s
e
software
est
i
m
ation
pla
ys
a
ver
y
pivotal
ro
le
in
pro
vid
i
ng
t
he
m
os
t
accurate
siz
ing
figure
f
or
buil
ding
co
nfi
de
nc
e
in
de
velo
pe
rs
an
d
sta
ke
ho
l
ders
relat
ion
s
hip
[
2]. A
cc
ur
acy
i
n
e
stim
ation
is a c
ru
ci
al
f
act
or
f
or p
la
nn
i
ng s
of
t
war
e
de
velo
pme
nt to
e
va
de
bu
dg
et
ov
e
rru
ns
a
nd
ta
rd
y
delivery
of
s
oft
wa
re.
E
stim
ation
of
t
est
effor
t
al
ong
wit
h
de
velo
pm
ent
is
con
s
ider
e
d
cru
ci
al
f
or
a
ppreh
e
ndin
g
qual
it
y
so
ftwar
e
[
3]
.
Dev
el
opm
ent
of
m
ob
il
e
so
f
tware
is
co
ns
id
ered
diff
e
re
nt
f
ro
m
dev
el
op
i
ng
t
ra
diti
on
al
s
of
tw
are
pe
rta
inin
g
to
it
s
disti
nctive
featu
res
s
uch
as
li
m
it
ed
m
e
m
or
y,
pro
cessi
ng
powe
r,
sm
al
l a
nd m
ulti
ple input inter
face,
m
ulti
ple con
necti
on
s
, ban
dwidt
h fact
or, a l
ower
batte
ry, etc.
[4
-
7].
W
it
h
the
a
dvent
of
A
gile
So
ft
war
e
De
ve
lop
m
ent
(A
S
D)
[8
-
9],
e
ntire
softwa
re
de
velo
pm
ent
com
m
un
it
y
has
bee
n
dri
ve
n
t
o
the
a
doptio
n
of
a
gile
m
e
thodo
l
og
y.
The
A
gile
esp
ou
sal
t
o
m
ob
il
e
app
li
cat
ion
dev
el
op
m
ent
is
co
ns
ide
red
a
s
a
nat
ur
al
fit
by
m
any
research
e
rs
[10
-
14]
.
The
e
ffo
rt
est
i
m
ation
of
sof
tware
dev
el
op
m
ent
in
an
a
gile
en
vir
on
m
ent
is
al
so
dif
fer
e
nt
from
tradit
ion
al
so
ftwa
re
es
tim
a
ti
on
[
8].
Effor
t
est
i
m
ation
is m
or
e e
xig
e
nt in an ag
il
e b
eca
use
o
f
em
plo
yi
ng d
ynam
ic
ch
ang
es in
require
m
ents [
15]
. There are
m
any
est
i
m
ati
on
te
ch
niques
for
tra
diti
on
al
so
ft
war
e
est
im
at
ion
i
n
a
n
a
gi
le
en
vir
onm
e
nt
s
uch
as
pla
nn
i
ng
poke
r,
E
xpert
judgm
ent, U
se
Ca
se Po
i
nts (U
CP) Met
ho
d,
et
c. [8,
16].
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
12
, N
o.
3
,
Dece
m
ber
2
01
8
:
1265
–
1272
1266
Mob
il
e
ap
plica
ti
on
s
are
different
f
ro
m
tradit
ion
al
softw
are
[4
-
7]
and
agile
appro
ac
h
to
m
ob
il
e
app
li
cat
io
n deve
lop
m
ent and e
stim
ation
needs a s
olid m
et
ho
do
l
og
ic
al
a
ppr
oach f
or pre
dicti
ng
e
ffor
t.
The
Pro
blem
:
-
The
ap
proac
hes
us
e
d
f
or
e
stim
ation
of
tr
aditi
on
al
softw
are
are
ada
pted
for
m
ob
il
e
app
li
cat
io
n
de
velo
pm
ent.
Bu
t
pr
ese
ntly
,
not
m
uch
wor
k
ha
s
bee
n
ded
ic
at
e
d
to
i
den
ti
fyi
ng
s
uitable
ap
pr
oach
e
s
exclusi
vely
f
or
ef
f
or
t
est
im
at
ion
of
m
obil
e
app
s
[
17]
.
The
est
im
a
tio
n
te
ch
niques
in
a
gile
s
of
t
war
e
dev
el
op
m
ent
for
tradit
io
nal
so
ftw
are
ca
nnot
be
a
dapt
ed
to
m
ob
il
e
do
m
ai
n
pert
ai
nin
g
to
dif
fer
e
nt
char
act
e
risti
cs
of
m
ob
il
e
apps.
Howe
ver
,
th
ere
is
ver
y
le
s
s
li
te
ratur
e
ava
il
able
on
the
e
stim
ation
of
m
ob
il
e
apps in
an agil
e en
vir
on
m
ent
Pr
op
os
e
d
so
l
ution
:
-
T
his
pa
per
c
on
t
rib
utes
m
ai
nly
by
e
xam
ining
the
sta
te
-
of
-
art
of
te
chn
iq
ue
s
app
li
ed
f
or
est
im
at
ion
of
m
obil
e
so
ftware
/a
ppli
cat
ion
s
in
tr
aditi
on
al
s
of
tw
are
de
velo
pm
e
nt
an
d
agil
e
softwa
r
e
dev
el
op
m
ent
by
us
i
ng
Syst
e
m
at
ic
Lit
erat
ur
e
Re
view
(
SLR).
SLR
w
il
l
fo
rm
baseli
nes
f
or
m
ob
il
e
ap
p
dev
el
op
e
rs
for
the
sel
ect
ion
of
a
ppr
opriat
e
est
i
m
ation
m
et
hod
acc
ordin
g
to
thei
r
nee
d.
T
his
will
al
so
hel
p
researc
hers
in
fill
ing
the
ga
p
by
pro
po
si
n
g
f
or
m
al
est
i
m
ati
on
m
od
el
s
f
or
m
ob
il
e
app
s
co
ns
ide
rin
g
it
s
spe
ci
fic
char
act
e
risti
cs.
The
rest
of
th
is
pa
per
is
order
e
d
a
s
f
ollo
ws:
Sect
io
n
2
presents
the
Re
search
m
et
h
od
use
d
f
or
cond
ucting
t
he
Syst
e
m
at
ic
Lit
eratur
e
Re
vi
ew
(S
LR
).
S
ect
ion
3
prese
nts
the
re
s
ults
of
S
LR
an
d
so
m
e
discuss
i
ons
on
sel
ect
ed
stu
die
s and
finall
y, S
ect
ion
4
c
oncl
udes t
he pape
r wit
h fu
t
ur
e
dir
ect
ion
s.
2.
RESEA
R
CH MET
HO
D
The
goal
of
this
stud
y
is
to
gain
a
n
un
der
sta
nd
i
ng
of
the
current
s
ta
te
-
of
-
art
in
m
ob
il
e
app
dev
el
op
m
ent
effor
t
est
im
a
ti
on
.
T
o
this
e
nd,
an
ex
plorat
ive
stud
y
is
co
nduc
te
d
us
i
ng
t
he
Syst
e
m
at
ic
Li
t
eratur
e
Re
view
(S
LR)
su
ccee
ding
th
e
guideli
nes
postulat
ed
by
Kitc
henha
n
a
nd
Cha
rters
[
18]
.
SLR
is
a
re
searc
h
m
et
ho
d
f
or
car
ryi
ng
out
a
li
te
ratur
e
re
view
i
n
a
syst
em
a
ti
c
way
of
c
h
a
rtin
g
well
-
de
fine
d
ph
a
ses.
SLR
m
et
hod
us
es
thr
ee
pha
ses
for
perf
orm
ing
li
te
ratur
e
rev
ie
w
incl
ud
i
ng
Pla
nn
i
ng
a
nd
sp
eci
fyi
ng
re
search
qu
e
sti
ons
as
a
first
ph
a
se,
the
second
phase
is
condu
ct
in
g
the
rev
ie
w
that
com
pr
ise
s
an
identific
at
io
n
of
search
stri
ng
&
d
at
a
so
urces
, s
el
ect
ing st
udie
s &
dat
a extracti
on a
nd the t
hir
d one bein
g res
ults
repor
ti
ng.
2.1
.
Pl
an
nin
g Ph
as
e
Fo
r
the
sm
oo
th
c
onduct
of
s
yst
e
m
atic
li
te
r
at
ur
e
rev
ie
w,
pro
per
pla
nn
i
ng
is
neces
sary.
The
resea
r
c
h
qu
e
sti
on
s
d
e
riv
e from
the en
ti
re s
yst
em
atic literat
ur
e
re
view
p
la
nnin
g p
has
e.
2.1.1. Rese
arc
h Qu
es
tio
ns
(
Rq
s
)
Affirm
ing
the
researc
h
ques
ti
on
s
is
the
vi
ta
l
par
t
of
an
y
syst
e
m
at
ic
r
eview.
In
acc
orda
nce
with
gu
i
delines
pro
po
s
ed
by
Pett
i
crew
a
nd
Ro
be
rts
[19],
the
crit
eria
to
fr
a
m
e
research
quest
io
ns
are
ba
sed
on
PI
COC
(Popul
at
ion
,
In
te
rv
e
nt
ion
,
C
om
par
ison,
Ou
tc
om
e
s,
an
d
C
on
te
xt
).
I
f
the
rese
arch
quest
io
n
is
no
t
ou
tl
ine
d pro
perl
y, the li
te
rature re
view
m
ay
t
urn o
ut off t
he c
ourse. F
or thi
s
stu
dy PI
C
OC
are defi
ned as
:
-
Popu
la
ti
on: M
ob
il
e
Applic
at
ion p
r
oj
ect
s
In
te
r
ve
ntion
:
E
ffor
t est
im
at
io
n
te
ch
niques/m
et
hods
/p
ro
ce
ss
Com
par
ison
: T
rad
it
io
nal m
ob
il
e app ef
fort
es
tim
a
ti
on
tech
ni
qu
e
s w
it
h
a
n
a
gile m
ob
il
e ap
p
est
im
ation
.
Ou
tc
om
es:
Est
i
m
ation
m
od
el
s
to
fo
ll
ow
f
or
m
ob
il
e
app
dev
el
op
m
ent
in
agile
and
tradit
ion
al
softwar
e
dev
el
op
m
ent.
Con
te
xt: Re
vie
w
the
ex
ist
in
g st
ud
ie
s
on esti
m
at
ion
of m
obil
e app
s
.
The
researc
h
quest
io
ns
ste
er
t
he
e
ntire
syst
em
at
ic
rev
ie
w
m
et
ho
dolo
gy.
The
m
ajo
r
ob
je
ct
ive
of
this
fi
nd
i
ng
is
to answe
r
the
foll
ow
i
ng r
esea
rch q
uestio
n:
-
RQ1
.
What
are
the
tradit
io
nal
effor
t
est
im
a
tio
n
te
c
hn
i
qu
e
s
us
e
d
f
or
m
ob
il
e
so
ft
war
e/
a
ppli
cat
ion
de
velo
pm
ent
and test
ing
?
RQ2
.
What
is
currently
known
e
ffor
t
est
i
m
at
ion
te
chn
i
ques
f
ollo
wed
i
n
agile
m
ob
il
e
so
ft
war
e/
a
pp
li
cat
ion
dev
el
op
m
ent a
nd test
ing
?
RQ3
.
Wh
at
a
r
e
the
est
im
ation
at
trib
utes
a
nd
acc
ur
acy
pa
r
a
m
et
ers
us
e
d
i
n
est
im
ation
te
chn
i
qu
e
s
f
or
m
ob
ile
app
li
cat
io
n
?
2.2
.
Conduc
t
ing
t
he
Re
view
Pha
se
2.2.1 Se
arch
S
trateg
y
The
ai
m
of
directi
ng
searc
h
strat
egy
is
to
i
den
ti
fy
st
ud
ie
s
per
ta
i
ning
to
answer
t
he
R
Qs.
Furthe
r
,
search
strat
eg
y
can
be
c
on
du
ct
e
d
in
t
hr
e
e
ph
a
ses;
I
de
ntifyi
ng
keyw
ords
a
nd
De
fining
Sea
rch
s
tring
s
,
Id
e
nt
ify
ing t
he
d
at
a s
ources
a
nd Sea
rch Proc
ess in Dat
a s
ou
rces.
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
Eff
or
t Est
imati
on in Tr
ad
it
io
nal a
nd A
gile M
ob
il
e A
ppli
cation Devel
opme
nt
&
Testi
ng
(
An
ur
eet
kaur
)
1267
a)
Ident
if
yin
g
k
e
ywords
and
D
efining
Se
arch
St
ri
n
gs
The
fi
rst
phase
com
pr
ise
s
f
orm
ing
the
sea
rc
h
strin
g.
The
s
earch
st
rategy
is
set
up
t
o
des
cribe
sea
rch
strings
a
nd
pri
m
ary
data
s
our
ces.
T
he
gu
i
del
ines
prov
i
ded
by
[
18]
are
fo
ll
ow
e
d
t
o
def
i
ne
the
sea
rch
stri
ng
by
analy
zi
ng
t
he
m
ai
n
keyw
ord
s
in
RQ
s,
sy
nonym
s
of
the
ke
ywo
rd
s
an
d
on
a
ny
ot
her
s
pe
ll
ing
s
of
t
he
words.
The
i
d
entifi
e
d keyw
ords
a
re show
n
in
T
a
ble
1.
Table
1.
Id
e
nti
fied Key
w
ords
and
sy
nonym
ou
s
Key
wo
rds
Sy
n
o
n
y
m
o
u
s Te
r
m
s
So
f
tware
So
f
tware,
p
roject,
syste
m
,
app
licatio
n
Ef
f
o
rt
Co
st, r
eso
u
rce,
si
z
e,
m
etric
Esti
m
atio
n
Esti
m
atin
g
,
esti
m
a
te,
p
redictio
n
,
p
red
ictin
g
,
p
redict, ass
ess
m
en
t,
f
o
reca
stin
g
,
f
o
reca
st, calcula
tio
n
,
calculate,
calculati
n
g
,
sizin
g
,
m
easu
r
e,
m
easu
ring
Mob
ile App
licatio
n
Mob
ile so
f
tware,
Mob
ile App
s, M
o
b
ile pro
ject
Dev
elo
p
m
en
t
I
m
p
rov
e
m
en
t
,
Pro
g
ress
Testin
g
Test,
v
erifi
catio
n
,
v
alid
atio
n
Ag
ile
Scru
m
,
XP
,
lean
,
c
ry
stal
Metho
d
Proces
s, techn
iq
u
es,
m
o
d
els,
ap
p
roach
es
Ba
sed
on
the
i
den
ti
fie
d
keyw
ords,
the
sea
rc
h
stri
ng
is
obta
ined
by
j
oi
ning
syn
onym
ou
s
te
rm
s
us
ing
the
log
ic
al
‘OR
’,
oth
e
r
keyw
ords
us
i
ng
lo
gi
cal
‘A
N
D’
a
nd
wildcar
d
chara
ct
er
(´
*
´
).
Here
wildcar
d
cha
racter
represe
nts 0,
1, or a
ny num
ber
of alp
ha
nu
m
eric char
act
e
rs.
T
he follo
wing s
earch
strin
g
is
ob
ta
ine
d:
(“Mo
bile
App
li
cat
ion
”
OR
“M
ob
il
e
so
ft
w
are”
OR
“M
obil
e
App”
OR
“M
ob
il
e
project
”)
AND
(“D
e
velo
p*”)
AND
(“
est
i
m
a
te
*”
OR
“
pr
e
dict*”
OR
“as
sessm
ent”
OR
“forecast
*”
OR
“cal
culat
e*”
O
R
“si
zi
ng
”
OR
“
m
e
asur
e*”
)
A
ND
(“I
m
prov
e
m
ent”
OR
“P
rogr
e
ss”
A
N
D
“Pr
ocess”
O
R
“t
echn
iq
ues
”
OR
“
m
od
el
s”
OR “
appr
oach
es”
) AN
D ( “a
gile”
OR “sc
ru
m
” OR
”X
P”
O
R “l
e
an”
Or “cryst
al
”)
b)
Dat
a
s
ou
rc
es
The
di
gital
databases
that
a
re
us
ed
t
o
searc
h
the
keyw
ord
s
are
S
pr
in
ge
r
Link,
IEE
E
X
plore,
ACM
Digital
Lib
rar
y
, Elsevie
r
Scie
nce
Direct,
Wile
y In
te
r
Scie
nc
e an
d Goo
gle S
cho
la
r.
c)
Se
arch
Proc
ess in
Dat
a
s
ou
rces
The
ne
xt
ph
ase
is
to
ap
ply
the
searc
h
strin
g
t
o
c
hosen
el
ect
r
on
ic
data
sourc
es
to
fin
d
al
l
th
e
entai
le
d
stud
ie
s.
This
phase
is
di
vid
e
d
into
tw
o
s
ub
-
ph
a
ses:
pri
m
ar
y
and
sec
onda
ry
searc
h
phas
e.
I
n
Pr
im
ary
Searc
h
Ph
ase
,
the
el
e
ct
ronic
data
s
ources
id
entifi
ed
are
searc
he
d
ba
sed
on
th
e
searc
h
strin
g
def
i
ned
e
arli
er.
T
he
resu
lt
s
f
ro
m
data
so
ur
ce
s
are
m
on
it
or
e
d
to
include
sea
rc
h
string
i
n
ti
tl
e
and
a
bs
tract
s.
The
searc
h
stri
ng
is
again
ref
i
ned
e
ach
ti
m
e
to
check
the
outc
om
e
an
d
a
naly
zed
f
or
bette
r
res
ults.
T
he
ai
m
of
this
m
ai
den
che
ck
i
s
to
eval
uate
the
eff
ic
acy
of
t
he
searc
h
stri
ng.
A
ddit
ion
al
ly
,
res
ults
are
res
tric
te
d
to
peer
-
rev
ie
wed
c
onf
eren
c
e
pap
e
rs
a
nd
j
ou
rn
al
arti
cl
es
w
hich
are
a
vaila
ble
betwee
n
2008
a
nd
2018
(l
ast
decad
e
).
T
he
duplica
te
ti
tles
an
d
abstracts
a
re
r
e
m
ov
ed
.
I
n
th
e
seco
nd
a
ry
s
earch
phase,
a
te
chn
i
qu
e
cal
le
d
sno
wb
al
l
tracki
ng
is
us
e
d
f
or
stud
yi
ng
al
l
the
ref
ere
nces
of
pr
im
ary
stud
ie
s
to
exp
l
oit
further
st
ud
ie
s
a
nd
inc
rease
the
chan
ces
of
inc
lusio
n
of
im
po
rtant pa
per
s in
the sy
stem
at
ic
literat
ur
e re
view
. Ta
ble 2
li
sts t
he
ref
ine
d
res
ults fr
om
d
at
a so
ur
c
es after
pr
im
ary and
se
conda
ry searc
h p
hase.
Table
2.Ov
e
rv
i
ew of
searc
h re
su
lt
s
Data Sou
rces
Relev
an
t Sear
ch
R
esu
lts
Sp
ring
erL
in
k
53
IE
E
E
Xp
lo
re
75
ACM Digital
L
ib
r
ary
57
Elsev
ier
Sc
ien
ce D
irect
66
Go
o
g
le Scho
lar
83
W
ile
y
I
n
te
rScienc
e
20
Total
354
2.2.2 Incl
usio
n/E
xclusion
Cri
teria for
Sel
ecting
Studies
The
res
ults
acqu
i
red
th
rou
gh
sea
rch
stri
ng
def
i
ned
previo
us
ly
in
the
el
ect
ro
nic
da
ta
bases
ar
e
analy
zed
acco
r
ding
to
the
I
nc
lusio
n/Ex
cl
us
ion
c
rite
ria.
Th
e
evaluati
on
of
the
pap
e
rs
is
done
by
re
adi
ng
the
ti
tl
e
and
abstra
ct
first
and
c
he
cked
i
f
it
is
relat
ed
to
t
he
iss
ues
ad
dresse
d
in
RQs.
T
he
n
the
decisi
on
is
m
ade
for
it
s acce
ptan
ce f
or
rea
ding t
he whole
pap
e
r
or is
rej
ect
e
d
t
her
ei
n.
T
he
i
nc
lusio
n
crit
eria:
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
12
, N
o.
3
,
Dece
m
ber
2
01
8
:
1265
–
1272
1268
1.
Stud
ie
s
that in
cl
ud
e e
stim
at
io
n
m
et
ho
ds o
f m
ob
il
e app
li
cat
ion
dev
el
opm
ent AN
D;
2.
Descr
i
bed in
E
ng
li
s
h AND;
3.
Pe
er
-
re
viewe
d pap
e
rs
a
re s
el
e
ct
ed.
The
e
xclusi
on
crit
eria:
1.
Stud
ie
s
h
a
ving
m
ob
il
e app
li
c
at
ion
dev
el
op
m
ent p
r
ocess
a
nd not a
re
no
t
consi
der
i
ng est
i
m
ation
OR;
2.
No
t
desc
ribe
d
i
n
E
ngli
sh
OR;
3.
No
t
pee
r
re
vie
wed
2.2.3 Q
ua
li
t
y Assessme
nt a
nd Da
ta Ex
traction
To
e
valuate
th
e
qual
it
y
of
s
hortli
ste
d
stu
di
es;
6
qu
e
sti
ons
are
prepa
re
d
to
be
a
nswer
ed
for
eac
h
sel
ect
ed
stu
dy.
The
quest
io
n
can
be
a
nswer
ed
as
‘
Y’,
‘M
‘
or
‘
N’.
T
he
s
c
or
es
f
or
‘
Y
’
=
1,
‘P’
=
0.5
,
a
nd
‘N’
=
0.
T
he
quest
io
nn
ai
re
is
de
ve
lop
e
d
by
us
i
ng
the
guide
li
ne
s
de
fine
d
by
Kitc
heh
am
an
d
Cha
rters
[
18]
.
The
qu
al
it
y sco
re
f
or each
p
a
per s
hould be
m
ini
m
u
m
3
to
be f
urt
her incl
ude
d
i
n
the
stu
dy.
Fo
ll
owin
g
a
re t
he qu
e
sti
on
s
in
the
qu
e
sti
onna
ire:
1.
Ar
e
the
researc
h
m
otives clea
r
ly
stat
ed
?
2.
Was t
he
stu
dy
desig
ne
d
to
ac
h
ie
ve
t
he
ai
m
s
?
3.
Ar
e
the esti
m
ation
tec
hn
i
ques
well
d
e
fine
d
?
4.
Is
the
r
esea
rc
h process
doc
ume
nted
a
de
quat
el
y
?
5.
Ar
e
all
r
esear
c
h qu
e
sti
on
s
ans
wer
e
d
a
de
qu
at
el
y
?
6.
Ar
e
the m
ai
n
fi
nd
i
ngs stat
ed
c
le
arly
in
te
rm
s o
f
cre
ditabil
it
y,
v
al
idit
y, a
nd re
li
abili
ty
?
The
aut
hor
s
carried
ou
t
a
qu
a
li
ty
assess
m
ent
fo
r
the
al
l
sel
ect
ed
stud
ie
s.
Four
stu
dies
ar
e
exclud
e
d
owne
d
by
t
heir
low
-
qual
it
y
scor
e
.
Th
e
data
e
xtracti
on
phas
e
involves
e
xtr
act
ion
of
data
of
t
he
fi
nal
sel
ect
ed
stud
ie
s
that
ad
dr
ess
th
e
pec
uliari
ti
es
of
RQ.
The
data
extra
ct
ion
f
or
al
l
con
cl
us
i
ve
sel
ect
ed
stu
dies
is
do
ne
i
n
an
MS
Excel
s
heet co
ntaini
ng:
-
Pa
per nam
e, year
of
public
at
ion
, a
utho
r’ nam
e, an
d pa
pe
r
U
RL.
3.
RESU
LT
S
AND DI
SCUS
S
ION
The
re
su
lt
s
of
the
Syst
em
a
ti
c
Lit
eratur
e
Re
view
(
SLR)
a
nd
ans
we
rs
f
or
Re
search
Q
ues
ti
on
s
(R
Qs)
are
pr
e
sente
d
in
this
sect
ion
.
Table
3
desc
ri
bes
the
stu
dies
that
are
sel
ected
after
f
ull
screeni
ng
a
nd
pa
ssing
qu
al
it
y assess
m
ent crite
ria.
Table3
. S
el
ect
ed pape
rs f
or
fi
nal stu
dies
Stu
d
y
I
D
Ref
erence I
D
Year
Stu
d
y
I
D
Ref
erence I
D
Year
S
tu
d
y
I
D
Ref
erence I
D
Year
S1
[
2
0
]
2013
S2
[
2
1
]
2015
S3
[
2
2
]
2015
S4
[
2
3
]
2014
S5
[
2
4
]
2016
S6
[
2
5
]
2013
S7
[
2
6
]
2014
S8
[
2
7
]
2016
S9
[
2
8
]
2017
S1
0
[
2
9
]
2017
S1
1
[
3
0
]
2014
S1
2
[
3
1
]
2016
S1
3
[
3
2
]
2014
S1
4
[
3
3
]
2008
S1
5
[
3
4
]
2014
S1
6
[
3
5
]
2013
S1
7
[
3
6
]
2015
S1
8
[
3
7
]
2017
S1
9
[
3
8
]
2017
S2
0
[
3
9
]
2018
S2
1
[
4
0
]
2016
3.1
.
Tr
ad
itio
na
l T
ech
nique
s for Estim
ati
ng
M
ob
il
e Ap
pli
cat
io
n
Dev
e
lopmen
t: R
Q1
Sixteen
st
udie
s
out
of
twe
nty
-
on
e
sel
ect
ed
st
ud
ie
s
i
nvest
iga
te
d
the
t
rad
it
io
nal
est
im
a
ti
on
te
chn
iq
ues
for
m
ob
il
e
ap
pl
ic
at
ion
s.
Ta
bl
e
4
li
sts
the
id
entifi
ed
te
c
hn
i
qu
e
s
w
he
re
a
n
agile
ap
proac
h
is
not
f
ollowe
d
for
dev
el
op
m
ent
of
m
ob
il
e
ap
ps
.
The
te
ch
niques
a
r
e
broad
ly
cl
assifi
ed
int
o
t
hr
ee
cat
eg
or
ie
s
[
20
]
i.e
.
Algorithm
ic
-
ba
sed
m
od
el
s,
Ex
per
t
Jud
gme
nt
base
d
m
od
el
s
and
a
nalo
gy
based
m
od
el
s.
COSMIC
F
un
ct
i
on
Size
Me
asur
e
m
ent
[2
1
-
28]
is
fr
eq
ue
ntly
us
e
d
for
est
im
at
ion
te
chn
i
qu
e
wh
ic
h
is
us
ed
to
m
ea
su
re
th
e
functi
on
al
siz
e
of
the
m
ob
il
e
a
pp.
Othe
r
ty
pe
s
of
est
im
at
ion
te
ch
niques
i
de
ntifie
d
a
re
Fun
ct
ion
P
oi
nt
A
na
ly
sis
[29,
30]
and
Use
Ca
se
Po
i
nt
[31]
wh
ic
h
is
al
go
rithm
i
c
-
base
d
m
od
el
s
that
m
easur
e
fu
nctio
nal,
te
chn
ic
al
factors
an
d
en
vir
on
m
ental
factor
s
f
or
est
im
at
ion
.
R
eg
ressi
on
-
Ba
se
d
te
ch
nique
[
32
]
us
e
s
a
par
am
et
ric
m
od
el
base
d
on
effo
rt
pr
e
dictors
a
nd
data
po
i
nts
co
ll
ect
ed
through
an
onli
ne
ques
ti
on
nai
re
w
hich
are
f
ur
t
her
use
d
i
n
the
re
gressi
on
m
od
el
.
Delp
hi
m
et
ho
d
[
33
]
is
base
d
on
e
xpe
rience
t
o
est
im
at
e
the
e
ffor
t
wh
e
reas
A
rch
it
ect
ur
e
Ba
sed
est
i
m
at
i
on
m
od
el
[
34
]
for
reli
abili
ty
a
nd
te
sti
ng
est
im
at
ion
of
the
m
ob
il
e
app
li
cat
ion
is
pro
pose
d
an
d
the
case
stu
dy
was
co
nducte
d
in
tw
o
c
ompanies
.
Anothe
r
al
gorit
hm
ic
appr
oach
f
or
e
stim
ating
the
cost
of
dev
el
op
i
ng
A
ndr
oid
m
ob
il
e
app
s
are b
ase
d
on
t
he
COC
O
MO
–
I
an
d
II
m
od
el
[3
5]. Analo
gy
-
based
e
stim
ation
plu
s
fun
ct
io
nal
size m
easur
em
ent [
36]
app
r
oach is al
so p
r
opos
e
d for m
ob
il
e ap
ps
.
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
Eff
or
t Est
imati
on in Tr
ad
it
io
nal a
nd A
gile M
ob
il
e A
ppli
cation Devel
opme
nt
&
Testi
ng
(
An
ur
eet
kaur
)
1269
Table
4.
T
ra
diti
on
al
Esti
m
at
io
n
Tec
hniq
ues f
or m
ob
il
e app
li
cat
ion
s
Esti
m
atio
n
T
echn
iq
u
es
Stu
d
y
ID
Ap
p
roach
T
y
p
e
COSMIC
Fun
ctio
n
Size
Measu
re
m
en
t
S1
,S2,S3
,S4,S5
,S7
,S11
,S21
Alg
o
rith
m
ic
-
b
ased
M
o
d
el
Fu
n
ctio
n
Poin
t An
aly
sis
S6
,
S1
5
Alg
o
rith
m
ic
-
b
ased
Mod
el
Delp
h
i
m
eth
o
d
S9
Exp
ert
Ju
d
g
m
en
t
Use Cas
e Poin
t
S1
0
Alg
o
rith
m
ic
-
b
ased
Mod
el
Hy
b
rid (
An
alo
g
y
b
ased
esti
m
atio
n
+
Fu
n
ctio
n
Size
Measu
re
m
en
t
)
S1
3
An
alo
g
y
and
Algo
rith
m
ic bas
ed
m
o
d
el
Reg
ressio
n
-
Bas
ed
S1
2
Alg
o
rith
m
ic
-
b
ased
Mod
el
Architectu
re
-
Bas
e
d
S1
4
Alg
o
rith
m
ic
-
b
ased
Mod
el
COC
OMO
–
I
an
d
II
S8
Alg
o
rith
m
ic
-
b
ased
Mod
el
3.2
.
Ag
il
e
tec
hniques
f
or
es
tima
ting mo
bil
e a
ppli
ca
tio
n
develo
pmen
t:
RQ2
A
gile
ap
proac
h
to
m
ob
il
e
app
li
cat
ion
de
vel
op
m
ent
est
i
m
a
ti
on
has
ve
ry
le
ss
nu
m
ber
of
stud
ie
s.
On
e
of
t
he
reas
on
cou
l
d
be
t
he
adoptio
n
of
a
gile
to
m
ob
il
e
con
te
xt
is
sti
ll
in
it
s
evo
lv
ing
ph
a
se
a
nd
m
an
y
pr
act
it
ion
e
rs
a
r
e
ada
pting
t
radi
ti
on
al
est
i
m
at
i
on
m
et
hodo
l
ogie
s
to
an
a
gile
env
i
ronm
ent
fo
r
m
ob
il
e
ap
ps.
Th
e
identifie
d
st
ud
i
es
are
li
ste
d
in
ta
ble
5.
Tra
diti
on
al
us
e
case
po
i
nt
m
e
tho
d
of
est
i
m
at
ion
is
exten
ded
by
ad
ding
eff
ic
ie
ncy
an
d
risk
fact
or
of
te
ste
rs
in
the
agile
te
a
m
[3
7]
.
Anothe
r
te
chn
i
qu
e
[38]
is
based
on
a
ste
pw
is
e
li
near
re
gr
es
sion
m
od
el
w
hi
ch
est
im
a
te
s
t
he
ef
fort
f
or
Androi
d
ap
ps
from
req
uire
m
ents
sp
eci
fic
at
ion
includi
ng
a
nu
m
ber
of
use
ca
ses,
act
ors,
et
c.
User
st
or
y
po
i
nt
[
39
]
is
re
fi
ne
d
by
c
onsider
ing
a
dd
it
io
nal
factors
al
ong
with
siz
e
an
d
com
plexity
.
The
qual
it
y
factor, Novelt
y
facto
r
a
nd
Type
fact
or
of U
s
er
Sto
ry
are
ad
ded
to
deliver
the
bes
t
est
i
m
at
ion
for
m
ob
il
e
app
de
velo
pm
ent.
An
ot
her
a
ppro
ac
h
[40]
us
es
Ea
rly
Use
Ca
se
Po
int
(EU
C
P
)
an
d
E
xten
ded
Use
Ca
se
Po
int
(E
X
UCP)
al
on
g
wi
th
COCOMO
dr
i
ver
s
at
dif
fe
ren
t
it
erati
on
l
evels
in
agile
m
ob
il
e
app
de
velo
pm
ent.
An
ex
pe
rience
-
dr
i
ven
a
ppro
ac
h
us
in
g
t
he
Delp
hi
te
c
hniqu
e
[
41
]
is
use
d
for
effor
t est
im
at
io
n hav
i
ng a m
ob
il
e ap
p
as
one
of the
ca
se st
udie
s.
Table
5.
A
gile Est
i
m
at
ion
Techn
i
qu
e
s fo
r
m
ob
il
e ap
plica
ti
on
s
Ag
ile E
sti
m
atio
n
T
echn
iq
u
es
Stu
d
y
I
D
Ap
p
roach
T
y
p
e
Use Cas
e Poin
t
S1
6
Alg
o
rith
m
ic
-
b
ased
m
o
d
els
Step
-
wise
Linear
Reg
ressio
n
S1
7
Alg
o
rith
m
ic
-
b
ased
m
o
d
els
User sto
r
y
Poin
t
S1
8
Exp
ert
Ju
d
g
m
en
t
Use Cas
e Poin
t+ C
OCOMO
S1
9
Alg
o
rith
m
ic
-
b
ased
m
o
d
els
Delp
h
i
S2
0
Exp
ert
Ju
d
g
m
en
t
3.3
.
Estim
ati
on
Attribu
tes and
A
cc
urac
y Par
ame
ters
f
or Mo
bil
e A
p
ps:
RQ3
The
est
im
ation
at
tribu
te
s
ide
nt
ifie
d
in
the
sel
ect
ed
stu
dies
a
re
m
os
tl
y
fo
cu
sed
on
siz
e
m
e
tric
wh
et
her
base
d
on
us
e
case,
f
un
ct
i
on
po
i
nt
an
d
sto
ry
point.
Ta
bl
e
6
li
sts
the
oth
e
r
est
im
a
tio
n
at
tri
bu
te
s
t
hat
are
identifie
d
f
or
est
i
m
ation
.
Ta
ble
7
li
sts
the
pa
ram
et
ers
us
ed
to
assess
the
accuracy
of
e
stim
ation
of
m
ob
il
e
app
li
cat
io
ns
. M
MR
E an
d P
r
ed(x) are
h
i
gh
l
y fo
ll
owe
d
in
m
os
t of
the
stu
dies.
Table
6.
E
stim
at
ion
att
rib
utes
for
m
ob
il
e a
ppli
cat
ion
s
Esti
m
atio
n
Att
ribu
tes
Stu
d
y
I
D
Size
S1
,S2,S3
,S4,S5
.S6
,S7,S1
0
,S11
,S13
,S15
,S16
,S17
,S18
(us
er
sto
ries)
,S19
,S21
Co
st
S8
,S13
,S19
Oth
ers
S9
(Score
Metr
i
c),
S1
2
(M
ean an
d
SD
o
f
collected
m
o
b
il
e app
s v
ariables),
S1
4
(ar
ch
itectu
re
b
ased
),
S20
(
m
e
an
ef
f
o
rt
b
ased
on
ex
p
erience)
Table
7
.
Param
et
ers
f
or m
easur
in
g
the
accu
ra
cy
o
f
estim
at
ion
Accurac
y
Pa
ra
m
et
ers
Stu
d
y
I
D
MRE(
Magn
itu
d
e of
Relativ
e E
r
ror)
S2
,S3,S1
8
MM
RE
(M
ean Ma
g
n
itu
d
e o
f
Relativ
e E
r
ror)
S2
,S3,S1
2
,S18
,S19
MdMRE
(M
ed
ian
MRE)
S2
,S3, S12
Pred(p
erce
n
tag
e
re
lativ
e er
ro
r
d
ev
iati
o
n
)
S2
,S3, S12
,S18
,S19
Linear Regr
ess
io
n
(
R2
)
S1
2
,S19
No
t Def
in
ed
S1
,S4,S6
,S7,S9
S,1
0
,S13
,S14
,S20
,S2
1
Oth
ers
S8
(web
-
b
ased
su
rvey), S1
1
(Co
m
p
ar
e
d
with actu
al eff
o
rt)
,
S1
5
(Co
m
p
ar
ed
with
actual ef
f
o
rt)
,
S16
(Co
m
p
a
riso
n
with actu
al eff
o
rt)
,
S17
(co
m
p
ared w
it
h
so
u
rce
co
d
e as a
so
f
tware
m
easu
re)
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
12
, N
o.
3
,
Dece
m
ber
2
01
8
:
1265
–
1272
1270
4.
CONCL
US
I
O
N
This
pa
per
r
ep
r
esents
a
Syst
em
at
ic
L
it
erature
Re
view
(S
L
R
)
concer
ning ef
f
or
t
est
im
a
tio
n
te
c
hn
i
qu
es
in
m
ob
il
e
so
ft
war
e/
a
pp
li
cat
ion
s
t
hat
are
de
velo
ped
ei
the
r
us
in
g
tra
diti
on
al
s
of
twa
re
dev
el
op
m
ent
or
agile
so
ft
war
e
devel
op
m
ent
m
et
ho
do
l
og
y.
In
it
ia
ll
y,
the
searc
h
s
tring
pro
du
ce
d
a
total
of
35
4
resu
lt
s
an
d
only
23
p
ape
rs
a
re
sel
ect
ed
f
or
pri
m
ary
stu
dies.
T
wo
m
or
e
pa
pe
rs
a
re
incl
ud
e
d
a
fter
perf
orm
ing
sec
ondary
sear
c
h
process
but
f
our
pa
pers
are
e
xclu
ded
due
t
o
low
-
qu
al
it
y
scor
es
in
qual
it
y
assessm
ent
crit
eria.
T
he
fin
din
gs
of
this
SLR
are
founde
d
on
21
pa
per
s
w
hose
da
ta
are
extracte
d
into
a
n
excel
s
heet
an
d
the
n
e
xcav
at
e
d
to
a
nswe
r
the RQs
for
m
ed durin
g
t
he plann
i
ng phase
.
The
tra
diti
on
al
est
i
m
a
ti
on
te
chn
i
qu
e
s
ap
plied
in
m
ob
il
e
app
li
cat
io
ns
ar
e
COSMIC
F
un
ct
io
n
Size
Me
asur
em
ent,
Functi
on
P
oin
t
A
naly
sis,
Del
ph
i
m
et
ho
d,
U
se
Ca
se
P
oin
t,
Hyb
rid
(Analo
gy
base
d
est
i
m
at
ion
plu
s
Functi
on
Size
Mea
sure
m
ent),
Reg
ress
ion
-
Ba
sed
, Arc
hitec
ture
Ba
se
d
a
nd COC
OMO
–
I a
nd
II.
The
agile
est
im
at
ion
te
chn
iq
ues
in
the
m
o
bile
do
m
ai
n
are
ver
y
few.
The
identifie
d
te
chn
i
qu
e
s
are
adap
ti
on
or
m
od
i
fica
ti
on
of
tradit
ion
al
es
tim
a
ti
on
te
ch
ni
qu
es
us
e
d
in
the
agile
en
vir
on
m
ent
to
m
ob
il
e
so
ft
war
e
.
The
t
echn
i
qu
e
s
are
Use
Ca
se
Po
i
nt
,
Step
-
wise
Linear
Re
gressi
on,
Us
er
sto
ry
Po
int
with
a
dd
it
ion
al
factors,
Use
C
ase
Po
int
plus
COCOMO
a
nd
Delp
hi
te
chn
i
qu
e
.
SLR
al
so
pr
ese
nte
d
eff
ort
at
tribu
t
es
and
accuracy m
easur
es
fro
m
each
selec
te
d
stu
dy
us
e
d
f
or assess
ing
e
stim
at
ion
accur
acy
.
Fr
om
the
cur
r
ent
stud
ie
s,
it
can
be
co
nclu
ded
that
the
adoptio
n
of
a
gile
so
ftwar
e
de
velo
pm
ent
in
m
ob
il
e
do
m
ai
n
is
prolife
rati
ng
from
la
st
dec
ade
[42
-
43]
an
d
hen
ce
the
re
i
s
a
dire
need
f
or
f
or
m
al
est
i
m
at
ion
m
od
el
s
for
m
ob
il
e
softwa
re.
The
m
ob
il
e
s
oft
war
e
cha
ract
erist
ic
s
play
ed
a
crit
ic
al
r
ole
duri
ng
est
im
a
ti
on
in
tradit
ion
al
soft
war
e
de
velo
pm
ent
[4
-
7
]
,
[
44]
.
It
is
pr
ese
nt
ed
in
SLR
tha
t
none
of
t
he
e
stim
ation
m
odel
s
f
or
agile
m
ob
il
e
app
li
cat
io
n
devel
op
m
ent
are
c
og
it
at
ing
s
peci
fic
cha
racteri
st
ic
s
of
m
ob
il
e
a
pp
s
.
T
his r
esea
rch
ga
p
pro
vid
es
ideas for
resea
rchers
to
de
vise
nove
l
m
od
el
s
for
es
tim
a
ti
on
of
m
ob
il
e
app
s
i
n
agi
le
con
te
xt o
r
e
xten
d
existi
ng es
ti
m
a
ti
on
a
ppr
oach
e
s of esti
m
ation
.
ACKN
OWLE
DGME
NT
T
he
a
uthor
s
ar
e
than
kful
t
o
th
e
De
par
tm
ent
of
R
IC,
I.K.
G.
Punj
a
b
Tech
ni
cal
Un
i
ver
sit
y,
Ka
purthala
,
Punj
a
b, I
ndia
a
nd pr
ov
i
ding a
n opp
or
t
un
it
y t
o
ca
rr
y
ou
t t
his
r
esea
rch w
ork.
REFERE
NCE
S
[1]
I.
Mala
vol
ta
,
S
.
Rubert
o,
T
.
Soru,
and
V.
Te
rra
gn
i.
“
End
users’
per
ce
pt
ion
of
h
y
bri
d
m
obil
e
apps
in
the
Google
Pla
y
Store”
.
IE
EE
Int
.
Conf
.
Mob
il
e
S
erv
ices
,
New
York,
NY
,
US
A,
Ju
n.
/Jul. 2015, pp.
25
–
32.
[2]
Y.
M.
G.
Soare
s
and
R.
A.
d.
A.
Fagundes.
“Softwa
re
ti
m
e
e
stim
at
ion
using
reg
ression
m
et
hods”.
IEE
E
Lati
n
Ame
rican
Con
fer
enc
e
on
Compu
tat
ional Int
el
l
igence
(
LA
-
CCI)
,
Arequi
pa
,
2017
,
p
p.
1
-
6.
[3]
Pulak
Sahoo,
J.
R.
Mohant
y
,
“
E
arly
T
est
Eff
ort
Predic
ti
on
using
UM
L
Diagr
ams
”,
Indone
sian
Jo
urnal
of
El
e
ct
ri
c
al
Engi
ne
ering
and
Computer
Sc
ie
n
ce
(
IJE
ECS)
.
201
7;
5(1):
pp.
220
-
228
[4]
L.
Corra
l
,
A.
Sil
li
tti
and
G.
Succ
i.
“
Software
dev
el
opm
ent
proc
es
ses
for
m
obil
e
sy
stems
:
Is
agi
le
rea
l
l
y
ta
king
ov
er
the
business
?
”
,
1st
Inte
rnational
Workshop
on
th
e
Engi
ne
ering
of
Mobil
e
-
Enab
le
d
Syste
m
s
(
M
OBS)
,
San
Franc
isco,
CA,
2013,
pp.
1
9
-
24.
[5]
de
Souza
L.
S.
,
d
e
Aquino
G.S.
“Mobil
e
Applicat
ion
Esti
m
at
e
th
e
Design
Phase”
.
In:
Mac
ia
sz
ek
L.
,
Fil
ipe
J.
(
eds)
Ev
aluation
o
f
N
ove
l
Approach
es
to
So
ft
ware
En
gine
ering
.
Com
municat
ions
in
Computer
and
I
nformation
Sc
ience
,
vol
551.
Springe
r,
Cham
[6]
Inukoll
u
VN
,
Kesham
oni
DD
,
Ka
ng
T,
Inukoll
u
M.
“
Fact
ors
infl
uenc
ing
qua
li
t
y
of
m
obil
e
apps:
Role
of
m
obil
e
a
p
p
deve
lopment
li
f
e
c
y
cle”
.
arXi
v
pr
eprint
arX
iv
:141
0.
4537
.
2014
Oc
t
16.
[7]
Flora
HK
,
W
ang
X,
Chande
SV
.
“
An
inve
stiga
t
i
on
on
the
ch
aract
er
isti
cs
of
m
obil
e
appl
i
ca
t
ions:
A
surve
y
stu
d
y
”
.
Inte
rnational
Jo
urnal
of
Mod
ern
Educ
a
ti
on
and
Computer
Scien
ce
.
2014
Oc
t; 6(
6).
[8]
Muham
m
ad
Usman,
Emili
a
Me
ndes,
Fran
ci
l
a
W
ei
dt,
and
R
ic
a
rdo
Britto
.
2014
.
“
Eff
ort
esti
m
at
i
on
in
agi
l
e
soft
war
e
deve
lopment
:
a
s
y
stematic
l
itera
ture
r
evi
ew”
.
In
Proceedi
ngs
of
the
10
th
Int
ernati
onal
Con
fe
re
nce
on
Predictiv
e
Mode
ls i
n
Soft
w
are
Engi
n
ee
ring
(
PR
OMISE
2014)
.
ACM
,
New York,
NY
,
US
A,
8
2
-
91.
[9]
Raj
esh
H
Kulkar
ni,
P
Padm
ana
bham
,
Mana
si
Harshe,
K
K
Basee
r,
Pall
avi
Pat
il,
"Investi
gating
Agi
le
Adaptation
fo
r
Projec
t
Dev
el
op
m
ent
",
Inte
rnat
i
onal
Journal
of
El
e
ct
rica
l
and
C
omputer
Engi
ne
ering
(
IJE
CE)
,
2
017;
7
(3):
1278
–
1285.
[10]
Abraha
m
ss
on,
P.,
H
anhi
nev
a,
A.
,
Hulkko,
H.
,
Ih
m
e,
T
.
,
Jä
äl
ino
ja
,
J.,
Kork
al
a
,
M.
,
Kos
kel
a,
J.,
K
y
l
löne
n,
P.,
Salo
,
O.
“
Mobile
-
D:
An
Agile
appr
oa
ch
f
or
m
obil
e applic
at
ion
developm
e
nt
”.
In
pro
ce
ed
i
ngs of
OO
PSLA
’
04.
2004
.
[11]
Jeong,
Y.J.,
Lee
,
J.H.,
Shin,
G.S.
“
Deve
lopment
proc
ess
of
m
obi
le
appl
i
cation
SW
base
d
on
Agile
m
et
hodolog
y
”.
10th
Int
l. Conf
.
on
Adv
an
ce
d
Co
mm
unic
ati
on
Te
ch
.
,
pp.
362
-
366
.
2008.
[12]
Rahi
m
ia
n,
V.
Rams
in,
R.
“
Desi
gning
an
Agile
m
et
hodolog
y
for
m
obil
e
software
deve
lopment
:
A
hy
br
id
m
et
hod
engi
ne
eri
ng
app
roa
ch”.
2nd
In
ter
nati
onal
Conf
e
renc
e
on
R
ese
ar
ch
Chall
enge
s
i
n
Information
S
ci
en
ce
,
2008,
p
p.
337
-
342.
[13]
Schar
ff,
C.
,
Ve
rm
a,
R.
“
Scrum
to
support
m
obil
e
application
developm
ent
p
roje
c
ts
in
a
just
-
in
-
ti
m
e
le
arn
in
g
cont
ex
t”
.
In
Pro
c
.
o
f the
ICSE
W
orkshop CHASE
2010,
pp.
25
-
31
.
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
Eff
or
t Est
imati
on in Tr
ad
it
io
nal a
nd A
gile M
ob
il
e A
ppli
cation Devel
opme
nt
&
Testi
ng
(
An
ur
eet
kaur
)
1271
[14]
Cunha,
T
.
,
Dant
as,
V
.
,
Andrade
,
R.
“
SLeSS
:
A
S
cru
m
and
Le
an
Six Sigma
int
egr
a
ti
on
appr
oa
ch
fo
r
the
developm
ent
of
software
cust
om
iz
at
ion
for
m
obil
e
phon
es”
.
2
5th
Brazili
an
Sy
mpos
ium
on
Soft
ware
Engi
n
ee
ri
ng
,
pp.
283
-
292
.
2011.
[15]
Bini
sh
Ta
nve
er,
Li
l
ia
na
Guzm
án,
and
Ulf
Martin
E
ngel.
“
Under
standi
ng
and
improving
eff
ort
es
ti
m
at
ion
in
Agil
e
software
develo
pm
ent
:
an
indus
tri
al
ca
se
stud
y
”
.
In
Proceedi
ngs
of
the
Int
ernational
Confe
renc
e
on
Soft
ware
and
Syste
ms
Proce
ss
(
ICSSP
'16)
.
ACM
,
New York,
NY
,
US
A,
41
-
50.
[16]
B.
Praka
sh,
V
.
Visw
ana
tha
n.
A
,
“
Survey
on
Soft
ware
Esti
m
ation
Te
chn
ique
s
in
Tra
di
ti
ona
l
and
Agile
Deve
lopment
Models”.
In
done
sian J
ournal
o
f El
e
ct
rica
l
Eng
in
ee
ring a
nd
Computer
Sc
ie
nc
e(
IJE
ECS)
.
2017;
7(
3):
867
-
876
.
[17]
A.
I.
W
asserm
a
n.
“
Software
engi
nee
r
ing
issues
for
m
obil
e
application
develop
m
ent
”.
The
FSE/SDP
workshop
on
Fut
ure
o
f
softwa
re
engi
n
ee
ring
research
.
ACM
,
2010,
pp
.
397
–
4
00
[18]
B.
Kitc
henh
am,
S.
Chart
ers
.
“
Guidel
in
es
for
Per
form
ing
Sy
st
ematic
Liter
at
ure
R
evi
ews
in
Softw
are
Eng
ineeri
ng
”.
EB
SE
Technical
Re
port
,
EBSE
-
2
007
-
2101,
2007
[19]
Pett
ic
rew
,
M.
&
Robert
s,
H.
“
S
y
stemati
c
Reviews
in
th
e
Soci
al
S
ci
en
ce
s:
A
Pr
ac
t
i
ca
l
Guide”,
Bl
ac
kwe
ll
Pub
li
shing
.
2006.
[20]
E. Mende
s.
“
Cost E
stim
ation
T
e
chni
ques
for
W
e
b
Projects”.
IGI
G
lobal
.
2007
.
[21]
A.
Nitze. “
Mea
suring
m
obil
e
ap
pli
c
at
ion
size usi
ng
cosm
ic
FP
”. I
n
DASMA
Me
tri
k
Kongress
,
vo
l.
11,
2013
.
[22]
L.
D’A
vanz
o,
F.
Ferruc
c
i,
C.
Gr
avi
no,
and
P.
Sa
lz
a
.
“
Cosm
ic
fu
nct
ion
al
m
ea
surem
ent
of
m
obil
e
appl
icati
ons
an
d
code
size
est
imati
on
”.
In
Pr
oce
e
dings
of
the
30th
Annual
ACM
Symposium
on
App
li
ed
Computing
.
ACM
,
2015,
pp
.
1631
–
1636.
[23]
F.
Ferruc
ci
,
C.
Gravi
no,
P.
Sal
za
,
and
F.
Sarr
o.
“
Inve
stiga
ti
n
g
func
ti
onal
an
d
code
size
m
ea
sures
for
m
obi
le
appl
i
ca
t
ions:
A
r
epl
i
ca
t
ed
stud
y
”
.
In
Inte
rnat
ional
Conf
ere
nc
e
on
Product
-
Fo
cuse
d
Soft
ware
Proc
ess
Impr
ove
ment
.
Springer
,
2015
,
pp.
271
–
287
.
[24]
N.
A.
S.
Abdulla
h,
N.
I.
A.
Rusli, a
nd
M.
F.
Ibra
him.
“
Mobile
gam
e
size
esti
m
ation
:
Cosm
ic
fs
m
rule
s,
um
l
m
appi
ng
m
odel
and
unity3d ga
m
e engi
ne
”
.
In
Open
S
yste
m
s (
ICOS
)
, 2
014
IEE
E
Confer
enc
e
on.
IEEE, 2014, pp. 42
–
47.
[25]
A.
Sellam
i,
M.
Haoue
s,
H.
Ben
-
Abdall
ah
,
A.
A
bra
n,
A
.
L
este
r
huis,
C.
S
y
m
on
s,
and
S
.
Trud
e
l.
“
Sizi
ng
na
tur
al
la
nguag
e/
um
l
r
e
quire
m
ent
s for
web
and
m
obil
e appl
i
ca
t
ions u
sin
g
cosm
ic
fsm
”
.
Tech.
Re
p
.
,
2016
.
[26]
H.
van
He
eri
nge
n
and
E
.
Van
G
orp.
“
Mea
sure
th
e
func
t
ional
siz
e
of
a
m
obile
app
:
Us
ing
the
cosm
ic
func
ti
on
al
si
z
e
m
ea
surem
ent
m
et
hod”
.
In
So
ftw
are
Me
asur
eme
nt
and
the
In
t
ernati
onal
Con
f
ere
nce
on
Sof
t
ware
Proce
ss
a
nd
Product
M
easurement
(
IWSM
-
MENSURA
)
,
2014
Joint
Confer
ence
of
th
e
Int
ern
a
t
iona
l
W
orkshop
on.
I
EE
E
,
2014
,
pp.
11
–
16
.
[27]
La
udson
Silva d
e
Souza and
Gib
eon
Soare
s d
e
Aquino
Jr.
“
MEFF
ORTMO
B:
A E
ffort
Size Mea
sur
ement
for
Mobil
e
Applic
a
ti
on
Dev
el
opm
ent
”
.
In
te
r
nati
onal Journal
of
So
ft
ware
Eng
ine
ering
&
App
li
cat
ions (
IJS
E
A)
.
2014:
63
–
81.
[28]
Vogele
z
ang,
Fra
nk,
Ja
y
akumar
Kam
al
a
Ramasubram
ani
and
Sri
kant
h
Arvam
udhan.
“
Esti
m
at
io
n
for
Mobile
an
d
Cloud
Envi
ron
m
ent
s”.
Mode
rn
Soft
ware
Enginee
ring
M
et
hod
ologi
es
for
Mo
bil
e
and
Cloud
Env
ironmen
ts
.
IGI
Global
,
2016.
61
-
87.
[29]
T.
Preuss
.
“
Mobile
appl
i
ca
t
ions,
func
ti
on
al
an
aly
sis,
and
the
cu
stom
er
expe
rie
n
ce
”
.
The
IF
PUG
Guide
to
IT
and
Soft
ware
M
easurement
,
IFP
UG
,
Ed
.
Bo
ca Ra
ton
FL,
US
A:
Auerb
ac
h
Publ
ic
a
ti
ons
,
2012
,
pp
.
408
–
433.
[30]
Tuna
li
V
.
“
Software
Size
Estim
at
ion
Us
ing
Functi
on
Point
A
naly
s
is
–
A
Case
Stud
y
for
a
M
obil
e
Appli
catio
n”.
Muhe
ndisik v
e
T
ek
nolo
ji
S
empozyumu
.
2014
.
[31]
Marie
m
Haoue
s,
As
m
a
Sell
ami,
and
Hane
n
e
Be
n
-
Abdall
ah
.
“
A
rap
id
m
ea
surem
ent
pro
ce
dure
fo
r
sizi
ng
web
an
d
m
obil
e
appl
i
ca
t
i
ons
base
d
on
C
OS
MIC
FSM
m
et
hod”.
In:
P
r
oce
ed
ings
of
the
27th
Inte
rnati
onal
Workshop
on
Soft
ware
Me
asu
rement
and
12th
Inte
rnational
C
onfe
renc
e
on
Soft
ware
Proce
ss
and
Product
Me
asur
eme
nt
(
IWS
M
Me
nsur
a
'17)
.
ACM
,
New York,
NY
,
US
A,
129
-
137.
[32]
S.
A.
Shahwaiz
,
A.
A.
Mali
k
a
nd
N.
Sabaha
t
.
“
A
par
ametr
ic
e
ffort
esti
m
at
ion
m
odel
for
m
obil
e
apps”
.
In
19t
h
Inte
rnational
M
ult
i
-
Topic
Conf
e
renc
e
(
INMIC)
,
I
slamabad, 2016,
pp.
1
-
6.
[33]
Cat
oli
no
,
G.,
P.
Salz
a
,
Gravi
no,
C.
,
&
Ferruc
ci,
F.
“
A
set
of
me
trics
for
the
ef
fort
esti
m
at
ion
of
Mobile
apps”
.
Proce
ed
ings o
f
t
he
IE
EE
/
ACM
4
t
h
Int. Conf.
Mob
.
Sof
tw. E
ng
.
Sys
t
.
MO
BILE
Soft
.
2017.
[34]
W
adhwa
ni
V,
Mem
on
F,
Ha
m
ee
d
MM
.
“
A
rch
itect
ur
e
base
d
rel
ia
b
il
i
t
y
an
d
te
sting
est
imati
on
for
m
obile
appl
i
ca
t
ions”. In
Inte
rnat
ional
M
ult
i
Topi
c
Con
fer
enc
e
2008
Apr
11
(pp. 64
-
75). Springer, Be
r
li
n,
Heide
lb
erg
.
[35]
Zuba
ir
As
ghar
M,
Habib
A,
H
abi
b
A,
Rab
ail
Za
hra
S,
Ism
ai
l
S.
“
AndorEstim
at
or:
Android
-
b
a
sed
Software
C
ost
Esti
m
at
ion
Appl
ic
a
ti
on”
.
arX
iv p
reprint
arXi
v:1
6
05.
02304.
2016
.
[36]
A.
Nitz
e,
A.
Sc
hm
ie
te
ndorf,
R
.
Dum
ke.
“
An
A
nal
og
y
-
Based
Ef
fort
Esti
m
at
ion
Approac
h
for
Mobile
Applicat
io
n
Deve
lopment
Projects”.
In
Softw
are
Me
asur
eme
nt
and
the
Int
ernati
onal
Conf
ere
nce
on
Sof
t
ware
Proce
ss
and
Product
M
easurement
(
IWSM
-
M
ENSURA
)
,
2014.
[37]
Parve
z,
A.W.M.
M.
“
Eff
iciency
f
ac
t
or
and
risk
fa
ct
or
b
ase
d
user
ca
se
po
int
te
st
e
ffort
est
imati
on
m
odel
compati
b
l
e
with
agile
softw
are
d
eve
lopment
”.
In
Proceedi
ng
s
of
the
Inte
rnat
ional
Conf
ere
nc
e
on
Informatio
n
Technol
ogy
an
d
El
e
ct
rica
l
Eng
in
ee
ring
-
ICITEE
’13.
Yog
y
ak
art
a
,
Indone
sia
.
2013
.
pp
.
113
–
118
.
[38]
R.
Franc
ese
,
C.
Gravi
no,
M.
R
isi,
G.
Scanniello,
G.
Tort
or
a.
“
On
the
use
o
f
r
equi
rement
s
m
e
asure
s
to
pr
edi
c
t
software
project
and
produc
t
m
e
asure
s
in
the
co
nte
xt
of
Android
m
obil
e
apps:
A
pre
li
m
ina
r
y
stu
d
y
”
.
In
Proc
.
41
st
Euromicro
Conf. Se
r
ie
s So
ft
w
.
En
g.
Ad
vanced App
l
.
2015
.
pp
.
357
-
364.
[39]
W
.
As
la
m
,
F.
I
j
az
,
M.
I
.
La
l
i,
a
nd
W
.
Mehm
ood,
“
Risk
-
awa
re
and
qua
li
t
y
enr
i
che
d
eff
ort
estim
at
ion
for
m
obi
le
appl
i
ca
t
ions i
n
d
istri
bute
d
ag
il
e
s
oftwa
re
d
evelop
m
ent
”.
J. I
nf
.
S
ci.
Eng
.
2017
;
33(
6):
1481
–
1500
[40]
K.
Qi
and
B.
W.
Boehm.
“
A
li
ght
-
weight
in
crem
ent
al
eff
ort
est
imati
on
m
odel
for
use
ca
se
driv
en
proje
c
ts”.
IE
EE
28th
Annua
l
Sof
t
ware
Technol
og
y
Conf
ere
nc
e
(
STC
)
,
Gait
her
sbur
g,
MD
,
2017
,
pp
.
1
-
8.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
12
, N
o.
3
,
Dece
m
ber
2
01
8
:
1265
–
1272
1272
[41]
Lusk
y
M.,
Pow
i
la
t
C.
,
Böhm
S.
“
Software
Cost
Esti
m
at
ion
for
U
ser
-
Cent
ere
d
Mobile
App
Deve
lopment
in
Large
Ent
erp
r
ises”.
In
:
Ahram
T.,
Karw
owski
W.
(
eds)
Advanc
es
in
H
uman
Fac
tors,
S
oft
ware,
and
S
ys
te
ms
Engi
ne
erin
g
.
AH
FE
2017.
Ad
vanc
es
in
In
te
l
ligent
S
y
s
te
m
s a
n
d
Com
puti
ng,
vo
l
598.
Springer
,
Cham
[42]
Corra
l,
L.,
Sill
itti
,
A.
,
&
Suc
ci
,
G.
“
Agile
software
dev
el
opm
ent
proc
esses
for
m
obil
e
s
y
stems
:
Acc
om
pli
shm
ent
,
evi
den
ce
and
ev
olut
ion
”.
In
Proce
ed
ings o
f
the
Mobil
e
We
b
Info
rm
ati
on
Syste
ms
.
2013
.
90
-
106
.
[43]
Rit
a
Franc
ese
,
Carmine
Gravi
no,
Miche
l
e
Ri
si,
Giuseppe
Scanniell
o
,
and
Genove
ffa
Tortora.
“
Mobile
a
pp
deve
lopment
an
d
m
ana
gement:
result
s
from
a
q
ual
itati
v
e
inv
estigati
on
”.
In
Proc
ee
dings
of
the
4
th
Inte
rnat
ional
Confe
renc
e
on
Mobil
e
Soft
ware
Engi
n
ee
ring
an
d
Syste
ms
.
2017.
IEEE
Press
,
Pis
ca
t
awa
y
,
NJ
,
US
A,
133
-
143.
[44]
Kum
ar,
N.
A.,
Krishna,
K.
T
.
H.
,
&
Manjula,
R.
“
Chal
le
ng
es
and
best
pra
ct
i
ce
s
in
m
obil
e
applic
at
i
on
deve
lopmen
t
”
.
Imp.
J
.
In
te
rdisc
ip.
Re
s
.
2016
;
2(
12):
1607
–
1611.
Evaluation Warning : The document was created with Spire.PDF for Python.