Indonesi
an
Journa
l
of El
ect
ri
cal Engineer
ing
an
d
Comp
ut
er
Scie
nce
Vo
l.
13
,
No.
3
,
Ma
rch
201
9
, p
p.
109
5
~
110
1
IS
S
N: 25
02
-
4752, DO
I: 10
.11
591/ijeecs
.v1
3
.i
3
.pp
109
5
-
110
1
1095
Journ
al h
om
e
page
:
http:
//
ia
es
core.c
om/j
ourn
als/i
ndex.
ph
p/ij
eecs
Eva
lu
atin
g
g
ende
r
-
b
ased
m
obile
s
hopp
ing
a
pp
lication
u
sing
e
ye
-
t
rack
ing
t
ec
hn
ology
Aslina
B
ahar
um
1
, Emel
ia
A
bdul
Rahim
2
, Roz
ita
H
anap
i
3
,
Noorsidi
Ai
z
uddin Mat N
oo
r
4
,
Nu
rh
af
iz
ah
M
oz
iya
n
a Mohd
Y
us
op
5
1
Facul
t
y
of
Com
puti
ng
and
Infor
m
at
ic
s,
Univ
ersi
ti
Ma
lay
sia
Saba
h,
Mal
a
y
s
ia
2
,3
Facul
t
y
of
Bus
ine
ss
Mana
geme
nt,
Univ
ersit
i
T
e
knologi
MA
RA
Cawa
ngan
Sar
a
wak
,
Mal
a
y
sia
4
UTM CRE
S,
Fa
cul
t
y
of
Bui
lt E
n
vironment
and
Surve
y
ing, Univers
it
i
T
eknol
og
i M
al
a
y
s
ia,
Ma
lays
ia
5
Facul
t
y
of
Def
e
nce
Sc
ie
n
ce
and
T
e
chnol
og
y
,
Nat
iona
l
Defe
n
ce Unive
rsit
y
of
Ma
l
a
y
si
a, Ma
l
a
y
s
i
a
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
Oct
5
, 2
018
Re
vised Dec
6
,
2018
Accepte
d Dec
20
, 201
8
The
m
obil
e
applications
(ap
ps)
deve
lopment
h
as
bee
n
ga
ini
ng
po
pula
rity
fo
r
the
past
few
y
e
ars.
Som
e
m
obi
le
apps
have
b
e
en
fai
l
ing
in
it
s
popula
r
i
t
y
bec
ause
of
the
u
nat
tr
ac
t
ive
user
int
erf
ac
e
and
ba
d
int
er
ac
t
ion
wit
h
the
apps
.
Thi
s
was
prove
n
b
y
pre
vious
res
ea
rch
ers
who
have
stat
ed
th
at
th
e
int
er
ac
t
ion
bet
wee
n
a
user
of
m
obil
e
app
lication
and
a
gr
aphi
c
al
user
interfa
c
e
cou
ld
le
ad
to
som
e
m
i
sundersta
ndings,
err
ors
and
frust
rat
ion
from
an
i
nabi
lit
y
to
ac
hi
eve
a
go
al
a
nd
coul
d
lead
to
fai
lur
e
of the
m
o
bil
e appl
i
cation
.
Mos
t
of
the
m
obil
e
applic
ation
develope
rs
have
b
ee
n
f
a
ci
ng
prope
r
gr
aphi
c
al
user
int
erf
ac
e
d
esign
rec
en
tly
.
The
r
e
i
s
a
noti
ceabl
e
lack
of
studie
s
and
rese
arc
h
in
the
are
a
of
m
o
bil
e
application
design
co
m
par
ed
to
w
eb
app
l
ic
a
ti
on
user
int
erf
ac
e
(UI)
d
esign.
Therefor
e
,
thi
s
rese
ar
ch
was
done
to
eva
luate
the
diffe
ren
t
gende
r
on
the
m
obil
e
shopping
appl
ic
a
ti
on
using
e
y
e
-
tracki
n
g
te
chno
log
y
.
This
rese
arc
h
had
proposed
guide
l
ine
s
base
d
on
gende
r
fo
r
m
obil
e
shop
ping
appl
icati
on
.
Th
ese
proposed
guide
li
n
es
ca
n
be
pre
fer
red
in
the
designi
ng
pr
oce
ss
of
a
m
obil
e
shopping
appl
i
ca
t
ion
as
it
could
provide
a
bet
t
er
user
int
e
rac
t
ion
and
would
pre
vent
fa
i
lure
in
it
s
pop
ula
rity
b
y
providi
ng
f
ast an
d
user
fr
ie
ndl
y
i
nte
rfa
ce.
Ke
yw
or
ds:
Ey
e
-
t
rack
i
ng
t
echnolo
gy
Gende
r
Mob
il
e
s
hopping
a
pps
Copyright
©
201
9
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
:
Aslina Ba
haru
m
,
Faculty
of Com
pu
ti
ng
and
I
nfo
rm
atics,
Un
i
ver
sit
i M
al
ay
sia
Sab
a
h,
88400 K
ota
Kinab
al
u,
Saba
h,
Ma
la
ysi
a.
Em
a
il
: asl
ina
@u
m
s.ed
u.m
y
1.
INTROD
U
CTION
The
us
e
of
c
om
pu
te
rs
an
d
m
ob
il
e
phone
ha
s
bro
ught
great
co
nvenie
nces
and
c
on
tri
bu
ti
ons
t
o
s
ociet
y
in
daily
act
ivit
ie
s.
Both
c
ompu
te
r
syst
em
s
and
m
ob
il
e
ap
plica
ti
on
s
(app
s)
ha
ve
bee
n
buil
t
to
m
ake
peo
ple
,
even
m
or
e,
co
nv
e
nience
in
m
any
aspects.
W
it
h
c
on
ti
nu
in
g
pen
et
rati
on
of
t
he
I
nter
net
into
daily
li
fe,
it
ha
s
change
d
th
e
de
finiti
on
of
c
om
pu
te
rs
an
d
m
ob
il
e
ph
ones
,
an
d
bec
om
e
an
e
ssentia
l
par
t
of
the
pe
op
le
s
’
com
m
un
ic
at
ion
a
nd
daily
li
fe.
Wh
il
e
the
we
b
a
nd
m
ob
il
e
te
chnolo
gies
ha
ve
bec
om
e
ever
yday
li
fe,
in
or
de
r
t
o
face
the
ne
w
m
ark
et
env
i
ronm
ent
wh
ic
h
is
in
const
ant
c
hange,
t
he
co
m
pan
y
m
us
t
pl
ace
the
custom
er
in
the
centre
of
it
s
at
te
ntion
[
1].
D
evelo
ping
a
w
ebsite
and
m
ob
il
e
app
li
cat
io
n
m
akes
it
po
ssible
an
exce
ll
ent
com
m
un
ic
at
ion
with
the
cl
ie
nts,
an
d
this
le
ads
to
a
c
on
sta
nt
ada
ptati
on
of
the
co
m
pan
y’s
of
fe
r
to
the
con
ti
nu
ously
changin
g
c
us
t
om
ers’
re
quest
s
.
D
ue
t
o
this
,
the
avail
a
bili
ty,
an
d
a
de
quac
y
assessm
ent
i
n
this
sit
uation
a
re
be
com
ing
incre
asi
ng
ly
re
quire
s.
I
n
orde
r
to
stud
y
how
us
e
rs
inte
ract
wit
h
t
he
vis
ual
el
e
m
ents
within
t
he
we
bs
it
e
an
d
m
obil
e
app
s,
the
Ey
e
-
tracki
ng
te
chnolo
gy
is
i
ncr
easi
ng
ly
a
ppli
ed
to
t
he
usa
bili
ty
stud
y. T
oday
, m
any p
os
si
bili
ti
es already exi
st wh
e
re th
e
usa
bili
ty
o
f
we
bsi
te
s an
d
m
ob
il
e app
s
, am
on
g o
thers,
can
be
as
sesse
d
with
t
he
hel
p
of
Ey
e
-
trac
ki
ng
[
2].
Ey
e
-
t
r
ackin
g
m
easure
s
w
her
e
a
pe
r
so
n
is
l
ooking,
oft
en
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.
13
, N
o.
3
,
Ma
rc
h 201
9
:
109
5
–
110
1
1096
us
e
d
to
m
easur
e
how
us
e
rs
i
nteract
with
t
he
visu
al
el
em
e
nts,
wit
h
the
a
i
m
s
of
i
m
pr
ov
ing
it
s
desi
gn
and
avail
abili
ty
[3
]
.
It
sh
ows
where
the
us
e
r’
s
at
te
ntion
is
focuse
d
an
d
w
hich
path
s
are
f
ollow
e
d,
pro
vid
es
a
n
unobtr
us
ive m
eans t
o
e
xam
in
e co
gn
it
ively
a
nd att
entio
n
to
deal with
.
In
gen
e
ral,
di
f
fer
e
nt
ge
nd
e
r
has
di
ff
e
ren
t
pe
rsp
ect
ive
s
in
m
os
t
ac
ti
viti
es
includi
ng
on
li
ne
sho
pp
i
ng
and
us
e
r
interf
ace
desig
n
[
4].
Re
centl
y,
on
li
ne
sho
pp
i
ng
ha
s
grown
in
t
heir
popula
rity
in
li
ne
with
gro
wi
ng
of
the
I
nter
net
an
d
s
hoppin
g
m
ob
il
e
a
pps
on
sm
a
rtph
on
e
s
[
5].
Mo
bile
ap
ps
a
re
desig
ne
d
a
nd
de
velo
pe
d
by
dev
el
op
e
rs
a
nd
are
a
vaila
ble
in
ap
p
st
or
es
.
Mob
il
e
use
r
i
nt
erf
ace
desig
n
is
an
esse
ntial
in
the
m
ob
il
e
ap
ps
dev
el
op
m
ent
process
.
A
poor
use
r
inter
act
ion
with
m
ob
il
e
app
s
cou
l
d
le
ad
to
fail
ur
e
of
a
pp
s
[6
]
.
User
I
nterf
a
ce
desig
n
is
a
c
r
uc
ia
l
par
t
in
m
ob
il
e
ap
ps
a
nd
web
sit
e
de
velo
pm
ent
and
ha
d
cl
aim
ed
that
m
ental
m
od
el
theor
y
help
dev
el
op
e
r
s
in
desi
gn
i
ng
a
us
er
-
f
rien
dly
and
str
ong
vi
su
al
hiera
rch
y
us
er
i
nterf
ace
for
a
m
ob
il
e app
a
nd
web
sit
e
[7
]
.
In
rece
nt
ye
ars
,
HTML
5
ha
s
e
m
erg
ed
as
a
ve
ry
popula
r
wa
y
fo
r
buil
ding
m
ob
il
e
app
s,
c
al
le
d
m
ob
il
e
web
a
pps.
M
obil
e
web
a
pps
l
ooke
d
ve
ry
m
uch
li
ke
na
ti
ve
apps
but
buil
t
at
a
m
uch
faste
r
an
d
c
heap
e
r
r
at
e
in
HTML/
CSS.
T
he
m
ob
il
e
web
app
s
can
be
re
ached
the
wide
st
ran
ge
of
the
dev
ic
es
by
on
l
y
on
e
desig
n.
I
n
this
stud
y,
tw
o
different
s
hoppi
ng
m
ob
il
e
web
a
pp
inter
face
wi
ll
be
dev
el
op
e
d
base
d
on
the
e
xisti
ng
ge
nder
-
base
d
us
ers
’
m
ental
m
od
el
patte
rn
gu
i
delines.
Th
e
ey
e
-
tracki
ng
te
chnolo
gy
wil
l
be
us
e
d
to
ev
al
uate
the
de
ve
l
op
e
d
interfaces
.
By
analy
sing
pa
rtic
ipants’
ey
e
m
ov
em
ent
patte
r
n,
val
uab
le
i
nsi
gh
t
i
nto
us
a
bi
li
ty
and
oth
e
r
i
ssu
es
can
be
acq
uire
d
[
8].
Be
sides
,
the
val
uab
le
i
nfor
m
at
ion
li
ke
s
co
nclusi
on
s
about
the
pos
it
ion
ing
of
el
e
m
ents
al
so
ca
n be
dr
a
wn [8]
.
Nowa
days,
we
bs
it
e
an
d
m
obil
e
app
s
de
vel
op
e
r
sti
ll
facin
g
the
prob
le
m
of
li
m
it
ed
gu
ideli
nes
for
pro
per
us
er
i
nterf
ace
desig
n.
Vala
et
al
.
[6
]
and
[
9]
m
entioned
t
hat
m
os
t
of
t
he
m
ob
il
e
apps
de
velo
pers
ha
ve
been
facin
g
prop
e
r
grap
hical
us
er
inter
face
(G
U
I)
desig
n
recen
tl
y.
So
m
e
of
the
existi
ng
guideli
nes
f
or
GUI
desig
n
are
onl
y
descr
ibin
g
f
undam
ental
patte
rn
s
or
us
e
c
ases
[6
]
.
Vala
et
al
.
[6
]
m
ent
ion
e
d
that
ther
e
is
a
no
ti
ceable
la
ck
of
stu
dies
in
the
area
of
m
ob
il
e
app
de
sig
n
com
par
ed
to
web
ap
p
use
r
interface
(
U
I)
des
ig
n.
More
ov
e
r,
[
6]
and
[
9]
al
so
sta
te
d
that
the
bad
interact
io
n
betwee
n
a
us
er
’s
of
m
ob
il
e
app
s
a
nd
a
G
UI
cou
l
d
le
ad
to
so
m
e
m
isun
de
rstan
di
ng
s
,
er
rors,
a
nd
frust
rati
on
from
an
inabili
t
y
to
achieve
a
go
al
a
nd
c
ou
l
d
le
ad
to
fail
ur
e
of
th
e
m
ob
il
e.
Thu
s,
the
U
I
desi
gn
is
a
cru
ci
al
pa
rt
of
t
he
de
vel
op
m
ent
proces
s,
an
d
it
cannot
be
des
pised.
Be
si
des,
[
10]
su
gge
ste
d
that
on
li
ne
sh
op
ping
we
bs
it
es
shou
l
d
f
ocus
m
or
e
on
f
e
m
al
e
section
instea
d
of
m
al
e
sect
ion
,
as
t
heir
res
e
arch
res
ults
shows
t
hat
fem
ale
hav
e
a
hi
gh
e
r
r
at
io
t
o
sho
p
on
li
ne
.
I
n
c
ontrast
,
[11]
identifie
d
m
al
e
custom
er
s
do
m
or
e
on
li
ne
s
hoppin
g
an
d
outl
ay
m
or
e
m
on
ey
than
w
om
en,
and
t
he
y
do
or
are m
or
e li
kely
to
sho
p
onli
ne
in
the fut
ur
e.
Be
sides, [1
2] s
ta
te
d,
ev
e
n
th
ough m
os
t of
t
he
shoppin
g
is
done by
wo
m
en,
on
li
ne
pur
c
hasin
g of
t
en
to
b
e
dom
inance
by m
al
e c
on
s
um
ers.
As a
r
es
ult, d
if
fer
e
nt
u
sers
’
e
xpect
at
ion
a
m
on
g
us
e
rs
in
onli
ne
s
hoppin
g
is
sign
i
ficantl
y
aff
ect
ed
by
ge
nder
.
Th
us
,
the
inte
rf
ace
desig
n
f
or
t
he
sh
op
ping a
pps
and w
e
bs
it
e s
houl
d no
t
f
ocu
s
on a si
n
gle
ge
nder
, but
nee
d
t
o
c
onsider
bot
h gende
r.
2.
METHO
D
OL
OGY
To
bii
ProX
2
-
30,
a
m
ob
il
e
sta
nd
de
vice
a
nd
a
Le
novo
y
400
la
ptop
fro
m
SD
S
Associ
at
es
has
bee
n
us
e
d
in
this res
earch
.
To
bii
Pro
X
2
-
30
is
a
sc
reen
-
ba
sed
ey
e
tracker
t
hat
ca
ptures
gaze d
at
a
at
30
H
z,
c
om
pac
t
and
af
fordable
syst
e
m
that
is
perfect
f
or
stu
dies
outsi
de
of
the
la
b
an
d
it
has
been
de
sig
ned
to
giv
e
us
instant
insig
hts
into
visu
al
at
te
ntio
n
in
high
-
le
ve
l
fixati
on
ba
sed
re
searc
h.
The
ey
e
-
trac
ki
ng
e
xperim
ent
was
cond
ucted
at
Un
i
ver
sit
i
Te
knol
og
i
Ma
ra
(
UiTM),
Me
la
ka
,
Jasin
,
a
publ
ic
un
ive
rsity
and
U
niv
e
rsity
Tena
ga
Nasio
nal
(Uni
TEN
),
a
pr
i
va
te
un
ive
rsity
.
A
total
of
31
par
ti
ci
pan
ts
was
invol
ved
in
the
ey
e
-
trackin
g
exp
e
rim
ent.
Th
e
par
ti
ci
pa
nts w
ere r
an
dom
s
tud
e
nts
w
ho
w
ere
pur
su
i
ng
und
e
r
gr
a
du
at
e pro
gr
am
s
from
Faculty
of
C
om
pu
te
r
Scie
nce
an
d
Ma
them
a
ti
cs,
UiTM
an
d
Coll
ege
of
C
om
pu
te
r
Scie
nce
an
d
Inform
ation
Tech
no
l
og
y,
U
niTEN
.
T
he
ey
e
-
tracki
ng
e
xperim
ent
at
UiTM
was
co
nduc
te
d
in
a
le
ct
ur
er'
s
room
.
Whereas
,
at
Un
iTEN
,
it
was
co
nducted
in
a
co
m
pu
te
r
la
b
.
The
ey
e
tracker
a
nd
m
ob
il
e
sta
nd
dev
ic
e
in
was
set
up
base
d
on
the
trai
ni
ng
pro
vid
e
d
by
S
DS
Associ
at
es
Sdn.
Bh
d
a
nd
li
gh
ti
ng
c
ondi
ti
on
s
were
c
ontrolle
d
i
n
a
wa
y
that
do
e
s
not
interf
ere
with
the
ey
e
tracker
.
Parti
ci
pan
ts
we
re
re
cru
it
ed
with
th
e
help
of
th
e
le
ct
ur
er
s
at
bo
t
h
of
the
un
i
ver
sit
ie
s
for
the
us
er
inte
rf
ace
te
sti
ng
of
adopted
Laz
ada
shoppin
g
m
ob
il
e
app
li
cat
ion
.
Th
e
par
ti
ci
pan
ts
who
ag
ree
d
to
involve
in
ou
r
researc
h
we
re
first
giv
e
n
a
con
s
ent
f
or
m
with
the
detai
ls
of
the
re
searc
h
to
be
sign
e
d
in
or
de
r
to
co
nf
irm
that
he
or
s
he
agr
ees
to
t
he
procedu
re
of
t
he
researc
h.
A
bri
ef
ex
planati
on
was
giv
e
n
on
how
t
he
ex
pe
rim
ent
is go
i
ng
t
o
be
c
onduct
ed
a
nd they we
re in
f
or
m
ed
that t
he
e
xp
e
rim
ent is g
oing to
involve
the
t
ra
ckin
g
of
their
ey
e
m
ov
e
m
ent
w
hile
they
a
r
e
doin
g
ta
s
ks
in
t
he
par
ti
cula
r
m
ob
il
e
ap
plica
ti
on
.
First,
the
cal
i
br
at
io
n
was
s
et
fo
r
t
he
part
ic
ipants.
D
ur
i
ng
t
his
proce
ss,
so
m
e
diff
i
culti
es
wer
e
f
ound,
wh
e
re
s
om
e
par
ti
ci
pan
t’s
ey
e
cannot
be
tr
acked
due
to
var
i
ou
s
reas
ons.
This
is
bec
ause
the
ey
e
t
rack
i
ng
te
chnolo
gy
has
been
de
velo
pe
d
to
work
we
ll
with
people
who
hav
e
heal
thy
ey
es
and
norm
al
visu
al
acuit
y.
In
ad
diti
on,
t
he
pa
rtic
ipants
who
w
or
e
glas
ses
we
re
m
ade
sure
t
heir
le
ns
are
in
go
od
conditi
ons
with
le
ss
scratches
as
it
can
ca
us
e
diff
i
culti
es
in
trac
ki
ng
t
heir
ey
e.
Af
te
r
the
cal
ib
rati
on
set
up,
s
om
e
ta
sk
s
were
give
n
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
Evalu
atin
g gender
-
base
d mob
il
e sh
op
ping
applicati
on
us
i
ng eye
-
tracki
ng
te
chnolo
gy
(
As
li
na
B
aharu
m)
1097
to
the
par
ti
ci
pa
nts
to
be
done
on
the
ad
opte
d
Lazada
shopp
ing
m
ob
il
e
app
based
on
fem
al
e
and
m
a
le
m
ental
m
od
el
patte
rn
s
and
t
heir
ey
e
m
ov
em
ent
wer
e
tracke
d
a
nd
rec
orde
d.
Aft
er,
they
ha
ve
finish
e
d
thei
r
ta
sk
s,
the
recordi
ng
was
stoppe
d
and
the
pa
rt
ic
ipants
wer
e
giv
en
s
om
e
so
uve
nirs
as
apprecia
ti
on
f
or
their
involvem
ent in
the eye
-
tr
acki
ng expe
rim
ent.
3.
RESU
LT
S
A
ND
FI
ND
I
NGS
The
ey
e
-
t
rack
i
ng
ex
pe
rim
ent
rec
orded
ev
er
y
ey
e
m
ov
em
e
nt
of
t
he
par
ti
c
ipants
w
hile
th
ey
perform
the
ta
sk
s
giv
e
n
in
both
fem
al
e
and
m
al
e
m
ental
m
od
el
s
base
d
us
er
i
nterf
ace
desi
gn.
The
ta
s
ks
giv
e
n
w
er
e
lim
it
ed
to o
nly
f
eat
ur
e
s that ha
ve
di
ff
e
ren
t l
oc
at
ion
in
the
use
r
inter
face d
es
ign
base
d
on id
entifi
ed
fem
al
e an
d
ma
le
m
ental
m
od
el
patte
r
ns
i
n
[13].
T
he
re
corde
d
data
w
ere
a
naly
zed
i
n
the
f
or
m
of
gaze
plo
ts.
Ga
ze
plo
ts
visu
al
iz
e
a
part
ic
ipant’s
gaze
patte
r
n
th
rou
gh
a
se
ries
of
dots
in
dicat
in
g
fi
xatio
n
an
d
fine
li
ne
s
in
di
cat
ing
sacca
des.
T
he
siz
e
of
the
dots
re
pr
ese
nts
th
e
durati
on
of
a
fixati
on.
S
hort
fixati
ons
are
ind
ic
at
ed
by
sm
a
ll
do
ts
,
w
hile
la
rg
er
do
ts
in
dic
at
e
a
lon
ge
r
f
ixati
on
.
Gaze
plo
t
re
veals
th
e
order
of
fi
xa
ti
on
s
in
dicat
ed
by
nu
m
ber
s
i
n
the
do
ts
w
hich
e
na
bles
resea
rc
he
rs
to
a
naly
ze
how
a
pa
rtic
ipa
nt
pe
rceive
d
a
n
inter
face
a
nd
wh
ic
h
el
e
m
ent
wer
e
l
ooke
d
at
first
a
nd
w
hich
el
em
ents
w
ere
gaze
d
at
la
st.
T
he
f
ixati
on
point
w
hile
they
ga
zi
ng
the
par
ti
cula
r
fe
at
ur
es
w
he
n
doing
the
ta
sk
s
on
differe
nt
us
e
r
inte
rf
ace
desi
gn
we
re
a
naly
zed
base
d
on
t
he
gaze
data obta
ine
d
i
n
the
eye
-
trac
ki
ng
e
xperim
ent.
The
Table
1
con
ta
in
s
a
naly
zed
gaze
plo
t
data
t
hat
wa
s
obta
ine
d
w
hi
le
the
pa
rtic
ipants
we
re
perform
ing
ta
sk
s
in
fem
al
e
a
nd
m
al
e
based
us
er
inte
rf
ace
desig
n.
T
her
e
wer
e
fi
ve
in
va
li
d
data
that
were
fou
nd
duri
ng
the
a
naly
sis
of
gaze
r
eplay
s.
F
ive
pa
rtic
ipant
s’
ey
e
m
ov
em
e
nt
we
re
fail
ed
t
o
be
detect
ed
by
the
ey
e
trackin
g
de
vice.
T
hose
fi
ve
pa
rtic
ipant
data
we
re
c
od
e
d
s
‘N
T
’
in
t
he
Table
1.
T
here
m
a
y
sever
al
r
easo
ns
for
wh
e
re
t
hat
the
par
ti
ci
pa
nt
’s
ey
e
m
ov
e
m
ents
can
no
t
be
trac
ke
d
at
al
l
or
t
hat
a
pa
rtic
ipant
c
ha
nges
her
seat
ing
posit
io
n
in
a
way s
o
t
hat h
e
r
ey
es a
r
e no lo
nger
w
it
hin
t
he
ey
e trac
king
box.
Table
1
.
Gaze
Plot Data
Participan
t
Gen
d
er
Internal lin
k
Search
Log
in
Vo
ice
Prod
u
ct Categ
o
ry
Sh
o
p
p
in
g
car
t
W
ish
list
F
M
F
M
F
M
F
M
F
M
F
M
F
M
P0
1
Fe
m
ale
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
P0
2
Male
NT
NT
4
NT
NT
NT
6
4
NT
5
NT
NT
4
NT
P0
3
Male
NT
NT
NT
9
NT
4
NT
4
7
7
5
7
4
7
P0
4
Fe
m
ale
NT
NT
15
NT
NT
NT
NT
NT
NT
8
NT
NT
NT
NT
P0
5
Male
NT
NT
3
5
2
5
5
6
9
8
6
5
5
4
P0
6
Fe
m
ale
NT
NT
7
6
6
5
NT
9
9
4
6
4
7
18
P0
7
Fe
m
ale
NT
NT
4
NT
5
NT
7
3
4
9
6
2
5
8
P0
8
Male
NT
NT
NT
5
NT
2
NT
8
8
4
4
6
6
13
P0
9
Male
NT
NT
3
4
3
5
5
2
4
3
8
NT
7
NT
P1
0
Fe
m
ale
NT
NT
6
2
2
NT
NT
4
NT
8
7
4
NT
11
P1
1
Fe
m
ale
NT
NT
7
2
3
NT
NT
5
6
4
5
NT
4
6
P1
2
Fe
m
ale
8
3
4
7
5
8
3
7
6
4
4
4
7
8
P1
3
Fe
m
ale
NT
3
NT
NT
NT
5
NT
5
NT
6
NT
5
NT
6
P1
4
Fe
m
ale
4
NT
2
NT
3
NT
5
NT
4
NT
6
NT
7
NT
P1
5
Male
7
3
12
8
6
3
3
4
5
6
5
7
4
8
P1
6
Male
3
5
8
5
4
3
8
6
4
6
5
10
4
4
P1
7
Male
9
7
7
3
8
5
7
15
3
6
5
6
4
7
P1
8
Male
5
3
4
4
6
6
6
3
4
6
4
5
6
4
P1
9
Fe
m
ale
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
P2
0
Male
5
NT
4
3
4
3
2
3
4
9
8
6
7
4
P2
1
Fe
m
ale
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
P2
2
Male
3
6
9
6
4
7
11
4
6
9
6
4
3
5
P2
3
Fe
m
ale
2
NT
6
NT
2
4
3
NT
4
2
NT
NT
NT
NT
P2
4
Male
4
4
4
5
2
3
NT
NT
3
3
10
2
6
5
P2
5
Male
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
P2
6
Male
NT
6
4
7
NT
3
NT
6
6
4
NT
9
6
7
P2
7
Fe
m
ale
2
NT
7
8
5
4
6
8
11
6
NT
2
4
9
P2
8
Male
5
3
7
5
9
7
7
6
9
7
10
4
3
5
P2
9
Fe
m
ale
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
P3
0
Fe
m
ale
NT
NT
3
4
2
NT
2
3
6
9
2
6
NT
5
P3
1
Male
NT
NT
6
3
NT
NT
4
3
3
3
9
2
6
NT
NT
=
Not T
rac
ked
;
F =
Fem
ale
I
nterf
ace;
M
= Ma
le
I
nte
rf
a
ce
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.
13
, N
o.
3
,
Ma
rc
h 201
9
:
109
5
–
110
1
1098
Lege
nd
:
Correct
Con
tra
st
Eq
ual
Sam
e Fixati
on
Fewe
r
Fi
xation
Seve
n
com
plete
data
wer
e
obta
ined,
wh
e
re
the
ey
e
m
ov
em
ent
of
t
ho
se
pa
rtic
ipants
was
su
ccess
fu
ll
y
tracke
d
com
plete
ly
fo
r
al
l
t
he
ta
sk
s
gi
ven.
The
com
plete
data
wer
e
f
r
om
par
ti
ci
pan
t
12
,
par
ti
ci
pa
nt
15,
par
ti
ci
pa
nt
16,
par
ti
ci
pan
t
17,
par
ti
ci
pa
nt
1
8,
pa
rtic
ipant
22
a
nd
par
ti
ci
pan
t
28.
The
r
e
m
ai
nin
g
data
wer
e
par
ti
al
ly
tracke
d
f
or
only
cert
ai
n
ta
sk
s
.
T
he
abbre
viati
on
“
NT”
in
dicat
es
the
fail
ure
in
t
r
ackin
g
the
pa
rtic
ular
featur
e
.
The
num
ber
s
in
the
Table
1
sho
ws
the
fixati
on
point
the
pa
rtic
ipants
re
q
uire
d
in
order
to
l
oc
at
e
the
par
ti
cula
r
feat
ur
e
i
n
that
s
pe
ci
fic
us
er
i
nter
face
desi
gn.
T
he
fi
xation
points
we
re
c
om
par
ed
for
the
fe
at
ur
e
in
bo
t
h
interf
aces
in
order
to
ev
al
uate
the
iden
ti
fied
fem
al
e
a
nd
m
al
e
m
ental
m
od
el
patte
rn
in
[13].
The
s
m
al
le
r
the
f
ixati
on
point
par
ti
ci
pa
nt
require
d
in
ord
er
to
locat
e
at
the
sp
eci
fic
fe
at
ur
e
on
t
he
use
r
inter
face
de
sign,
the
m
or
e
bette
r
is
t
he
l
ocati
on
of
the
feat
ure
on
t
he
i
nterfac
e
as
t
hey
on
ly
require
d
fe
w
fixati
ons
t
o
locat
e
the f
eat
ur
e
.
An
al
ysi
s
of
the
gaze
pl
at
sh
ows
that
the
m
ajo
rity
of
the
m
a
le
par
ti
ci
pan
ts
cou
l
d
fin
d
the
i
nter
nal
li
nk
on
the
m
al
e
mental
m
od
el
ba
sed
us
e
r
inter
f
ace
desig
n
as
they
on
ly
re
qu
i
red
fe
w
fi
xatio
n
in
orde
r
to
fi
nd
a
nd
gaze
at
the
int
ern
al
li
nk.
4
m
al
e
par
ti
ci
pan
ts
out
of
6
m
al
e
par
ti
ci
pan
ts
wit
h
s
ucce
ssfu
l
gaze
dat
a
for
t
he
internal
li
nk
w
ere
easi
ly
fou
nd
t
he
inte
rn
al
li
nk
with
the
sh
ort
est
fi
xatio
n
point
.
T
her
e
fore,
it
justi
fie
s
the
locat
ion
of
i
ntern
al
li
nk
in
m
al
e
m
ental
m
od
el
patte
rn
.
W
he
reas,
fem
al
e
on
ly
one
da
ta
has
bee
n
de
te
ct
e
d,
so
no c
om
par
ison can
b
e
m
ade
, as
s
how
n
i
n
Figure
1
.
Figure
1
.
Fixat
ion
of
par
ti
ci
pa
nt
15
f
or
i
nter
na
l l
ink
on f
em
al
e and m
al
e
UID
As
f
or
the
sea
rch
feat
ur
e
as
sh
ow
n
in
Fig
ure
2,
the
m
ajorit
y
of
the
m
a
l
e
par
ti
ci
pan
ts
pr
e
fer
s
th
e
search
butt
on
as
the
co
rr
ect
po
sit
io
n
as
in
m
al
e
m
ental
m
od
el
patte
rn
as
7
ou
t
of
th
e
12
m
al
e
par
ti
ci
pan
ts
easi
ly
locat
ed
t
he
searc
h
with
few
fi
xation
points.
A
s
for
fe
m
al
e
par
ti
ci
pan
ts,
they
pr
e
fer
the
search
but
ton
to
be
on
to
p
as
3
ou
t
of
the
6
pa
rtic
ipants
easi
ly
detect
ed
the
search
lo
cat
ion
with
few
f
ixat
ion
points
in
th
e
m
al
e
m
ental
m
od
el
base
d
interface
and
the
rem
ain
in
g
3
pa
rtic
ipants
wer
e
able
to
find
the
se
arch
butt
on
wi
th
few
fixati
on
p
oi
nts
in the fem
al
e
m
ental
m
od
el
base
d user i
nte
rf
ace.
Figure
2
.
Fixat
ion
of
par
ti
ci
pa
nt
22
f
or
searc
h on fem
al
e and
m
al
e
UID
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
Evalu
atin
g gender
-
base
d mob
il
e sh
op
ping
applicati
on
us
i
ng eye
-
tracki
ng
te
chnolo
gy
(
As
li
na
B
aharu
m)
1099
Ba
sed
on
Fig
ure
3,
t
he
m
ajo
rity
of
the
m
ale
par
ti
ci
pa
nts’
ey
e
m
ov
em
en
t
towards
the
l
og
i
n
butt
on
wer
e
bette
r
i
n
a
m
a
le
m
ental
m
od
el
based
use
r
inte
rf
ace
de
sign.
T
his
wa
s
pro
ve
d
th
r
ough
the
a
naly
ze
d
data
wh
ic
h
show
th
at
the
m
ajo
rity
of
the
m
al
e
par
ti
ci
pan
ts
we
r
e
able
to
locat
e
the
log
in
ea
sil
y
with
the
few
er
fixati
ons
in
m
al
e
us
er
int
erf
ace
desi
gn
com
par
ed
to
the
fem
al
e
m
ental
m
od
el
base
d
us
e
r
i
nterf
ace
.
This
justi
fies
t
he
locat
io
n
of
the
log
i
n
in
th
e
m
al
e
m
ental
m
od
el
patte
rn
wh
ic
h
is
on
th
e
top
le
ft.
As
f
o
r
th
e
fem
al
e
par
ti
ci
pan
ts,
2
pa
rtic
ipants
re
quire
d
f
ewer
fix
at
ion
s
in
or
der
to
fin
d
the
lo
gin
in
f
e
m
al
e
m
ental
m
od
el
patte
rn
ba
sed
use
r
i
nterf
ace
,
wh
e
reas,
the
ot
her
tw
o
pr
e
fers
the
locat
ion
of
log
in
in
m
a
le
m
ental
m
od
el
base
d
us
er
inter
face
desig
n.
Figure
3
.
Fixat
ion
of
par
ti
ci
pa
nt
15
f
or
l
og
i
n on fem
al
e and
m
al
e
UID
The
m
ajo
rity
of
the
m
al
e
p
arti
ci
pan
ts
onl
y
need
ed
fewer
fixati
ons
to
find
the
vo
ic
e
recogn
it
io
n
bu
tt
on
on
m
ale
us
e
r
i
nterf
a
ce
desi
gn
co
m
par
ed
to
t
he
fem
al
e
us
er
interface
desi
gn.
6
o
ut
of
10
m
al
e
par
ti
ci
pa
nts
f
ound
the
voic
e
recog
niti
on
on
m
al
e
m
ent
al
m
od
el
base
d
inter
face
wi
th
fewer
fi
xat
ion
s
.
Fo
r
e
xam
ple
in
Fig
ur
e
4,
par
ti
ci
pan
t
22
who
is
m
a
le
able
to
fin
d
voic
e
rec
ogniti
on
on
the
m
a
le
us
er
inte
rf
ace
at
the
4
th
fixati
on
p
oi
nt,
w
her
eas
in
fem
a
le
us
er
interface d
e
sign
he
on
ly
w
as
able
to
find
the
bu
tt
on
at
the
11
th
fixati
on
point.
This
re
veals
th
at
the
locat
ion
of
the
voic
e
re
cogniti
on
butt
on
in
the
m
al
e
m
ental
m
od
el
patte
rn
is
in
the
c
orrec
t
po
sit
io
n.
T
he
m
ajo
rity
of
the
fem
al
e
par
ti
cip
ants
al
s
o
on
ly
require
d
fe
we
r
fixati
ons
in
fi
nd
i
ng
the
voic
e
rec
ogniti
on
in
use
r
inter
face
t
ha
t
was
desig
ne
d
base
d
on
the
fem
al
e
m
e
ntal
m
od
el
pa
tt
ern
.
This
rev
eal
s t
ha
t fem
al
e p
ref
e
rs vo
ic
e
rec
ogni
ti
on
on t
op r
i
ght.
Figure
4
.
Fixat
ion
of
par
ti
ci
pa
nt
22
f
or
voic
e
recog
niti
on
on
fem
al
e and
m
a
le
UID
More
ov
e
r,
th
e
analy
zed
data
i
nd
ic
at
e
that
th
e
m
ajo
rity
of
t
he
m
al
e
par
ti
cip
ants
wer
e
a
bl
e
to
fin
d
an
d
do
ta
s
ks
quic
kl
y
relat
ed
to
the
product
cat
eg
or
y
in
fem
al
e
us
er
i
nterf
ace
com
par
ed
to
th
e
m
a
le
us
er
interfac
e
with
fe
wer
fi
xa
ti
on
s
,
as
s
how
n
in
Fig
ur
e
5.
This
ind
ic
at
es
that
m
al
e
pr
efe
rs
the
pro
duct
cat
egory
to
be
on
to
p
wh
ic
h
is
horizon
ta
l.
Wh
e
reas
,
on
the
ot
her
hand,
fem
al
e
pr
efe
rs
the
pro
duct
cat
egory
to
be
on
the
le
ft
as
in
m
al
e
us
er
inter
fac
e
desi
gn.
T
his
was
s
how
n
from
the
analy
zed
data
w
hic
h
sho
ws
that
th
e
fem
al
e
wer
e
able
to
identify
he
pro
du
ct
cat
e
gory
in
m
al
e
us
er
int
erf
ace
desig
n
with
fe
we
r
fix
at
ion
s
c
om
par
ed
on
t
he
fem
ale
use
r
interface
d
esi
gn.
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.
13
, N
o.
3
,
Ma
rc
h 201
9
:
109
5
–
110
1
1100
Figure
5
.
Fixat
ion
of
par
ti
ci
pa
nt
22
f
or
pro
du
ct
cate
gory on
fem
al
e and
m
a
le
UID
As
f
or
sho
pp
i
ng
cart
as
sh
ow
n
in
Figure
6,
the
m
ajo
rity
of
the
m
al
e
wer
e
able
to
find
th
e
featur
e
in
bo
t
h
fem
al
e
and
m
al
e
us
er
int
erf
ace desi
gn
a
t
few
e
r
fi
xatio
ns
as
t
he
s
hopp
ing
ca
rt
locat
io
n
in
both f
em
ale
an
d
m
al
e
us
er
interface
desig
n
is
in
the
top
rig
ht
of
the
inter
face
.
W
he
reas,
the
m
ajo
rity
of
the
fem
a
le
par
ti
cip
ant
s
wer
e
pr
e
ferrin
g
the
locat
io
n
of
the
sho
pp
i
ng
ca
rt
on
the
m
al
e
us
er
interface
desig
n
wh
ic
h
is
nea
r
to
the
internal
li
nk.
Most
of
t
hem
wer
e
su
cc
essf
ul
in
fi
nd
i
ng
the
sho
pp
i
ng
cart
with
few
e
r
fix
at
ion
s
i
n
th
e
m
al
e
us
e
r
interface
.
The
m
ajo
rity
of
t
he
m
al
e
p
arti
ci
pan
ts
ide
ntifie
d
the
wis
hlist
with
fe
w
er
fixati
ons
on
fem
al
e
us
er
interface
desi
gn
as
it
was
loc
at
ed
on
the
m
ai
n
use
r
i
nte
r
fac
e
,
as
s
how
n
i
n
Figure
6.
O
n
t
he
m
al
e
us
er
i
nt
erf
ace
desig
n,
t
he
wis
h
li
st i
s locate
d i
n
inter
nal li
nk which
ha
ve
c
ause
d
the inc
re
ase in f
i
xation po
i
nt in order
t
o
fi
nd
the
wish
li
st.
T
he
m
ajo
rity
of
the
fem
al
e
par
ti
ci
pan
ts
we
re
a
ble
to
ide
ntify
the
wish
li
st
on
the
fem
a
le
m
ental
m
od
el
based
use
r
inter
face
w
it
h
few
e
r
fixat
ion
s
c
om
par
ed
to
the
m
al
e
m
ental
m
od
el
base
d
use
r
i
nterf
ace
.
These a
naly
ze
d data i
ndic
at
e
that t
he wish
li
st sh
ou
l
d be
on the m
ai
n
us
e
r i
nterf
ace
instea
d
in
inter
nal li
nk.
Figure
6
.
Fixat
ion
of
par
ti
ci
pa
nt
22
f
or
wis
hlist
on f
em
al
e an
d
m
al
e
UID
The
ey
e
trac
ki
ng
ex
per
im
ent
j
ust
ifie
s
the
id
entifi
ed
m
al
e
m
ental
m
od
el
patte
rn
[
13
]
as
the
m
ajo
rity
of
t
he
m
al
e
p
arti
ci
pan
ts’
ga
ze
plo
t
re
veals
that
m
al
e
pr
efers
t
he
m
al
e
m
ental
m
od
el
base
d
us
e
r
int
erf
ace
.
This
is
beca
use
the
m
al
e
par
t
ic
ipants
wer
e
a
ble
t
o
fi
nd
the
m
os
t
of
the
fe
at
ur
es
exce
pt
pro
duct
cat
eg
ory
an
d
wish
li
st
in
m
al
e
us
er
interf
ace
design
with
few
e
r
fixati
on
s
c
om
par
ed
to
the
fem
a
le
us
er
inter
face
desig
n.
Wh
e
reas,
fem
al
e
par
ti
ci
pa
nts
wer
e
balanc
ed
in
both
m
al
e
and
fem
al
e
us
er
inter
faces
a
s
m
os
t
of
the
m
cou
ld
fin
d
the
m
os
t
of
t
he
featu
res
in
bo
t
h
inter
f
aces
with
fe
w
er
fi
xations
e
xc
ept
f
or
the
product
cat
eg
ory
an
d
sh
op
ping
cart.
They
pr
e
fer
th
e
locat
ion
of
pro
duct
cat
ego
r
y
and
shoppin
g
cart
as
in
the
m
al
e
us
er
interfac
e
desig
n wit
h
t
he
su
cces
s in
fin
di
ng
with
fe
wer
fixati
ons.
4.
CONCL
US
I
O
N
As
a
co
nclu
sio
n,
the
Ta
ble
2
sh
ows
t
he
nu
m
ber
of
pa
rtic
ipants
who
were
able
to
fin
d
the
featu
res
with
the
fe
wer
fixati
ons
i
n
th
e
pa
rtic
ular
use
r
inter
face.
T
he
higher
num
ber
f
or
t
he
part
ic
ular
feat
ur
es
in
t
he
ta
ble
j
us
ti
fies
t
he
locat
io
n
of
the
featu
res
on
a
us
er
inter
fa
ce
desig
n.
T
his
j
us
ti
ficat
io
n
identifie
s
a
sta
nd
a
r
d
gu
i
deline
th
at
can
be
util
iz
ed
by
dev
el
op
e
rs
in
the
us
e
r
i
nt
erf
ace
desig
n
process
of
m
ob
il
e
sho
pp
i
ng
apps.
Most
of
the
fea
tures
exce
pt
wi
sh
li
st
wer
e
found
with
fe
we
r
fixati
on
s o
n
th
e
m
al
e
us
er
interface
de
sig
n
by
the
par
ti
ci
pa
nts.
T
his
co
nf
irm
s
the
locat
io
n
of
the
par
ti
cula
r
featur
es
base
d
on
m
al
e
mental
m
od
el
pa
tt
ern
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
Evalu
atin
g gender
-
base
d mob
il
e sh
op
ping
applicati
on
us
i
ng eye
-
tracki
ng
te
chnolo
gy
(
As
li
na
B
aharu
m)
1101
gu
i
deline.
Wis
h
li
st
was
f
ou
nd
ea
sil
y
with
few
e
r
fi
xatio
ns
on
fem
al
e
m
ental
m
od
el
patte
rn
base
d
us
er
interface
de
sig
n.
The
re
f
ore,
t
he
l
ocati
on
of
the
wish
li
st
ha
s
bee
n
co
nf
i
r
m
ed
base
d
on
fem
al
e
m
ental
m
od
el
patte
rn
guideli
ne.
Ba
se
d
on
t
he
a
naly
zed
da
ta
,
a
sta
nda
rd
m
ental
m
od
el
patte
rn
w
hich
is
acce
pta
ble
f
or
bo
t
h
fem
al
e and
m
a
le
w
as i
den
ti
fie
d for s
hoppin
g m
ob
il
e app
us
e
r
inte
r
face
desi
gn.
Table
2.
A
naly
zed
Data Ba
se
d on P
hysic
al
Featur
es
Ph
y
sical Fe
atu
res
Fe
m
ale
Use
r
Inte
rf
ace
Male
User
Interf
ac
e
Internal lin
k
3
6
Search
8
11
Log
in
7
8
Vo
ice Recog
n
itio
n
7
8
Prod
u
ct Categ
o
ry
11
13
Sh
o
p
p
in
g
Cart
8
10
W
ish
list
13
5
The
c
ontrib
ution
s
m
ade
in
t
his
pap
e
r
help
to
pro
vid
e
a
bette
r
unde
rstand
i
ng
of
t
he
e
valuati
on
of
m
ob
il
e
web
app
inte
rf
aces
.
In
pa
rtic
ular
,
the
resea
rch
c
onside
rab
ly
ex
pa
nds
the
existi
ng
unde
rstan
din
g
of
how
use
rs
inte
ract
with
the
s
hoppin
g
m
ob
il
e
web
a
pp
a
nd
the
i
m
pact
of
Ey
e
-
trackin
g
data
in
re
-
desi
gn
the
sh
op
ping
m
ob
il
e
web
ap
p
interface
.
T
hes
e
con
tri
bu
ti
ons
can
hel
p
to
i
m
pr
ove
the
de
velo
pm
ent
of
m
or
e
eff
ect
ive
c
omm
ercial
sh
oppi
ng
a
pp
inter
fa
ces.
Fu
t
ur
e
w
orks
c
ou
l
d
inc
lud
e
diff
e
re
nt
Ey
e
-
tracki
ng
a
naly
sis
te
chn
iq
ue
incl
ud
i
ng
Sca
npat
h
Tre
nd
i
ng
A
naly
sis
(S
TA
)
and
A
O
I
m
a
y
us
e
to
analy
se
the
us
er’s
ey
e
m
ov
e
m
ent
data.
In
t
his
resea
rch,
the
Ey
e
-
tr
ackin
g
analy
sis
te
chn
iq
ue
in
cl
ud
in
g
Sca
np
at
h
+
Cued
R
TA
an
d
Heatm
ap
hav
e
been u
sed
to
e
valuate t
he use
r’
s
ey
e m
ov
em
ent
data.
REFERE
NCE
S
[1]
Ta
rc
a
A
.
L
.
,
e
t
a
l.
,
“
A nove
l
sign
al
ing
pat
hwa
y
i
m
pac
t
an
aly
sis (
SP
IA)
,”
Bi
oin
formatic
s
,
vo
l.
25
,
pp.
75
–
82
,
2009
.
[2]
B.
As
li
na,
et
al
.
,
“
Eva
lua
t
ion
of
l
oca
l
iz
a
ti
on
for
e
-
le
arn
ing
w
ebsit
e :
a
pre
l
iminar
y
stud
y
,”
Proc
eedings
of
the
6t
h
Inte
rnational
Co
nfe
renc
e
on
Co
mputing
and
In
f
orm
ati
cs
,
pp
.
54
1
–
546
,
2017
.
[3]
T.
Punith
a,
et
al
.
,
“
Us
ers’
Mental
Model
Pat
te
rn
for
Us
er
Inte
rfa
c
e
Design
of
Mobile
Shopping
A
pps
,”
Adv
an
ced
Sci
en
ce Let
t
ers,
vol
/i
ss
ue:
24
(
2
),
2018
.
[4]
N.
Li
and
P.
Zh
ang,
“
Consum
er
onli
ne
shopping
at
titudes
and
be
ha
vior:
An
assess
m
ent
of
rese
arch
,”
Informatio
n
Syste
ms
proce
ed
ings o
f Ei
ghth
A
meric
as Confe
re
nce
,
2002
.
[5]
K.
L
.
L
au
and
B
.
As
li
na,
“
E
y
e
-
Tr
ac
king
Anal
y
sis for
Us
er
Int
erf
a
c
e
Design
o
f
Sho
pping
Mobil
e
W
eb
Appli
ca
t
ion,
”
Inte
rnat
ional J
o
urnal
of Adv
an
c
ed
Eng
ine
ering
and
Manage
men
t,
v
ol
/i
ss
ue:
2
(
12
)
,
pp
.
287
-
301
,
2
017
.
[6]
R.
Vala,
et
a
l.
,
“
Design
of
a
software
tool
for
m
obil
e
app
lication
user
m
ent
al
m
ode
ls
col
l
ection
a
nd
visual
izati
on
,”
Appl
ie
d
Math
emati
cs,
Computational Scie
nc
e
an
d
Engi
n
ee
ring
,
p
p.
133
-
141
,
201
4
.
[7]
B.
As
li
na,
e
t
al.
,
“
Loc
al
isation
for
E
-
le
arn
ing
W
ebsit
e
of
Com
pre
hensive
Univer
siti
es
in
Malay
si
a
,”
P
ER
TANIK
A
Journal
of
Scien
ce
and
Techno
lo
gy,
vo
l
/i
ss
ue:
25
(S):
91
-
100
,
201
7
.
[8]
Y
.
Habuc
hi
an
d
H
.
Ta
keu
chi
,
“
The
cha
ra
cter
isti
cs
of
usabili
t
y
and
users’
e
y
e
m
ovements
in
sea
rch
ing
for
informati
on
in
a
hie
r
arc
h
ically
orga
nized
info
rm
at
ion
struct
u
r
e
,”
Proceedi
ngs
of
th
e
Annua
l
Me
eting
of
th
e
Cognit
ive
S
ci
en
c
e
Soc
ie
t
y
,
v
ol
/i
ss
ue:
34
(
34
)
,
pp
.
1
632
-
1637
,
2012
.
[9]
Li
eng
L
.
K
.
an
d
Baha
rum
A.
,
“
A
Preli
m
ina
ry
Stud
y
of
Us
er
Inte
rf
ace
De
sign
for
Shopping
Mobile
W
e
b
Applic
a
ti
on
Ge
nder
-
base
d
usin
g
E
y
e
-
Tr
ac
k
ing
Te
chno
log
y
,”
Annal
s
of
adva
nce
d
sci
ences
,
vol/
issue:
2(1)
,
pp.
22
-
33
,
2018
.
[10]
B.
A
slina
and
J.
Aziz
ah
,
“
Us
er
In
te
rfa
ce
Design:
A
Stud
y
of
Exp
e
ct
a
ti
on
-
Confirm
at
ion
The
or
y
,”
Proce
ed
ings
of
th
e
5
th
Inte
rna
ti
ona
l
Confer
ence
on
Com
puti
ng
and
Inform
at
ic
s
(ICOCI
2015).
11
-
13
Augus
t
2015,
Istanbul
,
Turk
e
y.
Univer
siti
Utar
a M
al
a
y
s
ia.
Ind
ex
ed
in
ISI Web
of
Scie
n
ce
,
20
15
.
[11]
Z.
Li
n
a,
et
al
.
,
“
Online
shopping
ac
c
ept
an
ce
m
odel
-
A
cri
t
ical
surve
y
of
cons
um
er
fac
tors
in
onli
ne
shopping
,”
Journal
of
Elec
t
ronic
Comm
erc
e
Re
search
,
vo
l/
is
sue:
8(1),
pp
.
41
-
62
,
2007
.
[12]
C.
Dennis,
e
t
a
l.
,
“
New
insight
s
int
o
th
e
impact
of
digital
sig
na
ge
as
a
retail
at
m
ospheric
to
ol
,”
Journal
o
f
Consum
er
Be
ha
vi
our
,
vo
l/
issue:
11(6),
pp
.
454
-
4
66
,
2012
.
[13]
B.
As
li
na,
e
t
al.
,
“
Eva
lua
ti
ng
Ge
nder
s’
Menta
l
Model
Patterns
for
Us
er
Inte
rfa
ce
Design:
Mobile
Shopping
Apps
,”
Me
dwell
Journal
s (
Journal
of
En
gine
ering
and
A
ppli
ed
Scienc
es)
Spec
ia
l
Iss
ue
(
1816949X,
18187
803)
,
2012
.
Evaluation Warning : The document was created with Spire.PDF for Python.