Int
ern
at
i
onal
Journ
al of Ele
ctrical
an
d
Co
mput
er
En
gin
eeri
ng
(IJ
E
C
E)
Vo
l.
9
, No
.
6
,
Decem
ber
201
9
, p
p.
5512~
5518
IS
S
N:
20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v
9
i
6
.
pp5512
-
55
18
5512
Journ
al h
om
e
page
:
http:
//
ia
es
core
.c
om/
journa
ls
/i
ndex.
ph
p/IJECE
Malaysi
an
s
i
gn
l
anguag
e
m
obile
l
earning
a
pp
lication:
a
recomm
endati
on app to
co
mm
un
icat
e
with
hearing
-
imp
aired co
mm
un
ities
Ha
n
ayanti
Hafit
,
Ch
i
am Wu
i
X
i
ang
,
M
uni
rah M
oh
d
Yu
so
f
,
No
r
fara
dil
la Wahid,
Sh
ah
reen
K
as
si
m
Facul
t
y
of
Sci
en
ce
Com
pute
r
an
d
Inform
at
ion
T
ec
hnolog
y
,
Univ
ersit
i
Tun
Hus
se
in
Onn Mal
a
y
s
ia,
Mal
a
y
s
ia
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
J
a
n
13
, 2
01
9
Re
vised
Ma
y
2
9
, 2
01
9
Accepte
d
J
un
27
, 201
9
Malay
s
ia
n
Sign
La
nguag
e
(MS
L)
is
an
important
la
nguag
e
that
is
used
as
the
primar
y
comm
unic
at
ion
m
e
th
od
for
the deaf
c
om
m
unit
ie
s with
the
o
the
rs
.
Curre
ntly
,
th
e
MS
L
is
poorl
y
known
b
y
th
e
Malay
s
ia
ns
and
the
exi
stin
g
pla
tform
of
le
a
rning
the
sign
la
nguage
is
ine
fficie
nt
,
not
to
m
ent
ion
the
in
complete
func
ti
on
al
i
t
y
of
the
exi
sting
m
obil
e
learni
ng
ap
pli
c
at
ion
in
the
m
ark
e
t.
H
en
ce
,
the
purpose
of
developing
t
his
appl
i
cation
i
s
ai
m
ed
to
inc
re
ase
the
kno
wledge
and
re
co
gnit
ion
of
the
p
ubli
c
towar
ds
th
e
MS
L
and
a
ll
ows
the
m
to
l
ea
rn
th
e
sign
la
n
guage
m
ore
eff
e
ct
iv
ely
.
One
of
t
he
fe
at
ure
s
in
thi
s
appl
i
catio
n
is
sign
de
tection,
which
coul
d
anal
y
ze
the
ima
ge
c
apt
ure
d
b
y
phon
e
c
amera
int
o
sign
m
e
an
ing
.
Th
e
app
li
c
a
ti
on
a
lso
comprised
var
ious
ca
t
egor
ie
s
of
the
sign
for
e
ffi
ci
e
nt
learni
ng
and
quiz
to
te
st
user
knowledge
aga
inst
their
l
earned
sign
la
ngua
ge.
B
eside
s,
the
r
e
is
a
fe
edback
m
odule
fo
r
the
user
to
expr
e
ss
the
ir
opin
ions
and
suggestions
towar
ds
the
app
li
c
at
ion
to
the
develope
r
.
Thi
s
appl
i
ca
t
ion
is
ai
m
ed
to
he
lp
th
e
public
to
le
arn
MS
L
eff
icientl
y
b
y
s
el
e
ct
ing
the
c
ategor
y
of
sign
t
he
y
wish
to
l
earn
and
te
s
t
the
m
selve
s
b
y
using
the
quiz
m
odu
le
.
Beside
s
,
the
application
coul
d
al
so
det
e
ct
the
unk
nown
sign
by
ca
pturi
ng
the
image
and
ana
l
y
z
e
i
t
.
The
applic
at
ion
hel
p
ed
to
r
ai
se
t
he
rec
ogn
it
ion
o
f
MS
L
among
th
e
publi
c
an
d
expose
the
pub
lic
to
th
e
s
ign
la
n
guage
knowledg
e.
It
h
ad
a
lso be
c
om
e
a
sm
al
l
hel
p
in
bre
aki
ng
the
b
arr
ie
r
of
co
m
m
unic
at
ion
be
t
wee
n
th
e
de
af
co
m
m
unit
ie
s
and
th
e
pub
li
c
.
Ke
yw
or
d
s
:
Hear
i
ng
i
m
paired
Mob
il
e
l
ear
ning
Sign
d
et
ect
io
n
Sign
l
an
gu
a
ge
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
:
Han
ay
a
nti bi
nti Haf
it
,
Faculty
of Com
pu
te
r
Scie
nc
e an
d Inform
ation
Tech
nolo
gy
,
Un
i
ver
sit
i T
un
Hu
s
sei
n O
nn
Ma
la
ysi
a
,
Parit R
aja, J
ohor,
Ma
la
ysi
a.
Em
a
il
:
han
a@
uth
m
.ed
u.
m
y
1.
INTROD
U
CTION
Sign
la
ngua
ge
is
a
body
la
nguag
e
us
in
g
hand
gest
ur
es
t
o
c
omm
un
ic
at
e
with
peoples
[
1],
as
opposed
to
s
poke
n
la
ng
uag
e
.
Si
gn
la
ngua
ge
is
oft
en
us
e
d
to
com
m
un
ic
at
e
with
t
he
pe
op
le
suffe
r
ed
from
deaf
ne
ss
or
hear
i
ng
im
paired
,
w
her
e
dea
f
ness
is
a
c
onditi
on
of
losi
ng
th
e
capa
bili
ti
es
to
hear
the
sou
nds.
Sig
n
la
ngua
ge
is
al
so
de
fine
d
as
a
la
nguag
e
th
a
t
us
ed
hands
a
nd
m
ov
em
ents
to
m
ake
a
sign
to
delive
r
a
m
e
anin
g
to
the p
e
op
le
s
in a c
omm
un
ic
at
ion
process
, t
arg
et
in
g
t
o dea
f or
hea
rin
g
-
im
paire
d perso
ns
[2
,
3].
World
Fe
der
at
ion
of
the
De
af
is
est
a
blish
ed
to
ta
ke
car
e
of
the
welf
a
re
a
nd
up
ho
l
d
the
rig
ht
of
the
dea
f
com
m
un
it
ie
s.
T
he
organ
i
zat
io
n
has
m
any
associat
ion
s
l
ocated
in
m
os
t
of
th
e
de
velo
ping
c
ountries
,
includi
ng
Ma
la
ysi
a.
W
he
n
the
Ma
la
ysi
a
Fede
rati
on
of
the
Dea
f
w
as
est
ablishe
d
in
the
ye
ar
1998,
the
fe
de
rati
on
has
s
et
the
sta
nd
a
r
d
c
omm
u
nicat
ion
sig
n
l
angua
ge
for
Ma
la
ysi
a,
wh
i
ch
is
Ma
la
ysi
an
Si
gn
Lan
gu
a
ge
(MS
L)
[4
]
.
This
is
beca
us
e
acc
or
ding
to
a
dif
fe
ren
t
c
ount
ry,
t
he
c
omm
un
ic
at
ion
la
ngua
ges
will
al
so
var
y
base
d
on
the
countr
y’s
sty
le
and
diale
ct
s,
including
sign
la
ng
uage.
Hen
ce
in
M
al
ay
sia
,
MSL
i
s
us
e
d
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Malaysi
an sig
n l
anguage
m
obil
e learn
in
g a
ppli
cation:
a
r
e
comm
e
ndatio
n app t
o
.
..
(
Han
ayanti H
afit
)
5513
as the prim
ary
com
m
un
ic
at
ion
m
et
ho
d
f
or
t
he
dea
f
c
ommun
it
ie
s w
it
h
t
he
o
the
rs.
Accor
ding to
the
sta
ti
sti
cs o
f
the
So
ci
al
W
el
far
e
De
pa
rtm
e
nt
of
Ma
la
ysi
a
in
the
ye
ar
2018,
the
num
ber
of
pe
ople
with
hea
rin
g
i
m
paired
in
Ma
la
ysi
a is esti
m
at
ed
up
t
o 7.6%
from
the tot
al
4
53,
258 pe
r
so
ns
with
disa
bili
ti
es [
5].
Ther
e
are
a l
ot
of
ways to
lear
n
MSL.
T
he
m
os
t com
m
on
and d
i
rect way to
learn
t
he
sig
n l
anguag
e
is
by
at
te
nd
i
ng
th
e
cl
ass
at
the
l
earn
i
ng
c
enter
.
Des
pite
that,
MSL
can
al
s
o
be
le
ar
nt
via
s
urfin
g
the
i
nter
net
f
or
relat
ed
we
bs
it
es,
or
by
p
ur
c
hasin
g
bo
ok
s
.
Mob
il
e
le
arn
i
ng
ap
plica
ti
on
f
or
sig
n
la
ng
ua
ge
has
al
s
o
be
com
e
a
tren
d
in
thes
e
few
ye
ar
s
as
an
al
te
rn
at
ive
way
to
re
place
al
l
the
oth
er
ways.
Mo
bile
le
arn
i
ng
a
ppli
cat
ion
is
so
ft
war
e
that
can
deliver
th
e
le
arn
i
ng
co
nt
ents
to
t
he
pe
op
le
s
us
i
ng
th
e
m
ob
il
e
phon
es.
Mo
bile
le
a
rn
i
ng
app
li
cat
io
n
has
ad
van
ta
ge
on
it
s
high
m
ob
il
it
y
and
flexi
bili
ty
in
the
acce
s
sing
of
t
he
c
onte
nts
w
hich
a
ll
ow
s
le
arn
er
s
to
acc
ess
stud
yi
ng
m
at
erial
s
and
res
ources
a
nytim
e
ever
yw
he
re
[
6
]
.
Mob
il
e
le
arni
ng
a
pp
li
cat
io
n
al
so
has
a
high
su
c
cessf
ul
rate
on
le
arn
in
g
com
par
e
d
to
nativ
e
le
arn
in
g
way
as
the
us
er
ca
n
ch
oose
w
he
n
they
wan
t
to
le
ar
n,
rather
tha
n
f
orci
ng
t
hem
wh
il
e
they
hav
e
no
de
sire
on
le
arn
i
ng
[
7
].
I
n
the
m
ob
il
e
le
a
rn
i
ng
app
li
cat
io
n,
th
e lea
rn
in
g
m
ater
ia
ls can co
m
pr
ise
of
var
i
ous ele
m
ents su
ch
as v
i
deos, im
ages and
s
ound
s r
at
he
r
than
plain
te
xt
to
al
low
m
ore
interact
ive
and
inte
resti
ng
le
arn
in
g
co
nte
nts
delivere
d
t
o
the
le
a
r
ne
rs
[
8
]
.
Ther
e
a
re
a
few
existi
ng
m
o
bile
le
arn
in
g
app
li
cat
io
n
for
MSL
[
9
-
1
1
]
,
ye
t
they
are
s
t
il
l
far
too
le
ss
a
nd
incom
plete
f
or
eff
ect
ive
learni
ng in
t
he
si
gn langua
ge [
12
].
Hen
ce
,
to
so
l
ve
the
prob
le
m
s
face
d,
the
re
is
a
neces
sit
y
to
buil
d
a
ne
w,
i
m
pr
ov
e
d
ver
si
on
of
MSL
m
ob
il
e
le
arn
ing
ap
p.
T
he
a
pp
will
con
ta
in
a sign
d
et
ect
or
m
echan
ism
wh
ic
h
c
an
ca
pture
the
i
m
age
of
t
he
sig
n
and
inter
pr
et
it
to
the
ap
p
us
e
r
s.
Ot
her
tha
n
t
hat,
the
ap
p
wi
ll
con
ta
in
us
e
f
ul
ph
rases o
f
si
gn
an
d
cat
e
gor
iz
ed
to
the
dif
fer
e
nt
si
tuati
on
t
o
im
p
rove
the
na
viga
ti
on
pr
ocess.
Fu
rt
her
m
or
e,
t
her
e
will
be
a
qu
iz
m
od
ule
t
o
te
st
the user
knowl
edg
e
on t
he
si
gn lan
guage
s th
ey
h
ave
lear
nt
wh
il
e
us
in
g
t
he
app.
2.
RESEA
R
CH MET
HO
D
Ma
la
ysi
an
Sign
Lan
guage
M
ob
il
e
Lear
ning
Applic
at
ion
is
the
app
prot
ot
ype
that
al
lows
ever
yo
ne
to
le
arn
the
MS
L
.
The
re
are
four
featu
res
int
egr
at
e
d
into
t
he
ap
p
,
w
hich
sign
detect
ion
are
us
i
ng
t
he
phone
ca
m
era,
le
arn
i
ng
by
cat
eg
ory
,
play
qu
iz
,
and
giv
e
fee
dback
.
I
n
the
app
de
velo
pme
nt
of
the
proj
ect
,
the
evo
l
ution
a
r
y
pr
oto
ty
pi
ng
-
base
d
m
et
ho
do
log
y
is
cho
se
n
as
the
m
od
el
t
o
de
velo
p
the
syst
e
m
.
Evo
luti
on
a
ry
prototy
pi
ng
m
od
el
is
a
syst
em
dev
el
opm
ent
m
od
el
that
allow
s
us
er
or
cl
ie
nts
to
te
st
the
syst
e
m
pr
otot
ype
in
m
ul
ti
ple
sta
ges
to
determ
ine
t
he
syst
e
m
'
s
req
ui
rem
ent
m
or
e
cl
early
and
enab
le
the
de
ve
lop
e
r
to
i
m
pr
ove
an
d
dev
el
op
t
he
de
sired
syst
em
of
the
cl
ie
nt
s.
Using
this
m
et
ho
dolo
gy,
t
he
dev
el
op
e
r
will
first
include
the
m
ini
m
a
l
fu
nctio
nalit
y
of
the
syst
em
to
te
st
the
us
e
r’s
opini
on
a
nd
acce
ptance
an
d
gathe
rin
g
t
he
ne
w
requirem
ent
from
the
us
ers
.
More
im
pr
ov
e
m
ent
and
f
un
ct
ion
al
it
y
will
be
add
e
d
in
the
s
econd
pr
oto
ty
pe
un
ti
l
the
final
proto
ty
pe
is
per
fect
and
rea
dy
to
be
i
m
ple
m
ented
as
the
real
syst
e
m
[1
3].
HTML,
Ty
pescri
pt
an
d
Ionic F
ram
ewo
rk are
us
e
d
t
o dev
el
op the
ap
p
wh
il
e
Ang
ular
Fire
base
is used as
the
data
base
for
t
he
a
pp
.
2.1.
Analysis
and
design
The
data
us
e
d
for
the
a
pp
de
velo
pm
ent
is
colle
ct
ed
f
rom
a
few
s
ourc
es.
First,
data
colle
ct
ion
i
s
carried
out
by
interviewi
ng
on
e
of
the
de
af
pe
op
le
.
Due
to
the
disab
il
ity
of
the
in
te
rv
ie
wee
to
s
pea
k,
the
intervie
w
proces
s
is
do
ne
by
us
in
g
ha
nd
gestu
res
an
d
pa
per
w
riti
ng
m
et
hod.
Th
e
res
ults
of
the
inte
rv
ie
w
are fu
nctional
r
equ
i
rem
ents an
d user
re
qu
i
re
m
ents f
or the
prop
os
ed
ap
p
.
S
econd,
Inter
net is u
se
d
as a
m
edium
to co
ll
ect
the
dat
a relat
e to the
ex
ist
in
g
MSL
.
How
e
ve
r,
t
he reso
urces
relat
ed
to
MSL a
re
scar
ce a
nd li
m
i
te
d.
An
al
ysi
s
ph
a
se
involve
s
the
proces
s
of
re
qu
i
rem
ent
analy
sis
to
determ
ine
the
goal
a
nd
f
unct
ion
al
it
y
of
t
he
propos
ed
a
pp
t
hat
will
be
de
vel
op
e
d.
It
in
volves
the
analy
sis
of
re
qu
i
re
m
ents
gathe
re
d
f
r
o
m
the
cl
ie
nts
in
the
inter
views
and
quest
i
on
naires
a
nd
t
he
resu
lt
is
f
our
m
od
ules
im
ple
m
ented
in
th
e
app
,
wh
ic
h
a
re
a
si
gn
detect
io
n
m
odule,
a
le
ar
nin
g
by
cat
e
gory
m
od
ule,
a
qui
z
m
od
ule,
a
nd
a
fee
db
ac
k
m
od
ule,
as sho
wn in
Figure
1.
Desig
n
phase
invol
ves
t
he
proces
s
of
de
sign
i
ng
the
wirefram
e
and
inter
face
of
the
propose
d
app
li
cat
io
n
.
W
i
refram
e
decides
the
bo
ne
structu
re
of
the
ap
plica
tio
n
on
w
hat
sh
oul
d
be
in
cl
ud
e
d,
the
ar
rangem
e
nt
of
the
el
e
m
ents
and
t
he
functi
onal
it
y
of
a
n
a
pp.
T
he
desig
n
of
a
wirefram
e
can
sa
ve
the
app
li
cat
io
n
dev
el
opm
ent
ti
m
e
and
pro
duce
a
con
sist
e
nt
la
yout
for
the
ap
plica
ti
on
[14].
In
the
propose
d
app
,
the
wi
refr
a
m
e
is
desig
ne
d
a
nd
sho
wed
in
Fig
ure
2.
T
he
str
uctu
re
of
the
propose
d
app
is
de
sig
ne
d
to
be
cl
ean
and
sim
ple
to
enh
a
nce
us
er
-
fr
ie
ndly
and
us
a
bili
ty
f
or
al
l
kinds
of
us
er.
T
he
hea
der
par
t
will
con
ta
in
a
back
butt
on
to
al
low
the
use
r
to
go
bac
k
to
the
pr
e
vious
ta
b
an
d
the
ta
b
ti
tl
e
will
be
sh
own
in
the
m
id
dle
of
the
hea
der
to
disp
la
y
the
lo
cat
ion
of
the
us
er
in
t
he
ap
p
.
T
he
co
nten
ts
of
the
ta
b
will
be
disp
la
ye
d
on
th
e m
idd
le
of the
scree
n,
w
hile t
he
na
vig
at
io
n
ba
r wil
l be loc
at
ed
at
the lower
par
t.
This
wirefram
e lay
o
ut w
il
l
be
a
pp
li
ed
to
m
os
t of
the
pa
ges
i
n
the
ap
p
t
o bu
il
d a c
onsist
ent lay
ou
t
f
or the
us
er
and e
nh
a
nce
us
a
bili
ty
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
6
,
Dece
m
ber
201
9
:
5512
-
5518
5514
Figure
1.
Use
c
ase dia
gr
am
f
or m
al
ay
s
ia
n
sig
n
la
nguag
e
m
ob
il
e lea
rn
in
g
a
pp
Figure
2
.
W
ire
fr
am
e d
esi
gn
2.2.
Implem
ent
ati
on
In
t
his
ph
ase
,
the
im
ple
m
e
ntati
on
w
orks
ha
ve
been
a
pp
li
ed
to
t
he
m
od
ules
as
discusse
d
i
n
the
previ
ous
se
ct
ion
to
t
urn
th
e
m
odules
into
functi
onal
acc
ordin
g
to
syst
em
fu
nctio
nal
re
qu
i
rem
ents
and
use
r
requirem
ents.
The
dev
el
op
e
d
app
co
ntaine
d
four
featu
res
as
discu
ssed
i
n
the
previ
ous
s
ect
ion
,
w
hich
i
s
sign
detect
ion,
le
ar
n
by
cat
eg
ory
,
quiz
,
a
n
d
fee
db
ac
k.
T
he
a
pp
is
im
ple
m
ented
by
us
i
ng
I
on
ic
Fr
am
ework
an
d
Ang
ular
Fire
ba
se.
The
databa
se
that
is
us
ed
in
the
pro
j
ect
is
Firebase
wh
i
ch
is
a
cl
oud
da
ta
base
an
d
re
qu
i
res
no
-
S
QL
c
omm
ands
to
que
ry
the
data.
Be
si
de
s,
the
re
a
re
se
ve
ral
A
PI
w
hich
ha
d
bee
n
u
se
d
to
de
velo
p
t
he
ap
p
.
Cam
era
AP
I
ha
s
been
im
ple
m
ented
in
the
app
to
al
lo
w
the
us
e
r’
s
phon
e
to
captur
e
th
e
i
m
age
in
the
sign
detect
ion
m
odules
in
the
ap
p
.
Cl
oud
Visio
n
AP
I
is
a
Goo
gle
Cl
oud
ser
vi
ce
that
can
detect
the
con
te
nt
s
and
la
bels
in
the
im
age.
In
this
pro
j
ect
,
Cl
oud
Vision
AP
I
ha
s
been
im
ple
m
ented
to
sig
n
detect
io
n
m
odule
to
detect
the
la
bel
s
of
sig
n
la
ngua
ge
c
on
ta
ine
d
i
n
the
im
age.
The
f
ollo
wing
s
ect
ion
s
s
how
the
im
ple
m
enta
ti
on
of
the
ap
p
i
nter
fa
ces.
2.3.
Sign
detec
tion
Sign
detect
io
n
m
od
ule
is
a
m
od
ule
that
al
lows
t
he
us
e
r
to
captu
re
the
i
m
age
and
det
ect
the
sign
m
eaning
.
I
n
this
m
od
ule,
c
a
m
era
plugin
will
be
us
e
d
t
o
ena
ble
the
abili
ty
of
us
e
r
’s
ph
on
e
t
o
i
niti
at
e
the
cam
era
fu
nc
ti
on
a
nd
ta
ke
the
pictu
re.
Be
sides,
cl
oud
vision
API
will
be
us
ed
t
o
detect
the
la
bel
co
ntain
e
d
in
the
im
age
captu
red
with
t
he
dataset
s
on
t
he
inter
net
to
de
te
rm
ine
the
m
os
t
su
it
a
ble
la
be
l
for
the
im
age
an
d
disp
la
y i
t.
Figure
3
s
how
s
the
interface
of
the
m
od
ule
after
the
sig
n
detect
ion
re
su
l
t
has
been
displ
ay
ed
base
d
on
the
im
age
captu
red
by
t
he
us
e
r.
Firstl
y,
us
e
rs
a
re
re
quire
d
to
ta
p
on
the
“t
ake
ph
oto
”
butt
on
to
ta
ke
the
phot
o.
Af
te
r
the
photo
is
t
aken,
it
will
be
au
tom
at
ic
ally
up
l
oad
e
d
to
th
e
fireb
a
se
sto
r
age
w
he
re
the
cl
oud
functi
on
w
hich
is
the
la
bel
de
te
ct
ion
f
ro
m
C
loud
Visi
on
A
PI
will
be
init
ia
te
d
to
sta
rt
de
te
ct
ing
the
la
be
l
that
m
igh
t
possibly
be
co
ntaine
d
i
n
the
im
age.
The
la
bels
c
on
ta
ined
in
t
he
im
a
ge
will
be
a
nal
yz
ed
by
com
par
in
g
the
existi
ng
da
ta
set
on
the
internet
to
obta
in
the
m
os
t
accurate
resu
lt
.
H
oweve
r,
since
th
ere
is
la
ck
of
da
ta
set
of
Ma
la
ysi
an
Sign
La
ngua
ge
co
ntents
on
the
inter
net,
the
si
gn
detect
i
on
m
od
ule
only
can
detect
f
ew
of
the sig
ns
c
urre
nt
ly
, for
e
xam
ple alph
a
bet
A,
B, an
d
E
.
2.4.
Le
arnin
g
b
y
c
at
e
gory
Learn
i
ng
by
cat
egory
m
od
ul
e
is
a
m
od
ule
that
al
lows
us
e
rs
to
le
arn
the
Ma
la
ysi
an
Sign
Lan
guage
accor
ding
to
t
he
cat
eg
or
y
th
at
is
us
efu
l
in
their
daily
li
fe.
In
the
a
pp
,
th
ere
are
sixtee
n
cat
ego
rie
s
c
urren
tl
y
include
d,
with
a
total
of
m
or
e
tha
n
tw
o
hund
red
sig
ns
c
on
te
nt
inclu
de
d
in
the
cat
e
gories.
Fig
ur
e
4
show
s
the
be
ginnin
g
interface
of
the
m
od
ule.
Wh
e
n
the
us
e
rs
ope
ned
t
he
app
,
cat
eg
ory
ta
b
is
disp
l
ay
ed
as
the
fi
rst
ta
b
f
or
the u
se
rs.
Use
rs
ca
n
s
cr
oll
an
d
fin
d
t
heir
des
ired
cat
e
gory
a
nd
ta
p
on
it
to v
ie
w
the sig
ns
li
st
i
n
that ca
te
gory.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Malaysi
an sig
n l
anguage
m
obil
e learn
in
g a
ppli
cation:
a
r
e
comm
e
ndatio
n app t
o
.
..
(
Han
ayanti H
afit
)
5515
Figure
5
s
how
s
the
sign
li
st
interface
afte
r
the
us
er
c
ho
s
e
the
cat
ego
ry.
In
this
case,
th
e
us
er
ha
d
chosen
t
he
“F
eel
ing
s”
cat
eg
or
y.
T
he
sig
n
li
st
pag
e
will
then
disp
la
y
the
e
ntire
si
gn
nam
e
that
rel
at
ed
t
o
the
cat
eg
or
y
a
s
a
li
st
to
the
us
er
t
o
sel
ect
the
sig
n
to
le
a
rn.
Fig
ure
6
s
how
s
t
he
sig
n
pa
ge
inter
fac
e
after
the
us
e
rs
ha
ve
sel
ect
ed
their
desire
d
sig
n
to
le
arn.
T
he
im
age
will
be
r
ead
f
ro
m
the
fireb
ase
databa
s
e
an
d
disp
la
y
to
the
us
er
.
De
pe
nd
i
ng
on
the
sig
n
cho
s
en
,
the
i
m
age
will
be
disp
la
ye
d
i
n
two
ty
pes
,
w
hi
ch
is
a
dynam
ic
i
m
age
in GIF f
or
m
at
, or
a stat
ic
im
a
ge of
JPE
G for
m
at
.
Figure
3
.
Sig
n detec
ti
on p
a
ge
interface
Figure
4
.
Ca
te
gories
pag
e
interface
Figure
5
.
Sig
n l
ist
p
age i
nterfac
e
Figure
6
.
Sig
n pag
e
inter
face
2.5.
Quiz
Qu
iz
m
od
ule
i
s
a
m
od
ule
tha
t
al
lows
us
e
rs
to
te
st
their
know
le
dg
e
on
th
e
sign
la
ng
uage
they
hav
e
le
arn
t
within
t
he
ap
p
.
Wh
e
n
the
us
e
r
sel
ect
s
the
qu
iz
ta
b,
the
qu
iz
im
age
will
be
gen
e
rated
ra
ndom
l
y
fr
om
the
colle
ct
io
ns
of
the
li
st
of
s
ign
s
c
onta
ine
d
in
the
data
bas
e
an
d
disp
la
y
to
the
us
e
r.
Fig
ur
e
7
sho
ws
th
e
quiz
pag
e
inter
face
wh
e
n
t
he
us
er
sel
ect
ed
the
quiz
ta
b
in
t
he
ap
p
.
A
rand
om
i
m
age
will
be
c
ho
s
en
from
the
si
gn
li
st
in
the
data
base
a
nd
dis
pl
ay
ed
to
t
he
us
e
r
to
a
nswer
the
quiz
.
User
s
ca
n
ei
the
r
ch
oose
to
e
nter
thei
r
a
ns
we
r
for
the
quiz
an
d
cl
ic
k
on
the
“check”
butt
on
to
validat
e
the
ir
ans
wer
or
c
hoose
to
sk
i
p
to
the
ne
xt
qu
e
sti
on
i
f
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
6
,
Dece
m
ber
201
9
:
5512
-
5518
5516
they
ha
ve
no
i
dea
on
t
he
a
nswer
of
the
qu
i
z.
The
ap
p
will
prom
pt
a
dialog
box
f
or
eac
h
of
the
opti
on
s
they
sel
ect
ed
to
op
erate
the
ne
xt
m
ov
e
based
on
their
decisi
on,
s
uch
a
s
ask
for
procee
ding
to
nex
t
qu
es
ti
on
or
rev
eal
the
a
nswer
of
that
quest
io
n.
T
he
va
li
dation
m
echan
ism
is
al
so
i
m
ple
m
ented
i
n
the
in
pu
t
ar
ea
of
the quiz a
nswe
r
to
en
s
ure the
us
er
answe
r
th
e questi
on
with
the
rig
ht for
m
at
.
2.6.
Feedb
ack
Feed
back
m
odule
is
an
im
po
r
ta
nt
m
od
ule
th
at
al
lows
the
c
omm
un
ic
at
ion
of
betwee
n
de
velo
per
a
nd
the
us
e
rs.
T
he
m
od
ule
is
use
d
as
the
bri
dge
to
c
onnect
the
de
velo
per
and
t
he
us
ers
of
t
he
ap
p
to
colle
ct
their
op
i
nions
and
sug
gestio
ns,
or
as
well
c
om
plaints
to
be
us
e
d
for
bette
r
im
pr
ov
em
ents
of
t
he
syst
e
m
in
the futu
re.
Figure
8
s
hows
the
fee
db
a
ck
pa
ge
inter
f
ace
wh
e
n
the
us
er
sel
ect
s
th
e
feedback
ta
b
in
the
ap
p
.
The
us
ers
a
re
on
ly
re
quire
d
to
enter
t
heir
e
-
m
ail
add
res
s
and
t
heir
op
i
nions
a
gainst
t
he
ap
p
a
nd
cl
ic
k
on
the
“se
nd
”
bu
t
ton
t
o
directl
y
forw
a
r
d
thei
r
op
i
nions
t
o
t
he
de
velo
per.
Va
li
dation
p
r
oces
s
is
im
ple
m
ent
ed
t
o
ens
ur
e
t
he
em
a
il
entered
by
th
e
us
er
is
valid
to
al
low
t
he
de
velo
per
t
o
c
omm
un
ic
at
e
with
the
us
e
rs
via
e
-
m
ai
l
if
necessa
ry.
The
se
nt
fee
dback
will
be
store
d
on
t
he
firebase
dat
abase
f
or
the
dev
el
oper
to
view
the
incom
ing
f
eedb
ac
k
f
ro
m
the
us
er
s
an
d
help
the
de
vel
op
e
r
to
colle
ct
essenti
al
su
gg
est
ion
s
f
ro
m
disti
nct
per
s
pecti
ve
t
o m
ake a
bette
r
versio
n of an
im
pro
ved
ap
p
i
n
t
he fut
ur
e.
Figure
7
.
Q
uiz
pag
e
inter
face
Figure
8
.
Fee
dback
p
a
ge
interface
3.
RESU
LT
S
A
ND
DI
SCUS
S
ION
The
ap
p
is suc
cessf
ully
b
uilt
b
y t
he
de
velo
pe
r.
U
se
r
te
sti
ng is carr
ie
d
out t
o
eval
uate the ap
p'
s d
esi
gn
and
t
he
ap
p'
s
fu
nctio
ns.
This
us
er
te
sti
ng
in
volve
d
30
use
r
’s
act
as
the
public.
The
us
e
r
te
s
ti
ng
was
c
ondu
ct
ed
at
U
ni
ver
sit
iTu
n
Hu
s
sei
n O
nn Mal
ay
sia
in
P
arit
Raja,
Joh
or, Mal
ay
sia
.
Table
1
sho
ws
the
res
ult
of
the
use
r
acce
pt
ance
te
st
on
th
e
ap
p'
s
desig
n.
The
nav
i
gatio
n
feat
ur
e
i
n
the
ap
p
has
s
cor
e
d
the
high
est
scor
e,
w
hich
is
99%
f
rom
us
ers
rati
ng
,
w
hich
m
eans
the
ap
p
is
easy
to
unde
rstan
d
an
d
us
e
by
the
public.
The
sco
ri
ng
is
f
ollow
e
d
by
the
second
crit
eria,
wh
ic
h
is
the
con
te
nt
la
yout
and
thir
dly,
th
e
app
inter
face
s.
Howe
ver,
te
xt
sty
le
has
ac
hieve
d
the
lowe
st
scor
in
g
w
hi
ch
is
69
%
only
fr
om
us
ers
r
at
in
g. T
he reaso
n be
hin
d t
his
w
as
pro
bab
ly
the
ina
ppr
opriat
e text s
iz
e comm
ented
by m
os
t of the
u
se
rs.
Table
1.
A
pp
li
cat
ion
desig
n
a
ccepta
nce
test
No
.
Sco
ring
Criter
io
n
Sco
re
(%
)
1
Ap
p
licatio
n
I
n
terfaces
89
2
Co
n
ten
t L
ay
o
u
t
97
3
Text St
y
les
69
4
Nav
ig
atio
n
99
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Malaysi
an sig
n l
anguage
m
obil
e learn
in
g a
ppli
cation:
a
r
e
comm
e
ndatio
n app t
o
.
..
(
Han
ayanti H
afit
)
5517
Table
2
sho
ws
the
res
ult
of
use
r
acce
pta
nce
of
t
he
syst
em
f
un
ct
io
ns.
Am
on
g
t
he
f
our
f
unct
ion
s,
s
i
gn
detect
ion
m
odules
achie
ved
t
he
hi
gh
est
sc
ore,
w
hich
is
87
%
of
th
e
us
e
r
r
at
ing
.
Howe
ve
r,
since
t
her
e
i
s
la
ck
of
dataset
of
MSL
co
ntents
in
the
inte
rn
e
t,
the
sig
n
detect
ion
m
od
ule
on
ly
ca
n
det
ect
few
of
t
he
signs
currently
,
f
or
exam
ple
al
ph
a
bet
A
,
B,
a
nd
E.
T
he
sc
or
i
ng
is
f
ollo
wed
by
a
quiz
m
od
ule,
w
hich
ac
hieve
d
a
77
%
sc
or
e
a
nd
the
le
ar
n
by
cat
ego
ry
m
o
du
le
w
hic
h
is
69%.
S
om
e
su
gg
e
sti
on
s
from
us
ers
wh
ic
h
the
le
ar
n
by
cat
egory
m
od
ule
sh
oul
d
inclu
de
som
e
sentences
instea
d
of
sin
gle
phrases
to
increase
the
le
arn
in
g
eff
ect
ive
ness
i
n
sig
n
la
ngua
ge.
Feed
bac
k
m
od
ules
achie
ved
the
l
ow
es
t
scor
e
f
ro
m
the
us
ers
,
wh
i
ch
is
on
ly
65%,
as
the
m
od
ule
is
just
us
ed
f
or
se
nd
i
ng
a
c
omm
ent
with
out
a
ny
oth
e
r
feat
ur
e
s
in
it
.
T
hese
da
ta
ha
d
been ta
ke
n
t
o
r
eview t
he fut
ure w
orks of t
he
app
.
Table
2
.
Syst
em
fu
nctional a
ccepta
nce
test
No
.
Sco
ring
Criter
io
n
Sco
re
(%
)
1
Sig
n
Detec
tio
n
87
2
Lear
n
b
y
Categ
o
r
y
69
3
Qu
iz
77
4
Feed
b
ack
65
4.
CONCL
US
I
O
N
In
ed
ucati
ng
de
af
a
nd
m
ute
people,
ther
e
a
re
th
ree
great
appr
oach
es
tha
t
can
be
su
cce
ssfu
ll
y
us
e
d:
the
bili
ngual
-
bi
cultural
(BiB
i)
ap
proac
h
w
hi
ch
m
akes
us
e
of
AS
L;
the
a
udit
or
y
verbal
appr
oach
t
hat
te
aches
the
En
glish
La
ngua
ge
th
rou
gh
resi
du
al
h
ear
ing
a
nd
s
pe
ec
h
instea
d
of
sig
n
la
nguag
e;
a
nd
total
com
m
un
i
cat
ion
that
com
bin
es
aud
it
ory
an
d
vi
su
al
com
m
un
i
cat
ion
f
or
instr
uction
[15].
In
this
pap
er
,
we
pr
ese
nted
a
m
ob
il
e
app
pr
oto
ty
pe
t
o
le
ar
n
MS
L.
This
a
pp
em
br
aces
a
f
unct
io
n
f
or
t
he
MSL
le
arn
e
r
to
detect
sign
m
eaning
us
i
ng
the
ph
on
e
ca
m
era.
Be
sides
,
to
ac
hieve
th
e
obj
e
ct
ive
of
i
m
pr
ovin
g
t
he
com
m
un
ic
at
ion
proces
s
be
tween
norm
al
peo
ple
with
the
dea
f
com
m
un
it
y,
the
app
has
gat
hered
a
lot
of
MS
L
to
pu
t
into
th
e
le
arn
in
g
cat
egory.
W
it
h
a
total
of
16
cat
eg
or
ie
s
a
nd
o
ve
r
20
0
sig
ns
include
d,
it
is
an
im
m
e
ns
e
help
to
weak
e
n
the
com
m
un
icati
on
ba
rr
ie
r
betwee
n
the
de
af
com
m
un
it
i
es
and
no
rm
al
people.
The
data
colle
ct
ed
fr
om
the
us
e
r
acce
ptance
te
st
ha
d
a
lso
ai
de
d
in
de
ci
din
g
t
he
project
’s
f
utu
r
e
w
orks
a
nd
di
recti
on
s
an
d
we
be
li
eve
that t
his a
pp w
i
ll
co
ns
i
der
a
bly i
ntr
oduce a
n
e
w
f
unct
io
n
c
om
par
ed
to t
he c
urren
t a
ppli
cat
ion
.
ACKN
OWLE
DGE
MENTS
We
w
ou
l
d
li
ke
to
thank
Un
i
ve
rsiti
Tun
H
usse
in
O
nn
Ma
la
ysi
a
fo
r
sup
porting
this
resea
rch
unde
r
the Co
ntract
G
ran
t
V
ot num
ber
W00
4.
REFERE
NCE
S
[1]
Merri
am
-
W
ebste
r
,
“
Sign
La
ng
uage
,
”
2017
,
[
Online
]
,
Availa
ble
:
h
tt
ps://
ww
w.m
err
ia
m
-
webster
.
com/dic
t
ionar
y
/
sign%20la
nguag
e
.
[2]
Medic
in
eNe
t
,
“
Sign
La
nguag
e
,
”
2017
,
[Onl
in
e]
,
Avail
ab
le
:
htt
ps://
ww
w.m
e
dic
in
ene
t
.
com/sc
ript
/main/art
.
asp
?
art
i
cl
ek
e
y
=3915
8
.
[3]
Cambridge
Dictionar
y
,
“
Sign
L
angua
ge
,
”
2017
,
[Online
]
,
Avail
able:
htt
p
:/
/d
ic
t
iona
r
y
.
ca
m
bridg
e.
org/d
ic
t
iona
r
y
/
engl
ish/sign
-
l
an
guage
.
[4]
Blogge
r
,
“
Histor
y
Mal
a
y
s
ia
n
Feder
at
ion
o
f
the
Dea
f
,
”
2006
,
[Online
]
,
Avail
able:
http://
waktusol
at.n
et
/
ogos1992.bl
ogspot.
com/2006/06
/
histor
y
-
m
ala
y
sia
n
-
fed
erati
on
-
of
-
dea
f_18.
h
tml
.
[5]
Depa
rtment
o
f
S
ta
ti
st
ic
s Ma
lay
si
a,
“
Soci
al
St
a
ti
st
ic
s Bullet
in
Mal
a
y
si
a
,
”
Dep
art
m
ent
of
Statis
tics
Malay
s
ia,
2018
.
[6]
M.
Micha
e
l
and
W
.
Donna,
“
Empow
eri
ng
Le
ar
ner
s
W
it
h
Mobile
Open
-
Acc
ess
Le
arn
ing
Initiati
ves
,
”
Ad
vances
in
Mobil
e
and
Dist
ance
Learning
I
GI Global
,
2017
.
[7]
A.
Khali
l
,
“
Mobile
L
ea
rn
ing
Te
chno
logi
es
,
”
Inte
rnational
Jo
urnal
of
El
e
ct
ri
cal
and
Computer
Engi
ne
ering
(
IJE
CE)
,
v
ol. 7
(
5)
,
pp
.
2833
-
283
7,
2017
.
[8]
F.
Gaz
za
we
,
“
Com
par
ison
of
website
s
and
mobi
le
applic
at
io
ns
for
e
-
le
arn
in
g
,
”
Inte
rnat
iona
l
Confe
renc
e
o
n
Technol
ogy
i
n
E
ducat
ion
2017
(
IACB
,
ICE
&
IC
TE)
,
New
York, 2017
.
[9]
Google
Pla
y
Store
,
“
M
y
BI
M
,”
2017
,
[Online
]
,
Avail
ab
le
:
htt
ps:
//
pl
a
y
.
google
.
com/store/
apps/d
et
a
il
s
?
id
=
m
y
.
com.m
obiz.
m
y
b
im
.
[10]
Google
Pla
y
St
ore
,
“
Belaja
r
B
aha
sa
Is
y
ar
at
,”
2017
,
[Online]
,
Avail
able:
ht
tps:/
/pla
y
.
goog
le.c
om
/store
/
apps/
det
a
il
s
?
id
=c
om
.
a
andt
.
b
el
a
ja
rb
ahas
ai
s
y
ar
at
.
[11]
Google
Pla
y
St
ore
,
“
Baha
sa
Is
y
ar
at
Ma
lay
sia
,”
2
017
,
[Onlin
e]
,
Avail
able:
h
tt
ps://
p
lay
.
googl
e.
com/store
/a
pps
/
det
a
il
s
?
id
=c
om
.
s
rda
nanja
y
a
.
bahas
ai
s
y
ar
at
.
[12]
S.
A.
Mokhtar
and
S
.
M
.
S
.
Anuar
,
“
Learni
ng
appl
i
ca
t
ion
for
Malay
sian
sign
la
nguag
e:
content
design
,
use
r
int
erf
ac
e
and
usa
bil
ity
,
”
IMCOM
,
2015
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
6
,
Dece
m
ber
201
9
:
5512
-
5518
5518
[13]
P.
Ta
vol
at
o
and
K.
Vince
n
a,
“
A
Protot
yp
ing
Me
t
hodology
and
I
t
s
Tool
,”
Budde
R.
,
Kuhl
enka
m
p
K.,
M
a
t
h
i
a
s
s
e
n
L
.
,
Z
ü
l
l
i
g
h
o
v
e
n
H
.
(
E
d
s
)
A
p
p
r
o
a
c
h
e
s
t
o
P
r
o
t
o
t
y
p
i
n
g
.
S
p
r
i
n
g
e
r
,
B
e
r
l
i
n
,
H
e
i
d
e
l
b
e
r
g
,
1
9
8
4
.
[14]
C.
Y.
W
ong,
et
al.
,
“
Inte
rfa
ce
d
esign
pra
ct
i
ce
a
nd
educat
ion
to
wards
m
obil
e
ap
ps
deve
lopment
,
”
Proc
edi
a
-
Soc
i
al
and
Be
ha
vi
oral
Sci
en
ce
s
,
vol.
51
,
pp
.
698
-
702
,
2
012
.
[15]
E
.
Aban
a
,
et
al
.
,
“
El
ec
tron
ic
G
lov
e:
A
Te
a
chi
ng
AID
for
the
Hea
r
i
ng
Im
pai
red
,
”
In
te
rnational
Jour
nal
of
Elec
tri
cal
and
Computer
E
ngine
ering
(
IJECE)
,
v
ol. 8
(4)
,
p
p.
2290
-
2298
,
2
018.
BIOGR
AP
H
I
ES
OF
A
UTH
ORS
Hanay
anti
Hafit
is
cur
ren
tly
a
lectur
er
in
Dep
artm
ent
of
W
eb
Te
chnol
og
y
,
Fa
cul
t
y
of
Com
puter
Scie
nc
e
and
Inform
at
ion
Te
chn
olog
y
,
Univer
sit
i
Tun
Hus
sein
Onn
Malay
sia
.
Her
are
as
of
int
er
est
in
cl
ude
Us
abi
li
t
y
&
Us
e
r
Expe
ri
ence
in
Hum
an
Com
put
er
Int
era
c
ti
on
,
w
eb
and
m
obile
appl
i
ca
t
ion.
Ch
iam
Wui
X
ia
ng
is a
stud
ent i
n
Depa
rtment
of
W
eb
Technol
og
y
,
Facu
lty
of
Co
m
pute
r
Scie
n
ce
and
Inform
at
ion
Te
ch
nolog
y
,
Unive
rsiti
Tun
Hus
sein
Onn
Malay
s
ia.
He
will
recei
ve
his
Bac
he
lor
Degr
ee i
n
Com
pute
r
Sci
enc
e
m
aj
oring
in
W
eb
T
ec
hnolog
y
in
2019
.
Munirah
Mohd
Yusof
is
cur
ren
tly
a
l
ec
tur
er
in
Depa
rtment
of
Software
Engi
n
eering,
Facult
y
of
Com
pute
r
Scie
n
ce
and
Inform
ation
T
ec
hnolog
y
,
Univer
siti
Tun
Hus
sein
Onn
Malay
s
ia.
Her
are
as
o
f
in
te
r
est inc
lud
e
Exp
ert Sy
stem
and
Softw
are
Engi
n
ee
ring
.
Norfar
adill
a
W
ahid
is
cur
ren
tly
a
senior
lectur
e
r
in
Depa
rtment
o
f
W
eb
Te
chnol
o
g
y
,
Facul
t
y
of
Com
pute
r
Scie
n
ce
and
Inform
ation
T
ec
hnolog
y
,
Univer
siti
Tun
Hus
sein
Onn
Malay
s
ia.
Her
are
as
o
f
in
te
r
est inc
lud
e
XM
L te
chnol
og
y
,
sem
an
ti
c
web and
In
te
r
net
of
Th
ings.
Sh
ahree
n
Kass
im
is
cur
ren
tly
an
associa
te
p
rofe
ss
or
in
Dep
art
m
ent
of
W
eb
Te
chnol
og
y
,
Facul
t
y
of
Com
pute
r
Scie
nce
and
Inform
at
ion
Te
chnol
og
y
,
Univer
siti
Tu
n
Hus
sein
Onn
Malay
s
ia.
Her
a
rea
s
of
int
er
est
inc
lud
e
bioi
nfor
m
at
ic
s,
soft
co
m
puti
ng,
data
m
ini
ng,
web
an
d
m
obi
le
app
li
c
ati
on.
Evaluation Warning : The document was created with Spire.PDF for Python.