I
AE
S In
t
er
na
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
1
5
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
6
,
p
p
.
20
~
43
I
SS
N:
2
2
5
2
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
15
.i
1
.
p
p
20
-
43
20
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
Ev
o
lutiona
ry
tren
ds in au
tom
a
tic
speech r
ecog
nition
wit
h
a
rtif
icia
l in
tellige
nce:
a
sy
stema
t
ic
literature
review
G
a
b
ri
el
O
lu
wa
t
o
bi
So
b
o
l
a
1
,
E
mm
a
nu
el
Ad
et
iba
1
,
2
,
O
l
a
b
o
de
I
d
o
wu
-
B
is
ma
r
k
1
,
Abdu
l
t
a
o
f
ee
k Ab
a
y
o
mi
3
,
4
,
Ra
y
m
o
nd
J
ules
K
a
la
5
,
Su
re
nd
ra
Co
lin
T
ha
k
ur
6
,
Sib
u
s
is
o
M
o
y
o
7
1
D
e
p
a
r
t
m
e
n
t
o
f
El
e
c
t
r
i
c
a
l
a
n
d
I
n
f
o
r
m
a
t
i
o
n
E
n
g
i
n
e
e
r
i
n
g
,
C
o
v
e
n
a
n
t
A
p
p
l
i
e
d
I
n
f
o
r
m
a
t
i
c
s
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
A
f
r
i
c
a
n
C
e
n
t
e
r
o
f
E
x
c
e
l
l
e
n
c
e
,
C
o
v
e
n
a
n
t
U
n
i
v
e
r
si
t
y
,
O
g
u
n
S
t
a
t
e
,
N
i
g
e
r
i
a
2
F
a
c
u
l
t
y
o
f
En
g
i
n
e
e
r
i
n
g
a
n
d
T
h
e
B
u
i
l
t
En
v
i
r
o
n
me
n
t
,
D
u
r
b
a
n
U
n
i
v
e
r
s
i
t
y
o
f
Te
c
h
n
o
l
o
g
y
,
D
u
r
b
a
n
,
S
o
u
t
h
A
f
r
i
c
a
3
F
a
c
u
l
t
y
o
f
En
g
i
n
e
e
r
i
n
g
,
B
u
i
l
t
E
n
v
i
r
o
n
me
n
t
a
n
d
I
n
f
o
r
mat
i
o
n
Te
c
h
n
o
l
o
g
y
,
W
a
l
t
e
r
S
i
s
u
l
u
U
n
i
v
e
r
si
t
y
,
Ea
s
t
Lo
n
d
o
n
,
S
o
u
t
h
A
f
r
i
c
a
4
D
e
p
a
r
t
me
n
t
o
f
I
n
f
o
r
mat
i
o
n
Te
c
h
n
o
l
o
g
y
,
S
u
mm
i
t
U
n
i
v
e
r
s
i
t
y
,
O
f
f
a
,
N
i
g
e
r
i
a
5
I
n
t
e
r
n
a
t
i
o
n
a
l
U
n
i
v
e
r
si
t
y
o
f
G
r
a
n
d
-
B
a
ssam,
G
r
a
n
d
-
B
a
s
sam,
C
ô
t
e
d
'I
v
o
i
r
e
6
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
U
n
i
v
e
r
s
i
t
y
o
f
S
o
u
t
h
A
f
r
i
c
a
,
P
r
e
t
o
r
i
a
,
S
o
u
t
h
A
f
r
i
c
a
7
D
e
p
a
r
t
me
n
t
o
f
M
a
t
h
e
m
a
t
i
c
a
l
S
c
i
e
n
c
e
s,
S
c
h
o
o
l
f
o
r
D
a
t
a
S
c
i
e
n
c
e
a
n
d
C
o
m
p
u
t
a
t
i
o
n
a
l
T
h
i
n
k
i
n
g
,
S
t
e
l
l
e
n
b
o
s
c
h
U
n
i
v
e
r
si
t
y
,
S
t
e
l
l
e
n
b
o
s
c
h
,
S
o
u
t
h
A
f
r
i
c
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
l 2
9
,
2
0
2
4
R
ev
is
ed
Oct
3
1
,
2
0
2
5
Acc
ep
ted
No
v
8
,
2
0
2
5
H
u
m
a
n
b
e
i
n
g
s
d
e
p
e
n
d
g
r
e
a
t
l
y
o
n
c
o
m
m
u
n
i
c
a
t
i
o
n
a
n
d
c
o
n
t
i
n
u
a
l
l
y
s
e
e
k
w
a
y
s t
o
o
v
e
r
c
o
m
e
la
n
g
u
a
g
e
b
a
r
r
i
e
rs
.
Au
t
o
m
a
t
i
c
s
p
e
e
c
h
re
c
o
g
n
i
t
i
o
n
(
A
S
R
)
h
a
s
e
m
e
r
g
e
d
a
s
a
v
i
t
a
l
t
o
o
l
f
o
r
e
n
h
a
n
c
i
n
g
h
u
m
a
n
i
n
t
e
r
a
c
t
i
o
n
.
E
a
r
l
y
A
S
R
r
e
s
e
a
r
c
h
r
e
l
i
e
d
o
n
p
r
o
b
a
b
i
l
i
s
t
i
c
m
o
d
e
l
s
,
p
a
rt
i
c
u
l
a
r
l
y
t
h
e
h
i
d
d
e
n
M
a
r
k
o
v
m
o
d
e
l
(
H
M
M
)
a
n
d
G
a
u
ss
i
a
n
m
i
x
t
u
r
e
m
o
d
e
l
(G
M
M
)
,
w
i
t
h
m
e
l
-
f
r
e
q
u
e
n
c
y
c
e
p
s
t
ra
l
c
o
e
f
f
i
c
i
e
n
t
s
(
M
F
C
C
s
)
f
o
r
fe
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
,
l
e
a
d
i
n
g
t
o
t
h
e
c
r
e
a
t
i
o
n
o
f
A
u
d
r
e
y
a
t
B
e
l
l
L
a
b
o
r
a
t
o
r
i
e
s
.
S
u
b
s
e
q
u
e
n
t
l
y
,
a
r
t
i
f
ic
i
a
l
i
n
t
e
l
l
i
g
e
n
c
e
(
A
I
)
a
p
p
r
o
a
c
h
e
s
,
e
s
p
e
c
ia
l
l
y
d
e
e
p
l
e
a
r
n
i
n
g
,
h
a
v
e
t
r
a
n
s
f
o
rm
e
d
A
S
R
a
n
d
p
r
o
d
u
c
e
d
s
y
s
t
e
m
s
s
u
c
h
a
s
Ja
s
p
e
r
,
W
h
i
s
p
e
r
,
G
o
o
g
l
e
A
s
s
is
t
a
n
t
,
M
i
c
r
o
s
o
f
t
C
o
r
t
a
n
a
,
A
p
p
l
e
S
i
r
i
,
a
n
d
A
m
a
z
o
n
A
l
e
x
a
.
T
h
i
s
p
a
p
e
r
p
r
e
se
n
t
s
a
s
y
s
tem
a
t
i
c
l
i
te
r
a
t
u
re
r
e
v
i
e
w
t
h
a
t
e
x
a
m
in
e
s
A
S
R
’
s
e
v
o
l
u
t
i
o
n
,
t
h
e
A
I
a
r
c
h
i
t
e
c
t
u
r
e
s
e
m
p
l
o
y
e
d
,
t
h
e
i
r
f
e
a
t
u
r
e
s
,
s
t
re
n
g
t
h
s
a
n
d
w
e
a
k
n
e
ss
e
s
,
a
n
d
t
h
e
p
e
r
f
o
rm
a
n
c
e
g
a
i
n
s
a
c
h
i
e
v
e
d
s
i
n
c
e
A
I
wa
s
i
n
t
e
g
r
a
t
e
d
i
n
t
o
p
r
o
b
a
b
i
l
i
s
t
i
c
m
o
d
e
l
l
i
n
g
.
A
s
n
o
w
b
a
l
l
i
n
g
a
p
p
r
o
a
c
h
w
a
s
u
s
e
d
t
o
i
d
e
n
t
i
fy
r
e
l
e
v
a
n
t
s
t
u
d
i
e
s
f
r
o
m
G
o
o
g
l
e
S
c
h
o
l
a
r
a
n
d
S
c
o
p
u
s
t
o
a
d
d
r
e
s
s
f
i
v
e
r
e
s
e
a
r
c
h
q
u
e
s
t
i
o
n
s
,
i
t
e
r
a
t
i
n
g
t
h
r
o
u
g
h
b
a
c
k
w
a
r
d
a
n
d
f
o
r
w
a
r
d
s
e
a
r
c
h
e
s
u
n
t
i
l
n
o
n
e
w
i
n
f
o
r
m
a
t
i
o
n
w
a
s
f
o
u
n
d
.
F
i
n
d
i
n
g
s
r
e
v
e
a
l
t
h
a
t
A
S
R
d
a
t
e
s
b
a
c
k
t
o
t
h
e
1
9
2
0
s
w
i
t
h
t
h
e
R
a
d
i
o
R
e
x
t
o
y
a
n
d
h
a
s
s
i
n
c
e
a
d
v
a
n
c
e
d
t
h
r
o
u
g
h
a
r
c
h
i
t
e
c
t
u
r
e
s
i
n
c
l
u
d
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
,
r
e
c
u
r
r
e
n
t
n
e
u
r
a
l
n
e
t
w
o
r
k
s
(
R
NN
)
,
s
u
p
p
o
r
t
v
e
c
t
o
r
m
a
c
h
i
n
e
s
(S
V
M
)
,
a
n
d
t
r
a
n
s
f
o
r
m
e
r
s
,
a
l
l
c
o
n
t
r
i
b
u
t
i
n
g
t
o
i
m
p
r
o
v
e
d
p
e
r
f
o
r
m
a
n
c
e
m
e
a
s
u
r
e
d
b
y
r
e
d
u
c
e
d
w
o
r
d
e
r
r
o
r
r
a
t
e
s
(
W
E
R
)
.
K
ey
w
o
r
d
s
:
Ar
tific
ial
in
tellig
en
ce
Au
to
m
atic
s
p
ee
ch
r
ec
o
g
n
itio
n
Dee
p
lear
n
in
g
Hid
d
en
Ma
r
k
o
v
m
o
d
el
L
an
g
u
ag
e
m
o
d
el
Ma
ch
in
e
lear
n
in
g
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
E
m
m
an
u
el
Ad
etib
a
Dep
ar
tm
en
t o
f
E
lectr
ical
an
d
I
n
f
o
r
m
atio
n
E
n
g
in
ee
r
in
g
C
o
v
en
an
t A
p
p
lied
I
n
f
o
r
m
atics a
n
d
C
o
m
m
u
n
icatio
n
Af
r
ica
n
C
en
ter
o
f
E
x
ce
llen
ce
,
C
o
v
en
a
n
t U
n
iv
er
s
ity
Km
.
1
0
I
d
ir
o
k
o
R
o
ad
,
C
an
aa
n
L
an
d
,
Ota,
Og
u
n
State,
Nig
er
i
a
E
m
ail: e
m
m
an
u
el.
ad
etib
a@
co
v
en
an
tu
n
iv
er
s
ity
.
ed
u
.
n
g
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
d
e
v
elo
p
m
e
n
t
o
f
au
t
o
m
at
ic
s
p
ee
ch
r
ec
o
g
n
itio
n
(
ASR
)
f
o
r
h
u
m
an
to
m
ac
h
in
e
c
o
m
m
u
n
icatio
n
s
tar
ted
o
v
er
4
0
y
e
ar
s
ag
o
[
1
]
–
[
3
]
.
T
h
e
n
,
it
ex
p
lo
r
e
d
th
e
p
r
o
b
ab
ilis
tic
ap
p
r
o
ac
h
es
s
u
ch
as
th
e
h
id
d
en
Ma
r
k
o
v
m
o
d
el
(
HM
M)
,
Gau
s
s
ian
m
ix
tu
r
e
m
o
d
el
(
GM
M)
co
u
p
le
d
with
m
el
-
f
r
eq
u
e
n
cy
ce
p
s
tr
al
co
e
f
f
icien
ts
(
MFC
C
s
)
f
o
r
f
ea
tu
r
e
e
x
tr
ac
tio
n
[
1
]
.
T
h
e
s
tate
o
f
th
e
ar
t
ASR
s
y
s
te
m
s
wh
ich
ar
e
HM
M
b
ased
with
s
tate
p
r
o
b
ab
ilit
y
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
E
vo
lu
tio
n
a
r
y
tr
en
d
s
in
a
u
to
m
a
tic
s
p
ee
ch
r
ec
o
g
n
itio
n
w
ith
…
(
Ga
b
r
iel
Olu
w
a
to
b
i
S
o
b
o
l
a
)
21
d
is
tr
ib
u
tio
n
m
o
d
elled
u
s
in
g
t
h
e
GM
M
co
u
ld
n
o
t
im
p
r
o
v
e
t
h
e
p
er
f
o
r
m
a
n
ce
o
f
an
ASR
s
y
s
tem
[
1
]
,
[
4
]
.
T
h
is
ca
n
b
e
en
h
a
n
ce
d
v
ia
t
h
e
ar
ti
f
icial
n
eu
r
al
n
etwo
r
k
s
(
ANN)
an
d
HM
M
h
y
b
r
id
s
y
s
tem
.
T
h
e
p
r
e
an
d
p
o
s
t
p
r
o
ce
s
s
in
g
tech
n
iq
u
es
i
n
v
o
l
v
in
g
f
ea
tu
r
e
ex
tr
ac
tio
n
,
la
n
g
u
ag
e
m
o
d
el,
p
r
o
n
u
n
ciatio
n
m
o
d
el,
an
d
m
o
d
el
ad
ap
tab
ilit
y
in
v
o
lv
in
g
s
p
ea
k
e
r
,
ch
a
n
n
el
a
n
d
n
o
is
e
v
ar
iatio
n
s
ar
e
th
e
id
en
tifie
d
s
tag
es
wh
er
e
p
er
f
o
r
m
a
n
ce
im
p
r
o
v
em
e
n
t is n
ee
d
ed
[
5
]
.
T
h
e
e
x
p
lo
r
a
ti
o
n
o
f
ar
tif
ici
al
i
n
te
lli
g
e
n
ce
(
AI
)
f
o
r
s
p
e
ec
h
r
e
co
g
n
iti
o
n
s
ta
r
t
ed
a
b
o
u
t
f
iv
e
d
ec
a
d
es
ag
o
[
2
]
,
[
3
]
w
h
il
e
t
h
e
u
s
e
o
f
ANN
i
n
v
o
lv
in
g
d
e
ep
n
eu
r
a
l
n
etw
o
r
k
s
(
DNN
)
(
a
s
u
b
s
et
o
f
m
ac
h
i
n
e
l
ea
r
n
in
g
)
w
h
ic
h
ar
e
m
o
s
tl
y
u
n
s
u
p
e
r
v
is
e
d
in
n
at
u
r
e
f
o
r
ASR
d
e
v
e
lo
p
m
e
n
t
s
ta
r
te
d
in
2
0
0
6
[
1
]
.
T
h
is
was
a
s
h
i
f
t
in
th
e
d
e
v
e
lo
p
m
e
n
t
tec
h
n
i
q
u
e
o
f
ASR
f
r
o
m
t
h
e
p
r
o
b
a
b
ilis
tic
ap
p
r
o
ac
h
.
Ma
ch
in
e
lea
r
n
i
n
g
(
a
s
u
b
s
et
o
f
A
I
)
is
th
e
a
b
il
it
y
o
f
m
ac
h
i
n
es
o
r
c
o
m
p
u
te
r
s
t
o
le
a
r
n
f
r
o
m
i
n
p
u
t
d
ata
wi
th
o
u
t
b
ei
n
g
ex
p
l
ici
tly
p
r
o
g
r
am
m
e
d
t
o
d
o
s
o
.
I
t
h
as
p
r
o
d
u
ce
d
l
o
ts
o
f
tec
h
n
o
lo
g
i
es
s
u
c
h
as
DNN
f
o
r
t
h
e
d
ev
el
o
p
m
en
t
o
f
s
p
e
ec
h
r
e
co
g
n
iti
o
n
.
T
h
er
e
ar
e
r
ec
e
n
t
a
d
v
a
n
c
es
in
m
ac
h
i
n
e
lea
r
n
i
n
g
u
s
in
g
D
NN
d
u
e
to
i
ts
a
b
ili
ty
t
o
a
cc
o
m
m
o
d
at
e
h
u
g
e
s
iz
e
o
f
tr
ai
n
i
n
g
d
at
a
a
n
d
i
ts
h
i
g
h
l
y
in
cr
ea
s
ed
p
r
o
ce
s
s
i
n
g
a
b
i
liti
es
o
f
c
o
m
p
u
te
r
ch
ip
s
[
1
]
.
I
t
c
o
m
es
i
n
th
e
f
o
r
m
o
f
r
e
cu
r
r
e
n
t
n
eu
r
al
n
et
wo
r
k
(
R
NN)
,
co
n
v
o
lu
ti
o
n
al
n
e
u
r
al
n
etw
o
r
k
(
C
NN
)
,
a
n
d
ti
m
e
d
el
ay
n
e
u
r
al
n
etw
o
r
k
(
T
DNN
)
f
o
r
th
e
m
o
d
el
li
n
g
o
f
ASR
,
a
n
d
s
o
m
eti
m
es
c
o
u
p
le
d
wi
th
t
h
e
H
MM
.
R
ese
ar
ch
w
o
r
k
s
h
av
e
b
e
en
ca
r
r
ie
d
o
u
t
o
n
d
i
f
f
e
r
e
n
t
ar
c
h
it
ec
t
u
r
es
o
f
h
y
b
r
i
d
HM
M
an
d
ANN
t
o
m
o
d
el
A
SR
s
y
s
t
em
s
.
T
h
ese
a
r
e
t
h
e
m
u
lt
i
-
b
a
n
d
h
y
b
r
i
d
HM
M
+A
N
N
,
a
ll
-
c
o
m
b
i
n
ati
o
n
s
m
u
lti
-
b
a
n
d
HM
M
+A
NN
h
y
b
r
i
d
,
all
-
c
o
m
b
in
ati
o
n
s
m
u
lti
-
s
t
r
ea
m
HM
M
+A
N
N
h
y
b
r
i
d
,
t
h
e
m
u
lt
i
-
s
t
r
ea
m
ta
n
d
e
m
HM
M+
ANN
h
y
b
r
i
d
,
a
n
d
t
h
e
n
ar
r
o
w
-
b
a
n
d
ta
n
d
e
m
H
M
M+
ANN
h
y
b
r
i
d
,
wh
ic
h
a
r
e
ex
p
l
o
i
te
d
o
v
e
r
t
h
e
p
r
o
b
a
b
i
lis
ti
c
a
p
p
r
o
ac
h
i
n
v
o
l
v
in
g
t
h
e
c
o
n
v
en
ti
o
n
al
h
y
b
r
i
d
HM
M
a
n
d
G
MM
[
5
]
in
o
r
d
er
to
im
p
r
o
v
e
t
h
e
p
e
r
f
o
r
m
a
n
c
e
o
f
ASR
.
Fi
r
m
s
l
ik
e
O
p
e
n
A
I
h
as
p
r
o
d
u
ce
d
th
e
W
h
is
p
e
r
ASR
t
h
a
t
e
x
p
l
o
r
e
d
d
e
ep
le
ar
n
i
n
g
f
o
r
m
o
d
elli
n
g
ASR
[
6
]
,
w
h
il
e
NV
I
D
I
A
p
r
o
d
u
ce
d
t
h
e
J
as
p
e
r
AS
R
th
a
t u
s
es
5
4
c
o
n
v
o
l
u
ti
o
n
a
l la
y
e
r
s
f
o
r
t
r
a
in
in
g
[
7
]
.
Oth
er
d
e
ep
l
ea
r
n
in
g
te
c
h
n
iq
u
e
s
t
h
at
h
a
v
e
b
ee
n
u
s
e
d
f
o
r
AS
R
a
r
e
t
h
e
s
el
f
-
o
r
g
a
n
i
zi
n
g
m
a
p
(
S
OM
)
,
r
a
d
ia
l
b
as
is
f
u
n
cti
o
n
(
R
B
F)
,
m
u
lt
ila
y
e
r
p
er
ce
p
t
r
o
n
(
ML
P
)
,
as
w
ell
as
s
u
p
p
o
r
t
v
e
ct
o
r
m
ac
h
i
n
e
(
SVM
)
[
2
]
.
T
h
e
ASR
s
y
s
tem
h
as
f
o
u
n
d
a
p
p
licatio
n
in
v
ar
io
u
s
f
ield
s
in
clu
d
in
g
h
ea
lth
,
m
ilit
ar
y
,
d
is
ab
led
p
eo
p
le,
an
d
tech
n
o
l
o
g
ies
[
3
]
.
Fo
r
in
s
tan
ce
,
in
th
e
h
ea
lth
s
ec
to
r
,
it
h
as
b
ee
n
im
p
lem
en
ted
f
o
r
m
ed
ical
d
o
cu
m
en
tatio
n
,
m
ak
in
g
tr
ea
tm
en
t
n
o
tes,
an
d
a
s
a
s
u
r
g
ical
ass
i
s
tan
t.
I
n
th
e
m
ilit
ar
y
,
it
ca
n
b
e
u
s
ed
to
co
m
m
an
d
an
a
u
to
p
ilo
t
,
s
et
r
ad
io
f
r
eq
u
e
n
cies,
co
n
tr
o
l
th
e
p
ar
am
eter
s
o
f
wea
p
o
n
s
b
ein
g
r
elea
s
ed
;
wh
ile
in
tec
h
n
o
lo
g
y
,
th
er
e
ar
e
d
if
f
er
en
t
ASR
s
y
s
tem
s
s
u
ch
a
s
th
e
Go
o
g
le
Ass
is
tan
t,
Mic
r
o
s
o
f
t
C
o
r
tan
a,
Ap
p
le
Siri,
an
d
Am
az
o
n
Alex
a
[
8
]
th
at
ar
e
b
ein
g
u
s
ed
b
y
in
d
iv
i
d
u
al
f
o
r
p
er
s
o
n
alize
d
lear
n
in
g
,
in
ter
ac
tiv
e
lear
n
in
g
,
ad
a
p
tiv
e
lear
n
in
g
,
lan
g
u
ag
e
lear
n
in
g
;
as
well
as
aid
in
g
r
ea
d
in
g
an
d
en
g
ag
em
e
n
t
o
f
d
is
ab
led
p
eo
p
le
with
h
ea
r
in
g
an
d
w
r
itin
g
d
if
f
ic
u
lties
in
d
is
cu
s
s
io
n
[
3
]
.
I
n
ad
d
itio
n
,
th
er
e
ar
e
also
o
th
er
o
n
lin
e
ASR
to
o
l
k
its
s
u
ch
as
th
e
HT
K,
S
p
h
in
x
,
Kald
i,
C
MU
L
M
to
o
lk
it,
an
d
SR
I
L
M
th
at
ar
e
av
ailab
le
to
b
u
ild
a
wo
r
k
in
g
ASR
s
y
s
tem
[
4
]
.
T
h
e
ty
p
e
o
f
d
ataset
an
d
th
e
av
ailab
ilit
y
o
f
d
ata
is
an
o
th
er
f
ac
to
r
th
at
n
ee
d
s
to
b
e
co
n
s
id
er
ed
in
AS
R
d
ev
elo
p
m
e
n
t
[
9
]
–
[
1
2
]
.
T
h
er
e
ar
e
lo
ts
o
f
s
p
ee
ch
d
ataset
lik
e
th
e
L
ib
r
esp
ee
c
h
d
ataset,
Natio
n
al
I
n
s
titu
te
o
f
Stan
d
ar
d
s
an
d
T
ec
h
n
o
lo
g
y
(
NI
ST
)
s
p
ee
ch
d
ataset,
Hu
b
,
T
ex
as
I
n
s
tr
u
m
e
n
ts
Ma
s
s
ac
h
u
s
etts
I
n
s
titu
te
o
f
T
ec
h
n
o
lo
g
y
(
TIMIT
)
,
T
E
D
-
L
I
UM
,
C
o
m
m
o
n
v
o
ice,
T
h
e
s
p
o
k
en
W
ik
ip
ed
ia,
C
STR
VC
T
K,
Ais
h
ell
-
I
,
Per
s
ian
co
n
s
o
n
an
t
v
o
wel
co
m
b
in
atio
n
(
PC
V
C
)
,
an
d
th
e
Ar
ab
ic
Sp
ee
ch
C
o
r
p
u
s
[
1
3
]
–
[
1
6
]
.
T
h
is
in
d
icate
s
th
at
lo
ts
o
f
ef
f
o
r
ts
h
av
e
b
ee
n
p
u
t
in
p
lac
e
to
war
d
s
ac
h
iev
in
g
th
e
u
ltima
te
g
o
al
o
f
an
ASR
s
y
s
tem
in
p
er
f
o
r
m
i
n
g
as
a
h
u
m
an
lis
ten
er
[
2
]
,
[
1
7
]
.
T
h
e
av
ailab
ilit
y
an
d
s
ize
o
f
d
ata
h
av
e
a
g
r
ea
ter
i
n
f
lu
en
ce
o
n
th
e
p
er
f
o
r
m
an
ce
o
f
an
ASR
s
y
s
tem
.
I
t
was
s
h
o
wn
in
[
4
]
,
th
at
th
e
p
er
f
o
r
m
an
ce
o
f
ASR
s
y
s
tem
im
p
r
o
v
e
d
s
in
ce
1
9
7
6
v
ia
a
r
ed
u
ctio
n
in
th
e
wo
r
d
er
r
o
r
r
ate
(
W
E
R
)
b
ec
au
s
e
o
f
th
e
av
ailab
ilit
y
o
f
m
o
r
e
d
ata
at
p
r
esen
t th
an
t
h
e
p
r
ev
io
u
s
y
ea
r
s
.
Dif
f
er
en
t
p
e
r
f
o
r
m
an
ce
m
ea
s
u
r
es
h
av
e
b
ee
n
e
x
p
lo
ited
t
o
ev
al
u
ate
th
e
p
er
f
o
r
m
an
ce
o
f
an
A
SR
s
y
s
tem
[
1
8
]
–
[
2
3
]
.
T
h
e
W
E
R
is
u
s
ed
to
ev
alu
ate
th
e
ac
cu
r
ac
y
,
a
n
d
t
h
e
r
ea
l
tim
e
f
ac
to
r
(
R
T
F)
to
e
v
alu
ate
th
e
s
p
ee
d
o
f
th
e
ASR
.
Oth
er
p
r
ec
is
io
n
m
e
asu
r
es
in
clu
d
e:
co
m
m
an
d
s
u
c
ce
s
s
r
ate
(
C
SR
)
,
f
r
am
e
er
r
o
r
r
ate
(
FER),
NI
ST
d
etec
tio
n
co
s
t
f
u
n
ctio
n
(
DC
F),
r
ec
o
g
n
itio
n
r
ate
(
R
R
)
,
co
n
ce
p
t
v
alu
e
e
r
r
o
r
r
ate
(
C
VE
R
)
,
co
n
ce
p
t
er
r
o
r
r
ate
(
C
E
R
)
,
p
h
o
n
e
r
ec
o
g
n
itio
n
er
r
o
r
r
ates
(
PR
E
R
)
,
an
d
f
r
am
e
class
if
icatio
n
(
FR
E
R
)
,
wo
r
d
r
ec
o
g
n
itio
n
r
ate
(
W
R
R
)
,
p
h
o
n
e
er
r
o
r
r
ate
(
PE
R
)
,
d
B
SIR
g
ain
,
lab
el
er
r
o
r
r
ate
(
L
E
R
)
,
p
h
o
n
e
m
e
class
if
i
ca
tio
n
p
er
f
o
r
m
an
ce
(
PC
P)
,
r
o
o
t
m
ea
n
s
q
u
ar
e
er
r
o
r
(
R
MSE
)
,
s
en
ten
ce
ac
cu
r
ac
y
,
q
u
er
y
er
r
o
r
r
ate
(
QE
R
)
,
u
n
wei
g
h
ted
class
if
icatio
n
ac
cu
r
ac
y
an
d
,
g
ai
n
in
d
B
[
1
]
,
[
2
]
.
T
h
e
r
em
ain
in
g
p
ar
ts
o
f
th
is
ar
ticle
is
s
tr
u
ctu
r
ed
as
f
o
llo
ws:
s
ec
t
io
n
2
co
n
tain
s
th
e
m
eth
o
d
o
lo
g
y
ap
p
lied
f
o
r
t
h
e
s
tu
d
y
.
T
h
e
r
e
s
u
lts
o
b
tain
ed
ar
e
p
r
esen
ted
i
n
s
ec
tio
n
3
.
T
h
e
d
is
cu
s
s
io
n
as
tailo
r
ed
to
ea
ch
r
esear
ch
q
u
esti
o
n
i
n
th
e
s
tu
d
y
is
co
n
tain
ed
in
s
ec
tio
n
4
.
Fin
al
ly
,
s
ec
tio
n
5
p
r
esen
ts
th
e
co
n
cl
u
s
io
n
.
2.
M
E
T
H
O
D
I
n
a
q
u
est
to
ca
r
r
y
o
u
t
a
th
o
r
o
u
g
h
s
y
s
tem
atic
r
ev
iew
o
n
th
e
ev
o
lu
tio
n
ar
y
tr
en
d
o
n
th
e
ASR
with
AI
,
th
er
e
is
th
e
n
ee
d
to
ad
d
r
ess
th
e
to
p
ic
b
y
d
ev
elo
p
in
g
r
esear
c
h
q
u
esti
o
n
s
to
g
u
id
e
an
d
s
er
v
e
as
th
e
f
o
cu
s
o
f
th
is
p
ap
er
.
T
h
ese
s
ets
o
f
q
u
esti
o
n
s
wer
e
d
ev
el
o
p
ed
to
h
av
e
d
ee
p
er
in
s
ig
h
t
o
n
th
e
to
p
ic.
L
o
ts
o
f
r
esear
ch
wo
r
k
s
h
av
e
b
ee
n
s
tu
d
ied
in
th
is
ar
ea
,
an
d
d
if
f
er
en
t r
esear
c
h
er
s
h
av
e
ap
p
r
o
ac
h
ed
th
e
ASR
s
y
s
tem
in
d
if
f
er
e
n
t w
ay
s
to
ac
h
iev
e
s
im
ilar
r
esu
lts
in
ter
m
s
o
f
g
en
er
atin
g
th
e
tex
t
eq
u
iv
alen
ts
o
f
th
e
s
p
ee
ch
s
ig
n
als.
Hen
ce
5
r
esear
ch
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
6
:
20
-
43
22
q
u
esti
o
n
s
wer
e
d
ev
elo
p
ed
an
d
ar
e
tailo
r
ed
to
war
d
s
th
e
e
v
o
l
u
tio
n
o
f
AI
with
ASR
,
th
e
d
if
f
er
en
t
ty
p
es
o
f
A
I
th
at
h
av
e
b
ee
n
ex
p
lo
r
ed
,
t
h
eir
b
e
n
ef
its
an
d
d
em
e
r
its
,
an
d
p
er
f
o
r
m
a
n
ce
en
h
an
ce
m
en
t
o
b
tain
ed
b
y
r
esear
ch
er
s
-
th
ese
ar
e
in
a
q
u
es
t to
m
ak
e
f
in
d
in
g
s
in
r
eg
ar
d
to
ASR
ev
o
lu
tio
n
with
AI
.
Af
ter
d
ev
el
o
p
in
g
th
ese
s
et
o
f
q
u
esti
o
n
s
,
th
e
s
elec
tio
n
o
f
p
ap
er
s
was
ca
r
r
ie
d
u
s
in
g
th
e
s
n
o
wb
allin
g
ap
p
r
o
ac
h
.
J
o
u
r
n
al
ar
ticles
an
d
r
ev
iew
p
a
p
er
s
wer
e
s
o
u
r
ce
d
an
d
s
elec
ted
f
r
o
m
th
e
d
ata
b
ases
o
f
Sco
p
u
s
an
d
Go
o
g
le
Sch
o
lar
.
T
h
ese
s
et
o
f
p
ap
er
s
f
o
r
m
e
d
th
e
in
itial
s
tar
t
s
et
p
ap
er
s
.
T
h
e
s
n
o
wb
allin
g
ap
p
r
o
ac
h
h
as
th
e
b
ac
k
war
d
s
n
o
wb
allin
g
ap
p
r
o
a
ch
an
d
f
o
r
war
d
s
n
o
wb
allin
g
a
p
p
r
o
ac
h
wh
ich
wer
e
e
x
p
lo
r
e
d
f
o
r
th
e
s
elec
tio
n
o
f
p
ap
er
s
f
r
o
m
th
e
in
itial
s
tar
t
s
ets.
T
h
e
b
ac
k
war
d
ap
p
r
o
ac
h
in
v
o
lv
es
s
elec
tio
n
o
f
p
ap
er
s
cite
d
in
th
e
wo
r
k
wh
ile
p
ap
er
s
wer
e
s
elec
ted
f
r
o
m
th
e
citatio
n
s
o
f
ea
ch
p
ap
er
o
n
th
eir
Sco
p
u
s
d
atab
ase
u
s
in
g
th
e
f
o
r
war
d
ap
p
r
o
ac
h
.
T
h
e
s
n
o
wb
allin
g
a
p
p
r
o
ac
h
is
illu
s
tr
ated
in
Fig
u
r
e
1
.
Pap
e
r
s
wer
e
an
aly
ze
d
an
d
s
elec
ted
o
r
r
ejec
ted
u
s
in
g
th
ei
r
titl
es,
au
th
o
r
s
o
f
th
e
p
ap
er
s
,
th
e
ab
s
tr
ac
t,
an
d
y
ea
r
s
.
T
h
e
s
elec
ted
p
ap
er
s
wer
e
r
ev
iewe
d
to
an
s
wer
th
e
r
esear
ch
q
u
esti
o
n
s
as
d
escr
ib
ed
in
th
e
r
esear
ch
q
u
esti
o
n
s
s
ec
tio
n
.
T
h
e
s
n
o
w
b
allin
g
ap
p
r
o
ac
h
was
r
u
n
th
r
o
u
g
h
3
iter
atio
n
s
f
r
o
m
th
e
1
3
s
tar
t
s
et
p
ap
er
s
s
elec
ted
f
r
o
m
th
e
in
itial
1
8
p
a
p
er
s
u
n
til
n
o
ca
n
d
id
ate
p
ap
er
was
f
it
f
o
r
s
elec
tio
n
.
T
h
e
s
tep
s
tak
en
t
o
s
ea
r
ch
f
o
r
an
d
s
elec
t
p
ap
er
s
f
r
o
m
Go
o
g
le
Sch
o
lar
an
d
Sco
p
u
s
d
ata
b
ases
wer
e
in
alig
n
m
en
t
with
th
e
s
n
o
wb
allin
g
ap
p
r
o
ac
h
laid
d
o
wn
in
[
2
4
]
.
Fig
u
r
e
1
.
Sn
o
wb
allin
g
ap
p
r
o
a
ch
[
2
4
]
2
.
1
.
Resea
rc
h
qu
estio
ns
2
.
1
.
1
.
When did
t
he
dev
elo
p
m
ent
o
f
ASR
wit
h
AI
s
t
a
rt
?
T
h
e
d
ev
elo
p
m
en
t
o
f
s
p
ee
ch
r
ec
o
g
n
itio
n
d
em
a
n
d
s
th
at
a
s
p
ir
in
g
an
d
in
ter
ested
r
esear
ch
er
s
h
av
e
u
n
d
er
s
tan
d
i
n
g
an
d
in
s
ig
h
ts
in
to
wh
en
s
p
ee
ch
r
ec
o
g
n
itio
n
s
tar
ted
p
ar
ticu
lar
ly
with
th
e
ex
p
lo
r
atio
n
o
f
AI
.
B
ef
o
r
e
th
e
i
n
teg
r
atio
n
o
f
AI
f
o
r
th
e
ASR
,
d
if
f
er
e
n
t
s
tatis
tic
al
ap
p
r
o
ac
h
es
h
a
v
e
b
ee
n
u
tili
z
ed
to
m
o
d
el
s
p
ee
ch
r
ec
o
g
n
itio
n
.
T
h
is
q
u
esti
o
n
ten
d
s
to
d
ig
d
ee
p
er
in
to
d
ate
o
f
v
ar
io
u
s
AI
ap
p
r
o
ac
h
es
f
o
r
th
e
d
ev
elo
p
m
e
n
t
o
f
ASR
.
T
h
e
q
u
esti
o
n
p
r
o
v
id
es
v
ar
io
u
s
an
s
wer
s
to
wh
e
n
e
ac
h
o
f
AI
ap
p
r
o
ac
h
es
wer
e
in
teg
r
ated
f
o
r
ASR
m
o
d
ellin
g
.
Fo
r
in
s
tan
ce
,
it a
n
s
wer
s
wh
en
DL
s
tar
ted
to
g
ain
p
o
p
u
lar
ity
f
o
r
th
e
d
ev
el
o
p
m
en
t o
f
AI
,
th
e
d
ay
s
o
f
p
o
p
u
lar
ity
in
th
e
u
tili
z
atio
n
o
f
R
NN,
an
d
v
ar
io
u
s
r
o
b
u
s
t f
ea
tu
r
e
ex
tr
ac
tio
n
m
eth
o
d
s
.
2
.
1
.
2
.
W
h
a
t
t
y
p
e
s
o
f
AI
t
e
ch
ni
q
ue
s
h
a
v
e
b
e
en
e
x
pl
o
r
e
d
b
y
v
a
r
i
o
u
s
r
es
e
a
r
ch
e
rs
t
o
d
e
v
e
l
o
p
a
n
AS
R
s
y
s
t
e
m?
Dif
f
er
en
t
ty
p
es
o
f
AI
h
a
v
e
b
ee
n
u
tili
z
ed
to
m
o
d
el
s
p
ee
ch
r
e
c
o
g
n
itio
n
.
T
h
is
q
u
esti
o
n
g
iv
es
in
s
ig
h
t
f
o
r
th
e
in
ter
ested
r
ea
d
e
r
in
to
t
h
e
v
ar
io
u
s
ty
p
es
o
f
AI
th
at
f
in
d
ap
p
licatio
n
in
ASR
m
o
d
ell
in
g
.
Dif
f
er
e
n
t
d
ee
p
lear
n
in
g
ar
c
h
itectu
r
es
s
u
ch
as
th
e
C
NN,
R
N
N,
SVMs,
SOM
,
T
DNN,
lo
n
g
s
h
o
r
t
-
ter
m
m
e
m
o
r
y
(
L
STM
)
h
av
e
b
ee
n
u
s
ed
to
m
o
d
el
ASR
.
C
o
u
p
led
with
th
is
is
th
e
u
til
i
z
ati
o
n
o
f
d
if
f
e
r
en
t
f
ea
tu
r
e
e
x
tr
ac
tio
n
tech
n
iq
u
es
s
u
c
h
as
th
e
wav
elet
tr
an
s
f
o
r
m
,
lin
ea
r
p
r
ed
ictiv
e
co
d
i
n
g
(
L
PC
)
,
an
d
r
elativ
e
s
p
ec
tr
al
f
ilter
in
g
(
R
ASTA
)
to
m
o
d
el
ASR
th
at
ca
n
w
ith
s
tan
d
th
e
v
ar
y
in
g
e
n
v
ir
o
n
m
en
tal
co
n
d
itio
n
,
s
p
ea
k
er
v
ar
iatio
n
,
an
d
i
n
to
n
atio
n
.
Via
th
is
q
u
esti
o
n
,
r
esear
c
h
er
s
an
d
r
ea
d
er
s
ar
e
ab
le
to
h
av
e
d
ee
p
u
n
d
er
s
tan
d
in
g
o
f
th
e
d
if
f
er
e
n
t
ty
p
es
o
f
AI
b
ei
n
g
u
tili
z
ed
to
th
e
m
o
d
ellin
g
o
f
A
SR
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
E
vo
lu
tio
n
a
r
y
tr
en
d
s
in
a
u
to
m
a
tic
s
p
ee
ch
r
ec
o
g
n
itio
n
w
ith
…
(
Ga
b
r
iel
Olu
w
a
to
b
i
S
o
b
o
l
a
)
23
2
.
1
.
3
.
Wha
t
a
re
t
he
benef
it
s
a
nd
dis
a
dv
a
nta
g
es o
f
ex
plo
ri
ng
ea
ch
o
f
t
ho
s
e
a
l
g
o
rit
hm
s
m
ent
io
ned in 2
?
Dif
f
er
en
t
AI
alg
o
r
ith
m
s
h
a
v
e
b
ee
n
ex
p
lo
r
ed
f
o
r
th
e
m
o
d
el
lin
g
o
f
ASR
.
E
ac
h
o
f
th
ese
a
lg
o
r
ith
m
s
co
m
es
with
it
i
s
o
wn
b
en
ef
its
an
d
d
is
ad
v
a
n
tag
es.
T
h
is
q
u
est
io
n
ad
d
r
ess
es
th
eir
b
en
ef
its
s
u
ch
as
p
er
f
o
r
m
an
ce
en
h
an
ce
m
e
n
t in
ter
m
s
o
f
r
ed
u
ctio
n
in
W
E
R
co
m
p
ar
ed
to
th
e
s
tatis
t
ical
ap
p
r
o
ac
h
es,
ab
ilit
y
to
o
b
tain
o
p
tim
i
z
ed
r
esu
lts
;
an
d
lim
itatio
n
s
o
f
A
I
u
s
ag
e
s
u
ch
as
th
eir
in
f
lex
i
b
ilit
y
to
h
an
d
le
tim
i
n
g
v
ar
iab
ilit
y
,
in
ab
ilit
y
to
m
o
d
el
ASR
th
at
ca
n
m
o
d
el
n
o
is
e
v
a
r
iab
ilit
y
.
T
h
e
q
u
esti
o
n
also
e
x
p
lain
s
th
at
th
e
ch
o
ice
o
f
d
if
f
er
en
t
AI
ap
p
r
o
ac
h
co
m
es with
th
eir
b
en
e
f
its
wh
ich
m
ig
h
t
b
e
a
lim
itatio
n
in
o
th
e
r
AI
ap
p
r
o
ac
h
es.
2
.
1
.
4
.
Do
es
t
he
e
x
plo
ra
t
io
n
o
f
AI
f
o
r
bu
ild
ing
a
n
ASR
s
y
s
t
em
ha
v
e
perf
o
rma
nce
im
p
ro
v
em
ent
o
v
er
t
he
t
ra
ditio
na
l pro
ba
bil
is
t
ic
a
pp
ro
a
ches?
T
h
e
tr
ad
itio
n
al
p
r
o
b
ab
ilis
tic
a
p
p
r
o
ac
h
es
in
v
o
l
v
in
g
th
e
HM
M,
GM
M,
alo
n
g
with
o
th
er
alg
o
r
ith
m
s
lik
e
th
e
f
o
r
war
d
al
g
o
r
ith
m
s
,
f
o
r
war
d
an
d
r
e
v
er
s
e
alg
o
r
ith
m
,
Viter
b
i
alg
o
r
ith
m
c
o
m
e
with
th
eir
b
en
ef
its
.
Ho
wev
er
,
s
in
ce
th
e
e
x
p
lo
r
atio
n
o
f
AI
,
e
n
h
an
ce
p
er
f
o
r
m
a
n
ce
h
av
e
b
ee
n
ac
h
iev
ed
.
AI
ap
p
r
o
ac
h
in
v
o
lv
es
tr
ain
in
g
with
lar
g
e
co
r
p
u
s
o
f
s
p
ee
ch
d
ata
an
d
r
o
b
u
s
t
s
p
ee
ch
r
ec
o
g
n
itio
n
m
o
d
el
h
av
e
b
ee
n
d
ev
elo
p
ed
.
T
h
e
q
u
esti
o
n
p
r
o
v
id
es a
n
s
wer
s
to
if
th
e
p
u
r
p
o
s
e
o
f
u
tili
z
in
g
AI
i
n
p
er
f
o
r
m
an
ce
en
h
an
ce
m
en
t in
ter
m
s
o
f
W
E
R
an
d
W
R
R
h
av
e
b
ee
n
ac
h
ie
v
ed
.
2
.
1
.
5
.
H
a
s
t
he
intr
o
du
ct
io
n
o
f
AI
in
t
he
dev
elo
p
m
ent
a
nd
m
o
dellin
g
o
f
ASR
been
a
ble
t
o
a
chiev
e
t
he
ultim
a
t
e
g
o
a
l o
f
ASR
o
f
perf
o
rm
ing
a
s
a
hu
ma
n lis
t
ener
?
R
esea
r
c
h
e
r
s
h
av
e
d
e
v
e
lo
p
ed
v
ar
i
o
u
s
s
p
e
ec
h
r
e
c
o
g
n
i
ti
o
n
s
y
s
t
em
to
r
ec
o
g
n
i
z
e
h
u
m
a
n
s
p
ee
c
h
s
o
th
at
th
e
o
u
t
p
u
t
t
e
x
t
wil
l
lo
o
k
li
k
e
a
h
u
m
a
n
t
r
an
s
cr
ib
e
r
’
s
o
u
tp
u
t
.
Va
r
i
o
u
s
ASR
s
h
a
v
e
b
e
en
d
e
v
e
lo
p
e
d
t
o
ac
h
i
ev
e
t
h
is
,
h
o
we
v
e
r
,
t
h
e
q
u
esti
o
n
t
h
at
ca
l
l
s
f
o
r
c
o
n
c
er
n
is
t
o
k
n
o
w
if
t
h
e
u
lt
im
ate
g
o
al
o
f
ASR
t
o
i
n
t
er
a
ct
l
ik
e
h
u
m
an
b
ei
n
g
h
as
b
e
en
ac
h
iev
ed
.
W
h
is
p
e
r
f
o
r
in
s
ta
n
c
e,
w
as
a
b
le
t
o
f
u
n
cti
o
n
a
b
o
v
e
h
u
m
a
n
tr
a
n
s
c
r
i
b
e
r
i
n
t
er
m
s
o
f
t
h
e
o
u
t
p
u
t
g
e
n
e
r
a
te
d
te
x
t
,
b
u
t
t
h
e
r
es
u
lt
is
n
o
t
t
h
e
i
d
e
al
ASR
r
es
u
l
t
th
at
f
u
l
f
ils
t
h
e
u
lti
m
a
te
g
o
al
o
f
a
n
A
SR
s
y
s
t
em
.
2
.
2
.
St
a
rt
s
et
T
h
er
e
wer
e
5
,
5
,
6
,
1
,
an
d
1
p
ap
er
s
elec
ted
,
w
h
ich
ad
d
r
ess
ed
r
esear
ch
q
u
esti
o
n
s
1
,
2
,
3
,
4
,
an
d
5
r
esp
ec
tiv
ely
.
Fo
r
r
esear
ch
q
u
esti
o
n
s
1
,
2
,
4
,
an
d
5
;
p
ap
er
s
wer
e
s
o
u
r
ce
d
f
r
o
m
th
e
Go
o
g
le
Sch
o
lar
d
atab
ase
wh
ile
th
o
s
e
o
f
r
esear
ch
q
u
esti
o
n
3
wer
e
o
b
tain
ed
f
r
o
m
t
h
e
Sco
p
u
s
d
atab
ase.
A
to
tal
o
f
1
8
p
ap
er
s
wer
e
s
elec
ted
.
T
h
e
p
ap
er
s
ar
e
illu
s
tr
ated
u
s
in
g
th
e
letter
“Q”
+a
n
u
m
b
er
+le
tter
“P”+
a
n
u
m
b
er
wh
er
e
“Q”
+a
n
u
m
b
er
r
ep
r
esen
ts
th
e
q
u
esti
o
n
n
u
m
b
e
r
b
ein
g
ad
d
r
ess
ed
,
wh
ile
“P”+
a
n
u
m
b
er
r
e
p
r
esen
ts
p
ap
e
r
n
u
m
b
er
in
r
elatio
n
t
o
th
e
r
esear
ch
q
u
esti
o
n
n
u
m
b
e
r
b
ein
g
ad
d
r
ess
ed
.
T
h
e
s
elec
tio
n
cr
iter
ia
f
o
r
s
elec
tin
g
p
ap
e
r
s
f
r
o
m
th
e
Go
o
g
le
Sch
o
lar
f
o
r
q
u
esti
o
n
s
1
,
2
,
4
,
an
d
5
wer
e
th
e
r
elatio
n
o
f
th
e
p
ap
er
v
ia
th
o
r
o
u
g
h
r
ea
d
in
g
o
f
th
e
ab
s
tr
ac
t,
th
e
titl
e
o
f
th
e
p
ap
er
,
th
e
n
u
m
b
e
r
o
f
citatio
n
s
,
an
d
y
ea
r
o
f
p
u
b
licatio
n
m
o
s
tly
with
in
1
0
y
ea
r
s
.
T
h
e
n
u
m
b
er
o
f
citatio
n
s
o
f
Q1
P1
,
Q1
P2
,
Q1
P3
,
Q1
P4
,
Q1
P5
,
Q2
P1
,
Q2
P2
,
Q2
P3
,
Q2
P4
,
Q2
P5
,
Q4
P1
,
an
d
Q5
P1
ar
e
9
2
4
,
1
5
4
,
4
2
,
8
,
2
1
0
,
1
0
2
,
7
8
,
1
1
6
,
5
0
0
,
2
1
1
,
4
9
,
a
n
d
1
1
1
r
es
p
ec
tiv
ely
o
n
Go
o
g
le
Sch
o
lar
as
at
th
e
d
ate
o
f
d
o
cu
m
e
n
tatio
n
(
Sep
tem
b
er
12
,
2
0
2
3
)
.
W
e
d
id
n
o
t
s
ea
r
ch
f
o
r
p
ap
er
s
p
u
b
lis
h
ed
s
o
m
e
2
0
-
3
0
y
ea
r
s
ag
o
f
o
r
s
o
m
e
r
esear
ch
q
u
esti
o
n
s
lik
e
Q1
,
a
n
d
Q5
b
ec
a
u
s
e
s
o
m
e
ea
r
lier
r
ev
iew
p
ap
er
s
u
s
ed
i
n
th
is
c
u
r
r
en
t
s
tu
d
y
alr
ea
d
y
h
av
e
an
s
wer
s
to
s
o
m
e
o
f
th
e
q
u
esti
o
n
s
n
ee
d
ed
.
On
t
h
e
o
th
er
h
an
d
,
Q3
p
ap
er
s
wer
e
s
o
u
r
ce
d
f
r
o
m
Sco
p
u
s
in
o
r
d
er
t
o
h
av
e
a
m
ix
tu
r
e
o
f
d
if
f
er
en
t d
atab
ases
f
o
r
p
ap
er
s
elec
tio
n
s
in
th
is
wo
r
k
.
T
h
e
t
h
o
r
o
u
g
h
s
e
a
r
c
h
i
n
g
c
o
m
m
e
n
c
e
d
b
y
e
n
t
e
r
i
n
g
t
h
e
s
e
a
r
c
h
w
o
r
d
s
w
i
t
h
t
h
e
p
l
u
s
o
p
e
r
a
t
o
r
a
s
:
A
u
t
o
m
a
t
i
c
+
s
p
e
e
c
h
+
r
e
c
o
g
n
i
t
i
o
n
.
T
h
e
n
,
s
o
m
e
s
e
t
o
f
k
e
y
w
o
r
d
s
a
s
s
h
o
w
n
i
n
T
a
b
l
e
1
w
e
r
e
s
p
e
c
i
f
i
e
d
t
o
f
i
l
t
e
r
t
h
e
s
e
a
r
c
h
e
d
r
e
s
u
l
t
s
.
T
h
e
r
e
s
u
l
t
s
o
f
t
h
i
s
,
r
e
d
u
c
e
d
t
h
e
n
u
m
b
e
r
o
f
l
i
s
t
e
d
p
a
p
e
r
s
t
o
2
1
,
0
8
3
.
T
h
e
s
u
b
j
e
c
t
a
r
e
a
a
n
d
d
o
c
u
m
e
n
t
t
y
p
e
w
e
r
e
a
l
s
o
s
p
e
c
i
f
i
e
d
a
s
p
r
o
v
i
d
e
d
i
n
T
a
b
l
e
2
t
o
f
i
l
t
e
r
t
h
e
r
e
s
u
l
t
s
t
o
1
8
,
1
6
7
d
o
c
u
m
e
n
t
s
.
N
e
x
t
i
s
t
h
e
r
e
f
i
n
e
d
s
e
a
r
c
h
w
i
t
h
t
h
e
s
e
a
r
c
h
w
o
r
d
a
u
t
o
m
a
t
i
c
s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
+
a
r
t
i
f
i
c
i
a
l
i
n
t
e
l
l
i
g
e
n
c
e
w
h
i
c
h
r
e
d
u
c
e
d
t
h
e
n
u
m
b
e
r
t
o
7
,
0
2
0
,
w
h
i
l
e
t
h
e
l
a
n
g
u
a
g
e
f
a
c
t
o
r
w
h
i
c
h
i
n
v
o
l
v
e
s
l
i
m
i
t
i
n
g
t
h
e
s
e
a
r
c
h
p
a
p
e
r
s
t
o
E
n
g
l
i
s
h
l
a
n
g
u
a
g
e
g
a
v
e
6
,
9
4
8
p
a
p
e
r
s
.
T
h
i
s
w
a
s
f
o
l
l
o
w
e
d
b
y
u
s
i
n
g
t
h
e
r
e
f
i
n
e
d
s
e
a
r
c
h
w
i
t
h
t
h
e
k
e
y
w
o
r
d
“
b
e
n
e
f
i
t
”
w
h
i
c
h
g
a
v
e
a
v
e
r
y
a
n
t
i
c
i
p
a
t
e
d
r
e
d
u
c
e
d
v
a
l
u
e
o
f
3
5
9
p
a
p
e
r
s
a
n
d
l
a
t
e
r
2
0
4
w
h
e
n
t
h
e
r
a
n
g
e
w
a
s
s
e
t
b
e
t
w
e
e
n
2
0
1
0
a
n
d
2
0
2
0
s
o
a
s
t
o
h
a
v
e
t
h
a
t
s
m
a
l
l
n
u
m
b
e
r
o
f
2
0
4
f
o
r
s
e
l
e
c
t
i
n
g
p
a
p
e
r
s
f
r
o
m
a
n
d
t
o
h
a
v
e
p
a
p
e
r
s
t
h
a
t
h
a
v
e
t
h
e
l
a
t
e
s
t
a
n
d
u
p
d
a
t
e
d
t
e
c
h
n
o
l
o
g
y
a
s
w
e
l
l
a
s
l
a
t
e
s
t
r
e
s
e
a
r
c
h
f
i
n
d
i
n
g
s
i
n
A
S
R
.
T
h
e
f
i
r
s
t
p
a
p
e
r
,
Q
3
P
1
w
i
t
h
8
c
i
t
a
t
i
o
n
s
i
n
S
c
o
p
u
s
a
n
d
1
5
c
i
t
a
t
i
o
n
s
i
n
G
o
o
g
l
e
S
c
h
o
l
a
r
w
a
s
s
e
l
e
c
t
e
d
a
f
t
e
r
r
e
a
d
i
n
g
t
h
e
a
b
s
t
r
a
c
t
t
o
e
n
s
u
r
e
i
t
c
o
n
t
a
i
n
e
d
t
h
e
n
e
e
d
e
d
i
n
f
o
r
m
a
t
i
o
n
.
O
t
h
e
r
p
a
p
e
r
s
s
u
c
h
a
s
t
h
e
Q
3
P
2
,
Q
3
P
3
,
a
n
d
Q
3
P
4
w
e
r
e
a
l
s
o
s
e
l
e
c
t
e
d
a
s
d
e
t
a
i
l
e
d
i
n
T
a
b
l
e
3
.
F
r
o
m
t
h
e
s
e
l
e
c
t
i
o
n
s
t
a
t
e
o
f
Q
3
P
3
,
u
s
i
n
g
k
e
y
w
o
r
d
s
s
u
c
h
a
s
d
e
e
p
l
e
a
r
n
i
n
g
,
f
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
,
DNN
,
n
e
u
r
a
l
n
e
t
w
o
r
k
,
R
N
N
,
C
N
N
,
S
V
M
,
B
a
y
e
s
i
a
n
n
e
t
w
o
r
k
s
,
A
N
N
,
t
e
m
p
o
r
a
l
c
l
a
s
s
i
f
i
c
a
t
i
o
n
,
m
u
l
t
i
l
a
y
e
r
n
e
u
r
a
l
n
e
t
w
o
r
k
s
,
t
r
a
n
s
f
e
r
l
e
a
r
n
i
n
g
,
L
S
T
M
,
a
n
d
p
a
t
t
e
r
n
r
e
c
o
g
n
i
t
i
o
n
b
r
o
u
g
h
t
t
h
e
n
u
m
b
e
r
o
f
p
a
p
e
r
s
d
o
w
n
t
o
1
8
8
d
o
c
u
m
e
n
t
s
,
a
m
o
n
g
w
h
i
c
h
Q
3
P
5
w
a
s
s
e
l
e
c
t
e
d
.
F
i
n
a
l
l
y
,
f
r
o
m
t
h
e
s
e
l
e
ct
i
o
n
s
t
a
te
o
f
Q
3
P
5
,
t
h
e
u
s
e
o
f
t
h
e
r
e
f
i
n
e
d
k
e
y
w
o
r
d
m
a
c
h
i
n
e
le
a
r
n
i
n
g
l
e
d
t
o
t
h
e
s
e
l
e
ct
i
o
n
o
f
Q
3
P
6
.
A
d
e
t
a
i
l
e
d
a
n
a
l
y
s
is
o
f
t
h
e
r
e
d
u
c
t
i
o
n
p
r
o
c
e
s
s
l
e
a
d
i
n
g
t
o
t
h
e
e
v
e
n
t
u
a
l
s
e
l
e
ct
io
n
o
f
p
a
p
e
r
s
Q
3
P
1
,
Q
3
P
2
,
Q
3
P
3
,
a
n
d
Q
3
P
4
a
r
e
g
i
v
e
n
i
n
T
a
b
l
e
s
1
t
o
3
w
h
e
r
e
t
h
e
n
u
m
b
e
r
o
f
p
a
p
e
r
s
l
e
f
t
a
t
e
a
c
h
s
t
a
g
e
o
f
t
h
e
f
i
l
t
e
r
i
n
g
a
r
e
i
n
d
i
c
a
t
e
d
w
i
t
h
a
n
a
r
r
o
w
p
o
in
t
i
n
g
r
i
g
h
t
.
T
h
e
s
el
e
c
t
i
o
n
p
r
o
c
es
s
es
o
f
Q
3
p
a
p
e
r
s
a
r
e
i
ll
u
s
t
r
at
ed
i
n
F
i
g
u
r
e
2
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
6
:
20
-
43
24
T
ab
le
1
.
Sear
ch
k
ey
wo
r
d
s
f
o
r
Q3
in
Sco
p
u
s
(
≥
2
1
,
0
8
3
d
o
c
u
m
en
ts
)
A
c
o
u
st
i
c
m
o
d
e
l
P
e
r
f
o
r
ma
n
c
e
A
c
o
u
st
i
c
m
o
d
e
l
l
i
n
g
R
e
v
e
r
b
e
r
a
t
i
o
n
A
r
t
i
f
i
c
i
a
l
i
n
t
e
l
l
i
g
e
n
c
e
R
o
b
u
st
s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
A
S
R
a
u
t
o
ma
t
i
c
s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
S
p
e
e
c
h
e
n
h
a
n
c
e
me
n
t
A
u
d
i
o
si
g
n
a
l
p
r
o
c
e
ssi
n
g
S
p
e
a
k
e
r
r
e
c
o
g
n
i
t
i
o
n
A
u
t
o
ma
t
i
c
r
e
c
o
g
n
i
t
i
o
n
S
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
A
u
t
o
ma
t
i
c
s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
S
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
so
f
t
w
a
r
e
A
u
t
o
ma
t
i
c
s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
(
A
S
R
)
S
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
sy
s
t
e
ms
A
u
t
o
ma
t
i
c
s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
sy
s
t
e
m
S
p
e
e
c
h
r
e
c
o
g
n
i
z
e
r
C
o
n
t
i
n
u
o
u
s s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
S
p
e
e
c
h
si
g
n
a
l
s
La
n
g
u
a
g
e
mo
d
e
l
W
ER
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
T
ab
le
2
.
Fil
ter
in
g
o
f
Q3
p
ap
er
s
b
y
s
u
b
ject
ar
ea
an
d
d
o
cu
m
en
t ty
p
e
in
Sco
p
u
s
(
≥
1
8
,
1
6
7
d
o
c
u
m
en
ts
)
S
u
b
j
e
c
t
a
r
e
a
D
o
c
u
me
n
t
t
y
p
e
C
o
m
p
u
t
e
r
sc
i
e
n
c
e
C
o
n
f
e
r
e
n
c
e
p
a
p
e
r
En
g
i
n
e
e
r
i
n
g
A
r
t
i
c
l
e
R
e
v
i
e
w
C
o
n
f
e
r
e
n
c
e
r
e
v
i
e
w
T
ab
le
3
.
Sear
ch
i
n
g
with
r
e
f
in
e
d
k
ey
wo
r
d
s
f
o
r
Q3
P1
,
Q3
P2
,
Q3
P3
,
an
d
Q3
P4
af
ter
th
e
s
tep
s
u
s
ed
to
o
b
tain
T
ab
les 1
,
2
,
a
n
d
3
in
Sco
p
u
s
R
e
f
i
n
e
d
k
e
y
w
o
r
d
s
≥
y
e
a
r
r
a
n
g
e
R
e
d
u
c
t
i
o
n
i
n
se
a
r
c
h
r
e
s
u
l
t
s
R
e
d
u
c
t
i
o
n
i
n
se
a
r
c
h
r
e
s
u
l
t
s
R
e
d
u
c
t
i
o
n
i
n
se
a
r
c
h
r
e
s
u
l
t
s
B
e
n
e
f
i
t
≥
2
0
1
0
-
2
0
2
0
3
5
9
≥
204
N
I
LL
N
I
LL
A
d
v
a
n
t
a
g
e
≥
2
0
1
0
-
2
0
2
0
N
I
LL
3
6
7
≥
185
3
6
7
N
u
mb
e
r
o
f
p
a
p
e
r
s se
l
e
c
t
e
d
Q
3
P
1
,
Q
3
P
2
Q
3
P
3
Q
3
P
4
Fig
u
r
e
2
.
Flo
wch
ar
t
o
f
s
elec
tio
n
o
f
p
ap
er
s
f
o
r
q
u
esti
o
n
3
T
h
e
1
8
p
a
p
er
s
in
itially
s
elec
t
ed
ar
e
lis
ted
in
T
ab
le
4
.
Af
ter
in
clu
s
io
n
/ex
clu
s
io
n
cr
iter
ia
s
u
ch
as
th
e
r
elev
an
ce
o
f
th
e
p
ap
er
s
to
th
e
p
ar
ticu
lar
q
u
esti
o
n
b
ei
n
g
ad
d
r
ess
ed
an
d
,
th
e
r
elev
an
ce
o
f
t
h
e
co
n
ten
ts
to
th
e
q
u
esti
o
n
,
h
e
n
ce
,
th
e
p
ap
er
s
ev
en
tu
ally
s
elec
ted
wer
e
Q1
P1
,
Q1
P2
,
Q1
P4
,
Q1
P5
,
Q
2
P1
,
Q
2
P2
,
Q2
P3
,
Q2
P4
,
Q2
P5
,
Q4
P1
,
Q5
P1
,
Q3
P3
,
an
d
Q3
P6
.
Fro
m
th
e
o
r
i
g
i
n
a
l
1
8
p
ap
er
s
s
elec
ted
,
p
ap
er
s
Q1
P3
,
Q3
P1
,
Q3
P2
,
Q3
P4
,
an
d
Q3
P5
wer
e
e
v
en
t
u
ally
d
is
ca
r
d
ed
.
T
h
ese
s
ets
o
f
p
ap
er
s
wer
e
r
ejec
ted
d
u
e
t
o
th
eir
in
ab
ilit
y
to
p
r
o
p
er
l
y
an
s
wer
th
e
q
u
esti
o
n
s
b
ein
g
ad
d
r
ess
ed
,
o
r
th
ey
co
n
tain
ed
r
ep
etitio
n
s
o
f
wh
at
h
ad
b
ee
n
g
o
tten
f
r
o
m
th
e
p
r
e
v
io
u
s
jo
u
r
n
als.
Fo
r
in
s
tan
ce
,
p
ap
er
Q1
P3
was
a
r
ev
iew
p
ap
er
th
at
c
o
n
tain
ed
t
h
e
m
ajo
r
ity
o
f
th
e
in
f
o
r
m
atio
n
th
at
we’
v
e
ac
q
u
ir
ed
f
r
o
m
th
e
p
r
ev
io
u
s
p
ap
er
s
,
wh
ich
ev
en
tu
ally
led
to
its
r
ej
ec
tio
n
.
Pap
er
Q3
P2
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
E
vo
lu
tio
n
a
r
y
tr
en
d
s
in
a
u
to
m
a
tic
s
p
ee
ch
r
ec
o
g
n
itio
n
w
ith
…
(
Ga
b
r
iel
Olu
w
a
to
b
i
S
o
b
o
l
a
)
25
was
also
r
ejec
ted
b
ec
au
s
e
its
ab
s
tr
ac
t
d
ev
iated
f
r
o
m
th
e
q
u
esti
o
n
s
o
f
th
e
r
esear
ch
wo
r
k
b
ein
g
ca
r
r
ied
o
u
t.
Hen
ce
a
to
tal
o
f
1
3
p
a
p
er
s
w
er
e
s
elec
ted
o
u
t
o
f
t
h
e
o
r
i
g
in
al
1
8
p
ap
er
s
s
o
u
r
ce
d
f
r
o
m
Go
o
g
le
Sch
o
lar
s
an
d
Sco
p
u
s
as
th
e
s
tar
t
s
et
o
f
p
ap
er
s
f
o
r
th
e
f
ir
s
t
iter
atio
n
s
t
ag
e.
T
h
e
s
elec
ted
p
ap
er
s
ar
e
lis
ted
in
T
ab
le
5
.
T
h
e
wh
o
le
s
elec
tio
n
p
r
o
ce
s
s
es o
f
th
e
1
3
s
tar
t set p
ap
er
s
ar
e
il
lu
s
tr
ated
in
Alg
o
r
ith
m
1
.
Alg
o
r
ith
m
1
.
Selectio
n
p
r
o
ce
s
s
f
o
r
1
3
in
itial p
ap
er
s
s
elec
ted
f
r
o
m
1
8
in
itial p
ap
e
r
s
s
elec
ted
1.
B
E
GI
N
2.
I
n
itialize
th
e
o
r
ig
in
al
a
r
r
ay
as
em
p
ty
lis
t
3.
Scan
/
r
ea
d
th
r
o
u
g
h
th
e
ab
s
tr
ac
t
/ title
o
f
p
ap
er
1
4.
I
ter
ate
th
r
o
u
g
h
p
ap
er
s
1
t
o
1
8
:
5.
I
f
f
ir
s
t p
ap
e
r
co
n
tain
s
r
ele
v
an
t
in
f
o
r
m
atio
n
:
6.
Ap
p
en
d
it to
th
e
ar
r
a
y
7.
E
ls
e
r
ejec
t
th
e
p
ap
er
8.
I
f
s
ec
o
n
d
p
ap
er
co
n
tain
s
r
elev
an
t in
f
o
r
m
atio
n
:
9.
Ap
p
en
d
it to
th
e
ar
r
a
y
10.
E
ls
e
R
ejec
t th
e
p
ap
er
11.
R
ep
ea
t step
s
4
to
5
f
o
r
p
ap
e
r
3
th
r
o
u
g
h
1
8
12.
Prin
t th
e
ar
r
ay
c
o
n
tain
in
g
all
t
h
e
s
elec
ted
p
ap
er
s
13.
E
ND
T
ab
le
4
.
T
h
e
1
8
in
itial set o
f
p
ap
er
s
s
elec
ted
f
r
o
m
Sco
p
u
s
an
d
Go
o
g
le
Sch
o
lar
R
e
f
.
n
o.
P
a
p
e
r
i
d
e
n
t
i
t
y
c
o
d
e
R
e
f
.
n
o.
P
a
p
e
r
i
d
e
n
t
i
t
y
c
o
d
e
[
1
]
Q
1
P
1
[
2
5
]
Q
2
P
5
[
2
]
Q
1
P
2
[
2
6
]
Q
3
P
1
[
2
7
]
Q
1
P
3
[
2
8
]
Q
3
P
2
[
3
]
Q
1
P
4
[
2
9
]
Q
3
P
3
[
4
]
Q
1
P
5
[
1
6
]
Q
3
P
4
[
3
0
]
Q
2
P
1
[
3
1
]
Q
3
P
5
[
3
2
]
Q
2
P
2
[
3
3
]
Q
3
P
6
[
1
7
]
Q
2
P
3
[
5
]
Q
4
P
1
[
3
4
]
Q
2
P
4
[
3
5
]
Q
5
P
1
T
ab
le
5
.
T
h
e
1
3
s
elec
ted
s
tar
s
et
p
ap
er
s
f
o
r
th
e
f
ir
s
t iter
atio
n
R
e
f
.
n
o.
P
a
p
e
r
i
d
e
n
t
i
t
y
c
o
d
e
R
e
f
.
n
o.
P
a
p
e
r
i
d
e
n
t
i
t
y
c
o
d
e
[
1
]
Q
1
P
1
[
3
4
]
Q
2
P
4
[
2
]
Q
1
P
2
[
2
5
]
Q
2
P
5
[
3
]
Q
1
P
4
[
2
9
]
Q
3
P
3
[
4
]
Q
1
P
5
[
3
3
]
Q
3
P
6
[
3
0
]
Q
2
P
1
[
5
]
Q
4
P
1
[
3
2
]
Q
2
P
2
[
3
5
]
Q
5
P
1
[
1
7
]
Q
2
P
3
2
.
3
.
I
t
er
a
t
io
n
2
.
3
.
1
.
I
t
er
a
t
io
n 1
T
h
e
1
3
s
et
o
f
p
ap
er
s
(
s
tar
t
s
et
)
s
elec
ted
f
r
o
m
th
e
1
8
in
itially
s
et
o
f
p
ap
er
s
s
o
u
r
ce
d
f
r
o
m
t
h
e
Go
o
g
le
Sch
o
lar
an
d
Sco
p
u
s
wer
e
r
u
n
th
r
o
u
g
h
th
e
f
o
r
war
d
an
d
b
ac
k
war
d
s
n
o
wb
allin
g
ap
p
r
o
ac
h
es
u
n
til
n
o
p
ap
er
was
f
o
u
n
d
to
b
e
in
clu
d
ed
in
th
e
lis
t
o
f
s
elec
ted
p
ap
er
s
.
T
h
e
s
tar
ti
n
g
s
et
o
f
p
ap
er
s
co
n
s
is
ted
o
f
1
3
p
ap
er
s
(
T
ab
le
5
)
s
elec
ted
f
r
o
m
th
e
1
8
in
itially
s
et
o
f
p
ap
er
s
(
T
ab
le
4
)
s
o
u
r
ce
d
f
r
o
m
Go
o
g
le
Sch
o
lar
an
d
Sco
p
u
s
.
T
h
e
s
elec
tio
n
o
f
th
ese
1
3
p
ap
e
r
s
was
b
ased
o
n
s
o
m
e
e
x
clu
s
io
n
an
d
in
clu
s
i
o
n
cr
iter
ia
b
ased
o
n
th
e
r
ele
v
a
n
ce
o
f
i
n
f
o
r
m
atio
n
in
r
elatio
n
to
th
e
r
esear
ch
q
u
esti
o
n
s
b
ein
g
ad
d
r
ess
ed
.
T
h
ese
p
ap
er
s
wer
e
s
elec
ted
tak
in
g
in
t
o
co
n
s
id
er
atio
n
th
e
y
ea
r
s
o
f
p
u
b
licatio
n
s
,
au
th
o
r
s
,
an
d
p
u
b
lis
h
er
s
with
a
k
ey
f
o
cu
s
o
n
th
eir
r
elev
a
n
ce
.
Alg
o
r
ith
m
1
s
h
o
ws
th
e
s
elec
tio
n
p
r
o
ce
s
s
o
f
th
ese
1
3
s
tar
ts
s
et
p
ap
er
s
.
2
.
3
.
2
.
B
a
ck
wa
rd
s
no
wba
llin
g
Fro
m
th
e
f
ir
s
t
p
ap
er
,
Q1
P1
,
ca
n
d
id
ates
wer
e
s
elec
ted
f
o
r
in
clu
s
io
n
u
s
in
g
b
ac
k
war
d
s
n
o
wb
allin
g
,
4
p
ap
er
s
wer
e
s
elec
ted
wh
ich
wer
e
ad
d
e
d
to
th
e
s
tar
t
s
et.
Fro
m
p
a
p
er
Q
1
P2
b
ac
k
war
d
s
n
o
wb
allin
g
ap
p
r
o
ac
h
p
r
o
d
u
ce
s
6
p
ap
er
s
wh
ich
wer
e
ad
d
e
d
to
th
e
s
tar
t
s
et
p
ap
er
s
.
Fro
m
p
a
p
er
Q
1
P4
,
2
p
a
p
er
s
w
er
e
s
elec
ted
v
ia
th
is
ap
p
r
o
ac
h
.
Q
1
P5
p
r
o
d
u
ce
s
6
p
ap
er
s
.
Q2
P1
p
r
o
d
u
ce
s
2
p
a
p
er
s
,
Q2
P2
p
r
o
d
u
ce
s
1
p
ap
e
r
.
Q2
P3
p
r
o
d
u
ce
s
3
p
ap
er
s
.
Pap
er
Q2
P4
p
r
o
d
u
c
es
3
p
ap
er
s
.
Pap
er
Q2
P5
p
r
o
d
u
ce
s
3
p
ap
er
s
.
Q4
P1
p
r
o
d
u
c
es
6
p
ap
er
s
,
Q5
P1
p
r
o
d
u
ce
s
7
p
ap
e
r
s
,
Q3
P3
g
en
er
ates
3
p
ap
er
s
,
wh
ile
th
e
last
p
ap
e
r
(
Q
3
P6
)
p
r
o
d
u
ce
s
2
p
a
p
er
s
.
T
h
en
a
to
tal
o
f
4
8
p
ap
er
s
wer
e
s
elec
ted
f
r
o
m
t
h
e
b
ac
k
war
d
s
n
o
wb
allin
g
ap
p
r
o
ac
h
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
6
:
20
-
43
26
2
.
3
.
3
.
F
o
rwa
rd
s
no
wba
llin
g
F
o
r
t
h
e
s
e
t
o
f
1
3
s
t
a
r
t
s
et
p
a
p
e
r
s
s
e
le
c
t
e
d
,
t
h
r
o
u
g
h
t
h
e
f
o
r
w
a
r
d
s
n
o
w
b
a
l
l
i
n
g
a
p
p
r
o
a
c
h
,
t
h
e
an
a
l
y
s
is
o
f
t
h
e
p
a
p
e
r
s
s
e
l
ec
t
e
d
u
s
i
n
g
t
h
e
i
n
c
l
u
s
i
o
n
/
e
x
cl
u
s
i
o
n
c
r
i
t
e
r
ia
a
r
e
d
e
t
a
i
l
e
d
t
h
u
s
:
p
a
p
e
r
Q
1
P
1
p
r
o
d
u
c
e
s
1
1
p
a
p
e
r
s
a
f
t
e
r
u
s
i
n
g
s
o
m
e
c
r
i
t
e
r
i
a
s
u
c
h
as
t
h
e
a
b
s
t
r
a
c
ts
o
f
t
h
e
p
a
p
e
r
s
,
t
h
e
t
i
t
l
e
s
,
a
n
d
t
h
e
n
u
m
b
e
r
o
f
c
i
t
a
t
i
o
n
s
o
f
t
h
e
p
a
p
e
r
s
.
T
h
e
y
a
r
e
l
i
s
te
d
a
s
Q
1
P
1
F
1
t
h
r
o
u
g
h
Q
1
P
1
F
1
2
.
P
a
p
e
r
Q
1
P
2
p
r
o
d
u
c
e
s
9
p
a
p
e
r
s
.
S
o
m
e
c
o
n
d
i
t
i
o
n
s
f
o
r
p
a
p
e
r
s
e
l
e
c
t
i
o
n
w
e
r
e
:
i
f
t
h
e
p
a
p
e
r
s
h
a
d
b
e
e
n
i
n
c
l
u
d
e
d
a
s
p
a
r
t
o
f
t
h
e
p
a
p
e
r
s
i
n
t
h
e
o
t
h
e
r
s
e
l
e
ct
i
o
n
s
,
i
t
w
a
s
r
e
jec
t
e
d
,
t
h
e
a
b
s
t
r
a
ct
o
f
t
h
e
p
a
p
e
r
,
a
n
d
t
h
e
r
e
l
e
v
a
n
c
e
o
f
t
h
e
p
a
p
e
r
s
t
o
t
h
e
q
u
e
s
t
i
o
n
b
e
i
n
g
c
o
n
s
i
d
e
r
e
d
.
N
o
p
a
p
e
r
w
a
s
s
e
l
e
c
t
e
d
f
r
o
m
Q
1
P
4
.
I
t
h
a
d
j
u
s
t
1
0
c
i
t
a
t
i
o
n
s
a
n
d
7
o
f
t
h
e
p
a
p
e
r
s
d
o
n
o
t
s
h
o
w
r
e
l
a
t
i
o
n
s
h
i
p
t
o
t
h
e
q
u
e
s
t
i
o
n
s
b
e
i
n
g
a
d
d
r
e
s
s
e
d
.
T
h
e
r
e
m
a
i
n
i
n
g
3
a
r
e
n
o
t
i
n
c
l
u
d
e
d
d
u
e
t
o
i
n
a
b
i
l
i
t
y
t
o
a
c
c
e
s
s
t
h
e
m
at
e
r
i
a
ls
.
F
o
r
p
a
p
e
r
Q
1
P
5
a
b
o
u
t
1
0
p
a
p
e
r
s
w
e
r
e
s
e
e
n
a
s
p
o
t
e
n
t
ia
l
p
a
p
e
r
s
i
n
r
e
l
a
ti
o
n
to
t
h
e
q
u
e
s
t
i
o
n
s
b
e
i
n
g
a
d
d
r
e
s
s
ed
n
o
n
e
o
f
w
h
i
c
h
c
o
n
t
a
i
n
e
d
s
o
m
e
t
h
i
n
g
u
n
i
q
u
e
f
r
o
m
t
h
e
i
n
f
o
r
m
a
t
i
o
n
i
n
i
t
i
al
l
y
g
a
t
h
er
e
d
f
r
o
m
t
h
e
p
r
e
v
i
o
u
s
p
a
p
e
r
s
.
B
e
s
i
d
e
s
,
s
o
m
e
o
f
t
h
e
c
o
n
t
e
n
ts
h
a
d
b
e
e
n
u
s
e
d
i
n
o
t
h
e
r
p
a
p
e
r
s
.
H
e
n
ce
,
n
o
p
a
p
e
r
w
a
s
s
e
l
ec
t
e
d
.
P
a
p
e
r
Q
2
P
1
w
it
h
1
0
6
c
i
t
a
t
i
o
n
s
h
a
d
1
p
a
p
e
r
f
o
r
in
c
l
u
s
i
o
n
b
u
t
it
w
as
r
e
j
e
c
t
e
d
b
e
ca
u
s
e
i
t
h
as
b
e
e
n
e
a
r
l
i
e
r
s
e
l
ec
t
e
d
,
h
e
n
c
e
n
o
p
a
p
e
r
w
a
s
s
el
e
c
te
d
.
O
t
h
e
r
p
a
p
e
r
s
a
r
e
e
it
h
e
r
r
e
p
e
a
t
ed
p
a
p
e
r
s
o
r
n
o
t
s
h
a
r
i
n
g
o
r
a
d
d
r
e
s
s
i
n
g
t
h
e
q
u
e
s
t
i
o
n
s
i
n
t
h
i
s
r
e
v
ie
w
.
F
o
r
p
a
p
e
r
s
Q
2
P
1
a
l
l
t
h
r
o
u
g
h
Q
5
P
1
,
n
o
p
a
p
e
r
w
a
s
s
e
l
e
ct
e
d
f
o
r
e
a
c
h
o
f
t
h
em
.
T
h
e
r
e
a
s
o
n
b
e
i
n
g
t
h
a
t
m
ajo
r
l
y
a
l
l
t
h
e
r
es
e
a
r
c
h
t
o
p
i
cs
s
h
a
r
e
s
i
m
il
a
r
i
ti
es
w
it
h
a
l
r
e
a
d
y
s
t
u
d
i
e
d
p
a
p
e
r
s
f
o
r
t
h
is
w
o
r
k
,
o
r
t
h
e
i
r
r
e
l
e
v
a
n
c
e
o
f
t
o
p
ic
o
r
a
b
s
t
r
a
c
t
t
o
t
h
e
q
u
e
s
t
i
o
n
s
b
ei
n
g
a
d
d
r
e
s
s
e
d
.
T
h
er
ef
o
r
e,
a
to
tal
o
f
2
0
p
a
p
er
s
wer
e
s
elec
ted
v
ia
th
e
f
o
r
wa
r
d
s
n
o
wb
allin
g
ap
p
r
o
ac
h
,
wh
i
ch
s
h
o
ws
th
at
a
to
tal
o
f
6
8
p
ap
er
s
h
as
b
ee
n
s
elec
ted
f
o
r
in
cl
u
s
io
n
in
th
is
wo
r
k
d
u
e
to
r
ele
v
an
ce
i
n
ad
d
r
ess
in
g
th
ese
q
u
esti
o
n
s
in
th
is
r
ev
iew.
Hen
ce
,
it
ca
n
b
e
co
n
clu
d
ed
th
a
t
f
r
o
m
th
e
f
ir
s
t
iter
atio
n
,
ca
n
d
id
ate
p
ap
er
s
wer
e
s
elec
ted
v
ia
th
e
b
ac
k
war
d
a
n
d
f
o
r
war
d
s
n
o
wb
allin
g
ap
p
r
o
ac
h
an
d
in
clu
d
ed
in
t
h
e
s
tar
t
s
et
s
elec
ted
lis
t
o
f
p
ap
er
s
m
ak
in
g
a
to
tal
o
f
1
3
(
s
tar
s
et
p
ap
er
s
)
+4
8
p
a
p
er
s
s
elec
ted
v
ia
b
ac
k
war
d
s
n
o
wb
allin
g
+2
0
p
a
p
er
s
s
elec
ted
v
ia
f
o
r
war
d
s
n
o
wb
allin
g
.
T
h
e
6
8
s
elec
ted
p
ap
er
s
ar
e
as sh
o
wn
in
T
ab
le
6
.
T
ab
le
6
.
I
ter
atio
n
1
s
elec
ted
li
s
t o
f
p
ap
er
s
(
6
8
)
R
e
f
.
no
.
P
a
p
e
r
i
d
e
n
t
i
t
y
c
o
d
e
R
e
f
.
n
o.
P
a
p
e
r
i
d
e
n
t
i
t
y
c
o
d
e
R
e
f
.
n
o.
P
a
p
e
r
i
d
e
n
t
i
t
y
c
o
d
e
R
e
f
.
n
o.
P
a
p
e
r
i
d
e
n
t
i
t
y
c
o
d
e
[
3
6
]
Q
1
P
1
B
1
[
3
7
]
Q
I
P
5
B
6
[
3
8
]
Q
4
P
1
B
5
[
1
0
]
Q
1
P
1
F
5
[
3
9
]
Q
1
P
1
B
4
[
4
0
]
Q
2
P
1
B
1
[
4
1
]
Q
4
P
1
B
6
[
4
2
]
Q
1
P
1
F
6
[
4
3
]
Q
1
P
1
B
5
[
2
2
]
Q
2
P
1
B
2
[
4
4
]
Q
5
P
1
B
1
[
1
4
]
Q
1
P
1
F
7
[
4
5
]
Q
1
P
1
B
6
[
4
6
]
Q
2
P
2
B
2
[
4
7
]
Q
5
P
1
B
2
[
4
8
]
Q
1
P
1
F
8
[
4
9
]
Q
1
P
2
B
1
[
5
0
]
Q
2
P
3
B
3
[
5
1
]
Q
5
P
1
B
3
[
1
5
]
Q
1
P
1
F
9
[
5
2
]
Q
1
P
2
B
3
[
5
3
]
Q
2
P
3
B
4
[
5
4
]
Q
5
P
1
B
4
[
8
]
Q
1
P
1
F
1
0
[
5
5
]
Q
1
P
2
B
4
[
5
6
]
Q
2
P
3
B
5
[
5
7
]
Q
5
P
1
B
5
[
5
8
]
Q
1
P
1
F
1
1
[
1
8
]
Q
1
P
2
B
5
[
5
9
]
Q
2
P
4
B
1
[
6
0
]
Q
5
P
1
B
6
[
6
1
]
Q
1
P
1
F
1
2
[
6
2
]
Q
1
P
2
B
6
[
6
3
]
Q
2
P
4
B
2
[
6
4
]
Q
5
P
1
B
7
[
9
]
Q
1
P
2
F
1
[
6
5
]
Q
1
P
2
B
7
[
6
6
]
Q
2
P
4
B
3
[
6
7
]
Q
3
P
3
B
1
[
2
1
]
Q
1
P
2
F
2
[
6
8
]
Q
1
P
4
B
1
[
1
9
]
Q
2
P
5
B
1
[
6
9
]
Q
3
P
3
B
2
[
7
0
]
Q
1
P
2
F
3
[
1
3
]
Q
1
P
4
B
3
[
7
1
]
Q
2
P
5
B
2
[
7
2
]
Q
3
P
3
B
3
[
7
3
]
Q
1
P
2
F
4
[
7
4
]
Q
I
P
5
B
1
[
7
5
]
Q
2
P
5
B
3
[
7
6
]
Q
3
P
6
B
1
[
2
3
]
Q
1
P
2
F
5
[
7
7
]
Q
I
P
5
B
2
[
7
8
]
Q
4
P
1
B
1
[
7
9
]
Q
3
P
6
B
2
[
8
0
]
Q
1
P
2
F
6
[
8
1
]
Q
I
P
5
B
3
[
8
2
]
Q
4
P
1
B
2
[
1
1
]
Q
1
P
1
F
2
[
8
3
]
Q
1
P
2
F
7
[
8
4
]
Q
I
P
5
B
4
[
8
5
]
Q
4
P
1
B
3
[
2
0
]
Q
1
P
1
F
3
[
8
6
]
Q
1
P
2
F
8
[
8
7
]
Q
I
P
5
B
5
[
8
8
]
Q
4
P
1
B
4
[
1
2
]
Q
1
P
1
F
4
[
8
9
]
Q
1
P
2
F
9
2
.
3
.
4
.
Su
m
m
a
r
y
o
f
it
er
a
t
i
o
n
1
Fro
m
th
e
13
-
s
tar
t
s
et
o
f
p
ap
er
s
u
s
ed
f
o
r
th
is
r
esear
ch
wo
r
k
,
a
to
tal
o
f
4
8
p
ap
er
s
h
a
v
e
b
ee
n
s
elec
ted
v
ia
th
e
b
ac
k
war
d
s
n
o
wb
allin
g
ap
p
r
o
ac
h
,
wh
er
e
p
ap
er
s
Q1
P1
,
Q1
P2
,
Q1
P4
,
Q1
P5
,
Q2
P1
,
Q
2
P2
,
Q2
P3
,
Q2
P4
,
Q2
P5
,
Q4
P1
,
Q5
P1
,
Q3
P3
,
an
d
Q3
P6
h
ad
2
3
2
,
1
7
9
,
1
0
,
4
3
,
9
,
2
4
,
2
6
,
5
1
,
7
5
,
8
6
,
7
5
,
5
1
,
an
d
1
3
5
r
e
f
er
en
ce
s
an
aly
z
ed
r
esp
ec
tiv
el
y
wh
ich
g
iv
es
a
to
tal
o
f
at
m
o
s
t
9
9
6
r
e
f
er
en
ce
s
f
o
r
th
e
4
8
p
a
p
er
s
in
cl
u
d
ed
in
th
e
s
tar
t
s
et
o
f
p
a
p
er
s
b
ased
o
n
th
e
i
n
clu
s
io
n
/ex
clu
s
io
n
cr
iter
ia.
Usi
n
g
th
e
f
o
r
war
d
s
n
o
w
b
allin
g
ap
p
r
o
ac
h
,
a
to
tal
o
f
2
0
p
a
p
er
s
wer
e
s
elec
ted
f
o
r
in
clu
s
io
n
in
th
e
s
tar
t
s
et
p
ap
e
r
s
.
Pap
er
s
Q1
P1
,
Q1
P2
,
Q1
P4
,
Q
1
P5
,
Q2
P1
,
Q2
P2
,
Q2
P3
,
Q2
P4
,
Q2
P5
,
Q4
P1
,
Q
5
P1
,
Q3
P3
,
an
d
Q3
P6
,
with
9
8
2
,
1
7
9
,
1
0
,
2
1
5
,
1
0
6
,
8
3
,
1
2
5
,
5
1
6
,
2
1
3
,
4
9
,
1
1
8
,
6
6
7
,
an
d
1
1
8
citatio
n
s
r
esp
ec
tiv
ely
s
h
o
wed
th
at
a
to
tal
o
f
at
m
o
s
t
3
,
3
8
1
r
ef
er
en
ce
s
w
er
e
an
aly
z
ed
u
s
in
g
ex
clu
s
io
n
/in
clu
s
io
n
cr
iter
ia
f
o
r
th
e
f
o
r
war
d
s
n
o
wb
allin
g
ap
p
r
o
ac
h
.
Hen
ce
,
a
to
tal
o
f
at
m
o
s
t
4
,
3
7
7
p
ap
er
s
wer
e
an
aly
z
e
d
a
n
d
a
to
tal
o
f
6
8
p
ap
er
s
h
ad
b
ee
n
in
clu
d
ed
in
th
e
s
tar
t
s
et
o
f
p
a
p
er
s
in
t
h
e
iter
atio
n
1
.
I
n
th
e
iter
atio
n
2
,
t
h
e
6
8
p
ap
e
r
s
wer
e
also
an
aly
z
ed
t
o
s
ee
wh
ich
p
ap
er
s
c
o
u
ld
b
e
ex
tr
ac
t
ed
f
r
o
m
ea
ch
v
ia
th
e
b
ac
k
war
d
a
n
d
f
o
r
war
d
s
n
o
wb
allin
g
ap
p
r
o
ac
h
.
2
.
3
.
5
.
I
t
er
a
t
io
n 2
I
t
s
h
o
u
ld
b
e
n
o
ted
th
at
at
th
e
s
tar
t
o
f
th
e
iter
atio
n
2
a
n
d
s
u
b
s
eq
u
en
t
o
n
es,
th
e
p
ap
er
s
ex
p
lo
r
ed
wer
e
lim
ited
to
2
0
2
2
an
d
2
0
2
3
.
T
h
i
s
is
d
u
e
to
th
e
f
ac
t
th
at
th
e
m
a
jo
r
ity
o
f
th
e
s
o
lu
tio
n
s
to
th
e
r
e
s
ea
r
ch
q
u
esti
o
n
s
in
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
E
vo
lu
tio
n
a
r
y
tr
en
d
s
in
a
u
to
m
a
tic
s
p
ee
ch
r
ec
o
g
n
itio
n
w
ith
…
(
Ga
b
r
iel
Olu
w
a
to
b
i
S
o
b
o
l
a
)
27
th
is
r
esear
ch
wo
r
k
h
a
v
e
b
ee
n
o
b
tain
ed
;
th
e
n
th
er
e
is
th
e
n
ee
d
to
r
ed
u
ce
th
e
s
ea
r
ch
p
ap
er
s
p
ar
ticu
lar
ly
f
o
r
p
ap
er
s
th
at
h
ad
lar
g
e
n
u
m
b
er
o
f
citatio
n
s
o
r
r
ef
e
r
en
ce
s
.
He
r
e
ea
ch
o
f
th
e
6
8
p
a
p
er
s
wer
e
an
aly
z
ed
to
m
a
k
e
s
elec
tio
n
s
o
f
p
ap
er
s
in
b
o
th
a
p
p
r
o
ac
h
es u
s
in
g
th
e
in
clu
s
io
n
a
n
d
ex
clu
s
io
n
cr
iter
ia.
2
.
3
.
6
.
B
a
ck
wa
rd
s
no
wba
llin
g
T
h
r
o
u
g
h
b
a
c
k
w
a
r
d
s
n
o
w
b
a
l
l
i
n
g
t
h
e
f
i
r
s
t
s
e
t
o
f
4
8
p
a
p
e
r
s
h
a
d
n
o
p
a
p
e
r
f
o
r
i
n
c
l
u
s
i
o
n
c
o
n
s
i
d
e
r
i
n
g
t
h
e
y
e
a
r
s
(
2
0
2
2
,
a
n
d
2
0
2
3
)
b
e
i
n
g
c
o
n
s
i
d
e
r
e
d
a
t
t
h
e
m
o
m
e
n
t
.
H
e
n
c
e
,
p
a
p
e
r
s
Q
1
P
1
B
1
t
h
r
o
u
g
h
Q
3
P
6
B
2
w
i
t
h
0
r
e
f
e
r
e
n
c
e
p
r
o
d
u
c
e
s
n
o
p
a
p
e
r
.
Q
1
P
1
F
2
,
w
i
t
h
2
r
e
f
e
r
e
n
c
e
s
,
h
a
s
n
o
p
a
p
e
r
f
o
r
i
n
c
l
u
s
i
o
n
d
u
e
t
o
t
h
e
t
i
t
l
e
s
o
f
t
h
e
p
a
p
e
r
s
.
Q
1
P
1
F
3
,
w
i
t
h
1
1
r
e
f
e
r
e
n
c
e
s
f
o
r
j
u
s
t
t
h
e
y
e
a
r
2
0
2
3
,
p
r
o
d
u
c
e
s
n
o
p
a
p
e
r
.
M
e
a
n
w
h
i
l
e
,
p
a
p
e
r
s
Q
1
P
1
F
4
,
Q
1
P
1
F
5
,
Q
1
P
1
F
6
,
Q
1
P
1
F
7
,
Q
1
P
1
F
8
,
Q
1
P
1
F
9
,
Q
1
P
1
F
1
0
,
Q
1
P
1
F
1
1
,
Q
1
P
1
F
1
2
,
Q
1
P
2
F
1
,
a
n
d
Q
1
P
2
F
2
w
i
t
h
0
,
1
2
,
0
,
1
,
0
,
2
1
,
0
,
0
,
0
,
0
,
a
n
d
0
r
e
f
e
r
e
n
c
e
s
r
e
s
p
e
c
t
i
v
e
l
y
p
r
o
d
u
c
e
n
o
p
a
p
e
r
f
o
r
i
n
c
l
u
s
i
o
n
.
I
n
a
d
d
i
t
i
o
n
,
p
a
p
e
r
s
Q
1
P
2
F
3
,
Q
1
P
2
F
4
,
Q
1
P
2
F
5
,
Q
1
P
2
F
6
,
Q
1
P
2
F
8
,
a
n
d
Q
1
P
2
F
9
w
i
t
h
1
,
1
,
3
6
7
,
1
,
0
,
a
n
d
0
r
e
f
e
r
e
n
c
e
s
r
e
s
p
e
c
t
i
v
e
l
y
p
r
o
d
u
c
e
s
n
o
p
a
p
e
r
w
h
i
l
e
p
a
p
e
r
Q
1
P
2
F
7
w
i
t
h
2
r
e
f
e
r
e
n
c
e
s
h
a
s
o
n
l
y
1
p
a
p
e
r
f
o
r
i
n
c
l
u
s
i
o
n
,
b
u
t
i
t
w
a
s
r
e
j
e
c
t
e
d
b
e
c
a
u
s
e
i
t
h
a
s
a
l
r
e
a
d
y
b
e
e
n
u
s
e
d
,
h
e
n
c
e
i
t
p
r
o
d
u
c
e
s
n
o
p
a
p
e
r
.
T
h
e
r
e
f
o
r
e
,
n
o
p
a
p
e
r
w
a
s
s
e
l
e
c
t
e
d
v
i
a
t
h
i
s
a
p
p
r
o
a
c
h
.
2
.
3
.
7
.
F
o
rwa
rd
s
no
wba
llin
g
T
h
r
o
u
g
h
t
h
e
f
o
r
wa
r
d
s
n
o
w
b
a
l
lin
g
a
p
p
r
o
ac
h
,
Q
1
P1
B
1
wi
th
1
9
cit
ati
o
n
s
p
r
o
d
u
ce
s
n
o
p
a
p
er
.
Pa
p
e
r
s
Q1
P
1
B
4
,
Q
1
P
1
B
5
,
Q1
P1
B
6
,
a
n
d
Q1
P
2
B
1
h
av
e
3
6
6
,
1
9
,
2
7
an
d
1
cit
ati
o
n
s
r
esp
ec
t
iv
el
y
p
r
o
d
u
ce
n
o
p
a
p
er
a
n
d
h
a
v
e
n
o
c
a
n
d
id
ate
(
s
)
f
o
r
i
n
c
lu
s
io
n
.
Q
1
P
2
B
3
,
was
i
d
e
n
ti
f
i
ed
wit
h
1
r
e
f
e
r
e
n
c
e
wh
ic
h
was
r
ej
ec
t
e
d
d
u
e
t
o
th
e
titl
e
;
h
e
n
c
e
n
o
p
a
p
e
r
w
as
s
ele
cte
d
f
o
r
i
n
cl
u
s
i
o
n
.
Q
1
P2
B
4
,
w
ith
3
3
cit
ati
o
n
s
,
p
r
o
d
u
ce
s
n
o
p
ap
er
.
Q
1
P
2
B
5
,
h
as
1
cit
ati
o
n
wh
ic
h
h
as
al
r
ea
d
y
b
e
en
i
n
c
lu
d
ed
i
n
t
h
e
s
el
ec
t
e
d
lis
t
o
f
p
ap
er
s
f
o
r
t
h
is
w
o
r
k
.
Hen
ce
n
o
p
a
p
e
r
was
s
ele
cte
d
f
r
o
m
it
.
Q
1
P2
B
6
wi
th
2
ci
tat
io
n
s
a
n
d
Q
1
P
2
B
7
,
wit
h
1
0
cit
ati
o
n
s
,
h
a
v
e
n
o
p
a
p
e
r
t
o
b
e
i
n
c
lu
d
ed
b
ase
d
o
n
t
h
e
c
r
it
e
r
ia
.
Als
o
,
Q
1
P4
B
1
wit
h
2
2
c
ita
ti
o
n
s
wi
th
in
t
h
e
y
e
ar
c
o
n
s
id
e
r
e
d
,
p
r
o
d
u
ce
s
n
o
p
a
p
e
r
.
Q
1
P
4
B
3
,
wit
h
3
p
a
p
e
r
s
o
n
s
p
e
ec
h
,
h
a
d
n
o
n
e
t
o
b
e
s
ele
ct
ed
f
o
r
in
cl
u
s
io
n
.
QI
P
5
B
1
h
as
4
ci
tati
o
n
s
a
n
d
Q
I
P
5
B
2
h
a
d
1
1
4
0
cit
ati
o
n
s
b
u
t
n
o
n
e
was
s
e
lect
e
d
d
u
e
t
o
t
h
e
f
ac
t t
h
a
t t
h
e
m
aj
o
r
it
y
o
f
t
h
o
s
e
p
a
p
er
s
h
ad
li
n
k
s
o
r
s
im
i
la
r
it
ies
to
o
t
h
e
r
in
cl
u
d
e
d
p
a
p
er
s
.
S
im
i
la
r
ly
,
Q
I
P
5
B
3
,
QI
P
5
B
4
,
a
n
d
Q
I
P5
B
5
h
a
v
e
8
,
8
,
an
d
1
8
ci
tat
i
o
n
s
r
es
p
e
cti
v
e
ly
b
u
t
h
a
d
n
o
ca
n
d
i
d
at
e
f
o
r
i
n
cl
u
s
i
o
n
w
h
il
e
QI
P
5
B
6
wi
th
6
c
itat
i
o
n
s
,
p
r
o
d
u
ce
s
n
o
p
a
p
e
r
as
w
ell
.
Q2
P1
B
1
,
wit
h
2
5
5
c
itat
io
n
s
p
r
o
d
u
ce
1
p
a
p
er
i
n
d
ic
ate
d
as
PG
1
.
Q
2
P
1
B
2
,
wit
h
2
ci
tat
io
n
s
,
p
r
o
d
u
c
es n
o
p
ap
er
.
Q
2
P
2
B
2
,
wi
th
4
cita
ti
o
n
s
,
p
r
o
d
u
ce
s
n
o
p
a
p
e
r
.
No
p
ap
e
r
w
as
p
r
o
d
u
c
e
d
b
y
Q
2
P
3
B
3
(
1
4
cit
ati
o
n
s
)
,
Q
2
P
3
B
4
(
1
0
2
0
ci
tati
o
n
s
)
,
Q2
P
3
B
5
(
4
5
ci
tati
o
n
s
)
,
Q
2
P
4
B
1
(
2
ci
tat
io
n
s
)
,
Q
2
P
4
B
2
(
5
c
ita
ti
o
n
s
)
,
Q
2
P
4
B
3
(
0
c
i
tati
o
n
s
)
,
Q2
P5
B
1
(
1
0
9
c
ita
ti
o
n
s
)
,
Q
2
P
5
B
2
(
0
c
itati
o
n
)
,
Q2
P
5
B
3
(
1
4
9
ci
tat
io
n
s
)
,
Q
4
P
1
B
1
(
0
c
ita
ti
o
n
)
,
Q
4
P1
B
2
(
0
c
ita
ti
o
n
)
,
Q4
P
1
B
3
(
3
c
itat
io
n
s
)
,
Q
4
P
1
B
4
(
0
cit
ati
o
n
)
,
Q4
P
1
B
5
(
4
0
c
ita
tio
n
s
)
,
Q
4
P
1
B
6
(
3
ci
tati
o
n
s
)
,
Q5
P
1
B
1
(
1
c
itat
io
n
)
,
Q5
P
1
B
2
(
2
cit
ati
o
n
s
)
,
Q
5
P
1
B
3
(
3
7
cit
ati
o
n
s
)
,
Q5
P
1
B
4
(
0
cita
ti
o
n
)
,
Q
5
P
1
B
5
(
5
6
c
ita
ti
o
n
s
)
,
w
h
i
le
Q5
P1
B
6
(
4
8
c
ita
ti
o
n
s
)
h
as
o
n
e
p
a
p
e
r
f
o
r
in
cl
u
s
i
o
n
b
u
t
was
r
ej
ec
te
d
b
e
ca
u
s
e
it
h
as
al
r
ea
d
y
b
e
e
n
u
s
e
d
,
h
e
n
ce
,
it
p
r
o
d
u
c
es
n
o
p
a
p
er
.
I
n
a
d
d
iti
o
n
,
Q
5
P
1
B
7
,
wi
th
1
1
cit
ati
o
n
s
,
p
r
o
d
u
c
es
n
o
p
ap
er
as
well
as
Q3
P
3
B
1
(
5
3
9
c
ita
ti
o
n
s
)
,
Q3
P
3
B
2
(
5
5
cit
ati
o
n
s
)
,
Q
3
P
3
B
3
(
3
3
cit
ati
o
n
s
)
,
Q
3
P
6
B
1
(
2
0
cita
ti
o
n
s
)
,
Q
3
P6
B
2
(
0
c
ita
ti
o
n
)
,
Q
1
P1
F
2
(
3
5
cit
ati
o
n
s
)
,
Q1
P
1
F
3
(
2
1
c
itati
o
n
s
)
,
Q1
P1
F
4
(
6
cit
ati
o
n
s
)
,
Q
1
P
1
F
5
(
1
ci
tat
io
n
)
,
Q
1
P1
F
6
(
1
3
cit
ati
o
n
s
)
Q1
P
1
F
7
(
2
cit
ati
o
n
s
)
,
Q
1
P
1
F8
(
6
cit
ati
o
n
s
)
,
Q
1
P
1
F
9
(
4
c
ita
ti
o
n
s
)
,
Q
1
P1
F
1
0
(
6
ci
tati
o
n
s
)
,
Q
1
P
1
F
1
1
(
1
ci
tati
o
n
)
,
Q1
P
1
F
1
2
(
0
cit
ati
o
n
)
,
Q
1
P
2
F
1
(
1
7
c
ita
ti
o
n
s
)
,
Q
1
P
2
F2
(
2
cita
ti
o
n
s
)
,
Q
1
P
2
F
3
(
3
c
ita
ti
o
n
s
)
,
Q
1
P2
F4
(
4
cit
ati
o
n
s
)
,
Q
1
P2
F5
(
2
cit
a
tio
n
s
)
,
Q
1
P
2
F
6
(
0
ci
tat
io
n
)
,
Q1
P
2
F
7
(
0
ci
tat
io
n
)
,
Q
1
P
2
F
8
(
0
ci
tat
io
n
)
,
an
d
Q1
P
2
F
9
(
0
cit
ati
o
n
)
.
T
h
e
o
n
l
y
p
a
p
e
r
(
P
GI
)
to
b
e
i
n
c
lu
d
e
d
wi
t
h
th
e
s
ta
r
t
s
et
is
c
o
n
tai
n
ed
i
n
T
ab
l
e
7
.
T
ab
le
7
.
I
ter
atio
n
2
s
elec
ted
p
ap
er
(
1
)
R
e
f
.
n
o
.
P
a
p
e
r
i
d
e
n
t
i
t
y
c
o
d
e
[
9
0
]
P
G
1
2
.
3
.
8
.
Su
m
m
a
r
y
o
f
it
er
a
t
i
o
n
2
A
p
a
r
t
f
r
o
m
t
h
e
1
3
-
s
t
a
r
t
s
e
t
o
f
p
a
p
e
r
s
u
s
e
d
f
o
r
t
h
i
s
r
e
s
e
a
r
c
h
w
o
r
k
,
o
n
l
y
1
p
a
p
e
r
h
a
s
b
e
e
n
s
e
l
e
ct
e
d
v
i
a
t
h
e
b
a
c
k
w
a
r
d
s
n
o
w
b
al
l
i
n
g
a
n
d
f
o
r
w
a
r
d
s
n
o
w
b
al
l
i
n
g
a
p
p
r
o
a
c
h
es
d
u
r
i
n
g
t
h
e
i
t
e
r
at
i
o
n
2
,
a
n
d
t
h
is
i
s
t
h
e
p
a
p
e
r
t
o
b
e
i
n
c
l
u
d
e
d
i
n
t
h
e
a
l
r
e
a
d
y
s
e
l
e
c
ted
6
8
p
a
p
e
r
s
.
A
t
o
t
a
l
o
f
4
2
0
p
a
p
e
r
s
w
e
r
e
a
n
a
l
y
z
e
d
v
i
a
t
h
e
b
a
c
k
w
a
r
d
s
n
o
w
b
a
l
l
i
n
g
a
p
p
r
o
a
c
h
a
n
d
a
t
o
t
a
l
o
f
4
,
2
6
7
p
ap
er
s
w
e
r
e
a
n
a
l
y
z
e
d
v
i
a
t
h
e
f
o
r
w
a
r
d
s
n
o
w
b
a
ll
i
n
g
a
p
p
r
o
a
c
h
.
T
h
u
s
,
a
t
o
t
al
o
f
4
,
6
8
7
p
a
p
e
r
s
w
e
r
e
a
n
a
l
y
z
e
d
i
n
t
h
e
s
e
c
o
n
d
i
t
e
r
a
ti
o
n
.
T
h
e
t
o
t
a
l
n
u
m
b
e
r
o
f
p
a
p
e
r
s
s
e
le
c
t
e
d
s
o
f
ar
i
s
6
9
.
I
n
t
h
e
n
e
x
t
i
t
e
r
a
ti
o
n
,
t
h
e
s
i
n
g
le
p
a
p
e
r
w
as
e
x
p
e
r
i
m
e
n
t
e
d
w
it
h
b
o
t
h
t
h
e
b
a
c
k
w
a
r
d
a
n
d
f
o
r
w
a
r
d
s
n
o
w
b
a
l
lin
g
a
p
p
r
o
a
c
h
e
s
.
2
.
3
.
9
.
I
t
er
a
t
io
n 3
I
n
th
is
iter
atio
n
3
,
th
e
s
in
g
le
p
ap
er
PG1
wh
ich
was
s
elec
ted
v
ia
th
e
in
clu
s
io
n
/e
x
clu
s
io
n
cr
i
ter
ia
f
r
o
m
th
e
p
r
e
v
io
u
s
iter
atio
n
w
as
als
o
an
aly
z
e
d
u
s
in
g
t
h
e
f
o
r
war
d
an
d
th
e
b
ac
k
war
d
s
n
o
wb
allin
g
a
p
p
r
o
ac
h
es
.
T
h
e
p
r
o
ce
s
s
es o
f
th
is
p
r
esen
t stag
e
wen
t th
u
s
:
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
6
:
20
-
43
28
‒
B
ac
k
war
d
s
n
o
wb
allin
g
:
f
o
r
p
a
p
er
PG1
n
o
p
ap
e
r
was
id
en
tif
ied
with
in
t
h
e
tim
e
f
r
am
e
o
f
2
0
2
2
to
2
0
2
3
,
h
en
ce
,
n
o
p
a
p
er
was selec
ted
.
‒
Fo
r
war
d
s
n
o
wb
allin
g
:
p
ap
er
P
G1
h
ad
n
o
citatio
n
,
h
en
ce
n
o
p
ap
er
was selec
ted
v
ia
th
is
ap
p
r
o
ac
h
.
T
h
e
s
u
m
m
ar
y
o
f
iter
atio
n
3
is
th
at
n
o
p
ap
er
was a
n
aly
z
ed
v
i
a
th
e
b
ac
k
war
d
an
d
f
o
r
war
d
s
n
o
wb
allin
g
ap
p
r
o
ac
h
in
th
is
iter
atio
n
,
h
en
ce
,
n
o
s
elec
tio
n
was
m
ad
e,
an
d
n
o
p
ap
e
r
was
an
aly
z
e
d
in
iter
atio
n
3
.
Ho
wev
er
,
f
o
r
th
e
p
u
r
p
o
s
e
o
f
ef
f
icien
c
y
,
a
s
in
g
le
p
a
p
er
(
2
0
1
3
)
was
a
n
aly
z
ed
v
ia
th
e
b
ac
k
war
d
s
n
o
wb
allin
g
ap
p
r
o
ac
h
,
h
o
wev
er
it
was
n
o
t
s
elec
ted
b
ec
au
s
e
th
e
co
n
ten
ts
wer
e
o
n
f
ea
tu
r
e
ex
tr
ac
tio
n
tech
n
iq
u
es
wh
ich
h
av
e
alr
ea
d
y
b
ee
n
s
tu
d
ied
in
th
e
p
r
ev
i
o
u
s
lite
r
a
tu
r
es
.
T
h
u
s
,
a
to
tal
o
f
6
9
p
ap
er
s
h
as
b
ee
n
s
elec
ted
v
ia
th
e
f
o
r
war
d
a
n
d
b
ac
k
war
d
s
n
o
wb
allin
g
ap
p
r
o
a
ch
es
af
ter
th
e
th
ir
d
iter
atio
n
a
n
d
ad
d
in
g
th
is
to
t
h
e
s
tar
t
s
et
lis
t
o
f
p
a
p
er
s
(
1
3
)
g
iv
es
a
to
tal
o
f
8
2
p
ap
er
s
s
elec
ted
in
th
is
wo
r
k
as
p
r
o
v
id
e
d
in
T
ab
le
8
.
T
h
e
a
u
t
h
o
r
s
o
f
t
h
e
s
el
e
c
t
e
d
p
a
p
e
r
s
u
t
i
l
i
z
e
d
i
n
t
h
i
s
c
u
r
r
e
n
t
s
t
u
d
y
w
e
r
e
n
o
t
c
o
n
t
a
c
t
e
d
t
o
p
r
e
v
e
n
t
d
e
l
a
y
o
f
t
h
i
s
s
t
u
d
y
c
o
n
s
i
d
e
r
in
g
t
h
e
f
a
c
t
t
h
a
t
t
h
e
r
e
a
s
o
n
f
o
r
c
o
n
t
a
c
t
i
n
g
,
is
t
o
s
o
u
r
c
e
f
o
r
a
d
d
i
t
i
o
n
a
l
m
at
e
r
i
a
ls
(
i
n
r
e
l
a
t
i
o
n
t
o
t
h
i
s
s
t
u
d
y
)
f
r
o
m
t
h
e
m
.
T
h
e
r
e
c
a
n
a
ls
o
b
e
a
n
e
l
e
m
e
n
t
o
f
b
i
as
h
e
n
c
e
t
h
e
r
e
a
s
o
n
f
o
r
n
o
t
c
o
n
t
a
c
t
i
n
g
a
n
y
o
f
t
h
e
a
u
t
h
o
r
s
.
T
ab
le
8
.
T
o
tal
s
elec
ted
lis
t o
f
p
ap
er
s
in
clu
d
i
n
g
th
e
1
3
s
tar
t set p
ap
er
s
(
8
2
)
R
e
f
.
n
o
.
P
a
p
e
r
i
d
e
n
t
i
t
y
c
o
d
e
R
e
f
.
n
o
.
P
a
p
e
r
i
d
e
n
t
i
t
y
c
o
d
e
R
e
f
.
n
o
.
P
a
p
e
r
i
d
e
n
t
i
t
y
c
o
d
e
R
e
f
.
n
o
.
P
a
p
e
r
i
d
e
n
t
i
t
y
c
o
d
e
[
1
]
Q
1
P
1
[
6
2
]
Q
1
P
2
B
6
[
7
5
]
Q
2
P
5
B
3
[
2
0
]
Q
1
P
1
F
3
[
2
]
Q
1
P
2
[
6
5
]
Q
1
P
2
B
7
[
7
8
]
Q
4
P
1
B
1
[
1
2
]
Q
1
P
1
F
4
[
3
]
Q
1
P
4
[
6
8
]
Q
1
P
4
B
1
[
8
2
]
Q
4
P
1
B
2
[
1
0
]
Q
1
P
1
F
5
[
4
]
Q
1
P
5
[
1
3
]
Q
1
P
4
B
3
[
8
5
]
Q
4
P
1
B
3
[
4
2
]
Q
1
P
1
F
6
[
3
0
]
Q
2
P
1
[
7
4
]
Q
I
P
5
B
1
[
8
8
]
Q
4
P
1
B
4
[
1
4
]
Q
1
P
1
F
7
[
3
2
]
Q
2
P
2
[
7
7
]
Q
I
P
5
B
2
[
3
8
]
Q
4
P
1
B
5
[
4
8
]
Q
1
P
1
F
8
[
1
7
]
Q
2
P
3
[
8
1
]
Q
I
P
5
B
3
[
4
1
]
Q
4
P
1
B
6
[
1
5
]
Q
1
P
1
F
9
[
3
4
]
Q
2
P
4
[
8
4
]
Q
I
P
5
B
4
[
4
1
]
Q
5
P
1
B
1
[
8
]
Q
1
P
1
F
1
0
[
2
5
]
Q
2
P
5
[
8
7
]
Q
I
P
5
B
5
[
4
7
]
Q
5
P
1
B
2
[
5
8
]
Q
1
P
1
F
1
1
[
2
9
]
Q
3
P
3
[
3
7
]
Q
I
P
5
B
6
[
5
1
]
Q
5
P
1
B
3
[
6
1
]
Q
1
P
1
F
1
2
[
3
3
]
Q
3
P
6
[
4
0
]
Q
2
P
1
B
1
[
5
4
]
Q
5
P
1
B
4
[
9
]
Q
1
P
2
F
1
[
5
]
Q
4
P
1
[
2
2
]
Q
2
P
1
B
2
[
5
7
]
Q
5
P
1
B
5
[
2
1
]
Q
1
P
2
F
2
[
3
5
]
Q
5
P
1
[
4
6
]
Q
2
P
2
B
2
[
6
0
]
Q
5
P
1
B
6
[
7
0
]
Q
1
P
2
F
3
[
3
6
]
Q
1
P
1
B
1
[
5
0
]
Q
2
P
3
B
3
[
6
4
]
Q
5
P
1
B
7
[
7
3
]
Q
1
P
2
F
4
[
3
9
]
Q
1
P
1
B
4
[
5
3
]
Q
2
P
3
B
4
[
6
7
]
Q
3
P
3
B
1
[
2
3
]
Q
1
P
2
F
5
[
4
3
]
Q
1
P
1
B
5
[
5
6
]
Q
2
P
3
B
5
[
6
9
]
Q
3
P
3
B
2
[
8
0
]
Q
1
P
2
F
6
[
4
5
]
Q
1
P
1
B
6
[
5
9
]
Q
2
P
4
B
1
[
7
2
]
Q
3
P
3
B
3
[
8
3
]
Q
1
P
2
F
7
[
4
9
]
Q
1
P
2
B
1
[
6
3
]
Q
2
P
4
B
2
[
7
6
]
Q
3
P
6
B
1
[
8
6
]
Q
1
P
2
F
8
[
5
2
]
Q
1
P
2
B
3
[
6
6
]
Q
2
P
4
B
3
[
7
9
]
Q
3
P
6
B
2
[
8
9
]
Q
1
P
2
F
9
[
5
5
]
Q
1
P
2
B
4
[
1
9
]
Q
2
P
5
B
1
[
1
1
]
Q
1
P
1
F
2
[
9
0
]
P
G
1
[
1
8
]
Q
1
P
2
B
5
[
7
1
]
Q
2
P
5
B
2
2
.
4
.
E
f
f
iciency
o
f
s
y
s
t
em
a
t
ic
lite
ra
t
ure
re
v
iew
us
ing
s
no
wba
llin
g
I
n
ter
m
s
o
f
th
e
n
u
m
b
e
r
o
f
p
ap
er
s
u
tili
z
ed
f
o
r
th
is
r
esear
ch
,
th
e
ef
f
icien
c
y
o
f
u
tili
z
ati
o
n
ca
n
b
e
co
m
p
u
ted
as:
i)
Star
t
s
et:
1
8
p
ap
er
s
(
o
r
ca
n
d
id
ates)
f
o
r
in
itially
s
elec
ted
an
d
1
3
s
tar
t
s
et
p
ap
er
s
wer
e
in
clu
d
ed
,
i.e
.
=
13
18
=
72%
.
ii)
I
ter
atio
n
1
:
4
,
3
7
7
ca
n
d
id
ates
f
r
o
m
s
n
o
wb
allin
g
f
r
o
m
th
e
s
tar
t
s
et,
an
d
6
8
p
a
p
er
s
wer
e
in
clu
d
e
d
,
i
.
e.
=
68
4377
=
1
.
6%
.
iii)
I
ter
atio
n
2
:
4
,
6
8
7
ca
n
d
id
ates
f
o
r
in
clu
s
io
n
wer
e
g
en
er
ate
d
in
th
e
b
ac
k
war
d
an
d
f
o
r
war
d
s
n
o
wb
allin
g
,
1
p
ap
er
was in
clu
d
ed
,
i.e
.
=
1
4687
=
0
.
02%
.
i
v
)
I
t
e
r
a
t
i
o
n
3
:
1
c
a
n
d
i
d
a
te
w
as
e
x
a
m
i
n
e
d
,
a
n
d
n
o
p
a
p
e
r
w
a
s
i
n
c
lu
d
e
d
,
a
n
d
h
e
n
c
e
t
h
e
e
f
f
i
c
i
e
n
c
y
b
e
c
o
m
e
s
:
0%
T
h
e
o
v
e
r
all
ef
f
icien
cy
b
ec
o
m
e
s
:
=
[
(
13
+
68
+
1
+
0
)
18
+
4377
+
4
6
8
7
+
1
]
×
100
=
0
.
90
%
.
Acc
o
r
d
in
g
to
W
o
h
lin
[
2
4
]
,
an
o
v
er
all
ef
f
icien
c
y
o
f
3
.
7
%
w
as
o
b
tain
ed
.
I
n
th
e
wo
r
k
,
th
e
n
u
m
b
er
o
f
p
ap
er
s
an
aly
ze
d
wer
e
n
o
t
as
v
o
lu
m
in
o
u
s
co
m
p
a
r
ed
t
o
wh
at
is
av
ailab
le
in
th
is
r
esear
c
h
w
o
r
k
.
B
esid
es,
if
th
e
y
ea
r
r
an
g
e
with
in
wh
ich
p
ap
e
r
s
wer
e
s
elec
ted
f
o
r
th
e
iter
ati
o
n
2
h
ad
b
ee
n
ex
ten
d
ed
b
ey
o
n
d
th
e
two
y
ea
r
s
o
f
2
0
2
2
to
2
0
2
3
,
th
e
ef
f
icien
c
y
wo
u
ld
h
av
e
in
cr
ea
s
ed
b
ey
o
n
d
th
e
0
.
9
0
%
o
b
tain
ed
.
I
n
ad
d
itio
n
to
th
is
,
it
was
o
b
v
io
u
s
th
at
t
h
er
e
was
a
s
u
r
g
e
in
th
e
n
u
m
b
er
o
f
p
a
p
er
s
s
elec
ted
f
r
o
m
th
e
s
tar
t
s
et,
in
th
e
it
er
atio
n
1
,
h
o
wev
er
,
th
e
n
u
m
b
er
s
elec
ted
f
o
llo
ws
a
d
escen
d
in
g
tr
e
n
d
as
th
e
n
u
m
b
er
o
f
iter
atio
n
s
in
cr
ea
s
es,
a
n
d
co
n
s
eq
u
en
tly
th
e
en
er
g
y
e
x
p
en
d
ed
in
s
ea
r
ch
in
g
an
d
s
elec
tin
g
p
a
p
er
s
d
r
o
p
s
,
wh
ich
im
p
lies
an
in
cr
ea
s
ed
a
n
d
o
p
tim
i
z
e
d
way
o
f
s
elec
tin
g
p
ap
er
s
f
o
r
r
ev
iewin
g
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
E
vo
lu
tio
n
a
r
y
tr
en
d
s
in
a
u
to
m
a
tic
s
p
ee
ch
r
ec
o
g
n
itio
n
w
ith
…
(
Ga
b
r
iel
Olu
w
a
to
b
i
S
o
b
o
l
a
)
29
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
3
.
1
.
Da
t
a
e
x
t
ra
ct
i
o
n
Af
ter
a
th
o
r
o
u
g
h
r
ev
iew
o
f
ea
ch
o
f
th
e
9
6
p
a
p
er
s
s
elec
ted
f
o
r
th
is
cu
r
r
en
t
s
tu
d
y
.
R
elev
an
t
in
f
o
r
m
atio
n
an
d
f
in
d
i
n
g
s
r
eg
a
r
d
in
g
ASR
in
r
elatio
n
to
ea
ch
o
f
th
e
r
esear
ch
q
u
esti
o
n
s
wer
e
ex
tr
ac
ted
.
T
h
ese
ar
e
p
r
esen
ted
in
T
ab
les 9
to
1
3
in
th
e
ap
p
en
d
ix
.
3
.
2
.
Resul
t
s
o
f
ex
t
ra
ct
ed
da
t
a
f
ro
m
s
elec
t
ed
pa
pers
T
h
e
r
esu
lts
o
f
an
d
f
in
d
in
g
s
f
r
o
m
ex
tr
ac
ted
d
ata
f
r
o
m
s
elec
ted
p
ap
er
s
r
eg
a
r
d
in
g
th
e
r
esear
ch
q
u
esti
o
n
:
i)
Q1
:
wh
en
d
id
t
h
e
d
ev
elo
p
m
e
n
t o
f
ASR
with
AI
s
tar
t?
is
p
r
es
en
ted
in
T
ab
le
9
in
th
e
ap
p
e
n
d
ix
.
ii)
Q2
:
wh
at
ty
p
es
o
f
AI
h
av
e
b
ee
n
ex
p
lo
r
ed
b
y
v
ar
io
u
s
r
es
ea
r
ch
er
s
to
d
ev
elo
p
ASR
?
is
p
r
esen
ted
in
T
ab
le
1
0
in
th
e
ap
p
en
d
ix
.
iii)
Q3
:
wh
at
ar
e
th
e
b
en
ef
its
an
d
d
is
ad
v
an
tag
es
o
f
ex
p
lo
r
in
g
e
ac
h
o
f
th
o
s
e
alg
o
r
ith
m
s
m
en
ti
o
n
ed
in
2
?
is
p
r
esen
ted
in
T
a
b
le
1
1
i
n
th
e
a
p
p
en
d
i
x
.
iv
)
Q4
:
d
o
es
th
e
ex
p
l
o
r
atio
n
o
f
AI
f
o
r
b
u
ild
in
g
a
n
ASR
s
y
s
te
m
h
av
e
p
e
r
f
o
r
m
an
ce
im
p
r
o
v
e
m
en
t
o
v
er
th
e
tr
ad
itio
n
al
p
r
o
b
ab
ilis
tic
ap
p
r
o
ac
h
es?
is
p
r
esen
ted
in
T
ab
le
1
2
in
th
e
a
p
p
en
d
ix
.
v)
Q5
:
h
as
th
e
in
tr
o
d
u
ctio
n
o
f
AI
in
th
e
d
ev
elo
p
m
en
t
an
d
m
o
d
ellin
g
o
f
ASR
b
ee
n
ab
le
to
ac
h
iev
e
its
u
ltima
te
g
o
al
to
p
e
r
f
o
r
m
as a
h
u
m
an
lis
ten
er
?
is
p
r
esen
ted
in
T
ab
le
1
3
in
th
e
ap
p
en
d
ix
.
Hav
en
’
t
d
o
n
e
a
th
o
r
o
u
g
h
r
ev
i
ew
o
n
th
e
r
esear
ch
q
u
esti
o
n
s
1
-
5
as
en
u
m
er
ated
,
th
e
r
esu
lts
o
f
v
ar
i
o
u
s
s
tu
d
ies
s
u
r
v
ey
ed
wer
e
lis
ted
i
n
T
ab
les
9
to
1
3
.
I
t
ca
n
b
e
s
ee
n
f
r
o
m
th
e
tab
les
th
at
v
ar
io
u
s
ap
p
r
o
ac
h
es
an
d
m
eth
o
d
o
l
o
g
ies
in
v
o
lv
i
n
g
AI
,
p
ar
ticu
lar
ly
m
ac
h
in
e
lear
n
i
n
g
an
d
d
ee
p
lear
n
in
g
,
h
a
v
e
b
ee
n
ex
p
lo
r
ed
b
y
v
ar
io
u
s
r
esear
ch
er
s
to
ac
h
iev
e
th
e
co
n
v
er
s
io
n
o
f
s
p
ee
ch
to
tex
t.
T
h
e
tab
les
f
u
r
th
er
illu
s
tr
ates
o
th
er
ex
is
tin
g
m
eth
o
d
s
lik
e
th
e
p
r
o
b
ab
ilis
tic
ap
p
r
o
ac
h
[
1
7
]
,
[
2
5
]
,
[
3
5
]
,
tem
p
late
b
ased
ap
p
r
o
ac
h
[
9
1
]
.
Mo
r
e
r
esear
ch
wo
r
k
s
ar
e
o
n
g
o
in
g
[
6
]
,
[
7
]
to
d
e
v
elo
p
a
m
o
r
e
r
o
b
u
s
t
s
p
ee
ch
r
ec
o
g
n
itio
n
th
at
ca
n
with
s
tan
d
v
ar
io
u
s
v
ar
iab
ilit
ies
lik
e
th
e
en
v
ir
o
n
m
en
t,
n
o
is
e,
s
p
ea
k
er
a
n
d
m
o
s
t
im
p
o
r
tan
tly
t
o
ac
h
ie
v
e
th
e
u
ltima
te
g
o
al
o
f
s
p
ee
ch
r
ec
o
g
n
itio
n
.
I
t
ca
n
also
b
e
s
ee
n
th
at
f
o
r
ex
tr
ac
tin
g
f
ea
tu
r
es
f
r
o
m
s
p
ee
ch
s
ig
n
als
,
m
ajo
r
ity
o
f
th
e
r
esear
ch
er
s
e
x
p
lo
r
ed
th
e
MFC
C
s
[
1
]
,
[
3
2
]
,
b
u
t
w
h
en
v
ar
iab
ilit
y
is
co
n
s
id
er
ed
,
th
en
o
th
er
m
eth
o
d
s
o
f
f
ea
t
u
r
e
e
x
tr
ac
tio
n
s
u
ch
as
t
h
e
wav
elet
tr
an
s
f
o
r
m
,
R
ASTA
,
an
d
ze
r
o
c
r
o
s
s
in
g
p
ea
k
am
p
litu
d
e
(
Z
C
P
A)
ar
e
ex
p
l
o
r
ed
[
1
7
]
,
[
3
2
]
.
Acc
o
r
d
in
g
to
Ma
lik
et
a
l.
[
2
]
,
it
was
n
o
ted
th
at
h
y
b
r
id
MFC
C
,
p
er
ce
p
tu
al
lin
ea
r
p
r
ed
ictiv
e
(
PLP
)
,
an
d
L
PC
ca
n
b
e
ex
p
l
o
r
ed
m
o
s
t
im
p
o
r
tan
tly
f
o
r
n
o
is
y
e
n
v
ir
o
n
m
e
n
ts
.
I
n
[
1
]
,
it
was
s
h
o
wn
th
at
th
e
W
E
R
r
ed
u
ce
d
b
y
3
0
%
wh
en
AI
is
e
x
p
lo
r
ed
o
v
er
th
e
p
r
o
b
ab
ilis
tic
ap
p
r
o
ac
h
es
ev
en
th
o
u
g
h
s
o
m
e
s
tu
d
ies
s
ti
ll
p
r
ef
er
to
u
s
e
t
h
e
c
o
n
v
e
n
tio
n
al
HM
M
in
ex
tr
ac
tin
g
f
ea
tu
r
es
a
n
d
s
u
b
s
eq
u
en
t
tr
ain
in
g
u
s
in
g
m
ac
h
in
e
lear
n
in
g
[
1
]
,
[
1
7
]
,
[
3
5
]
.
Oth
er
r
esear
ch
s
tu
d
ies
e
x
p
lo
r
ed
d
ee
p
n
etwo
r
k
s
lik
e
C
NN,
R
NN
f
o
r
f
ea
tu
r
e
e
x
tr
ac
tio
n
an
d
class
if
icatio
n
o
f
th
eir
m
o
d
els
[
3
]
.
I
n
[
2
]
,
it
was
s
h
o
wn
th
at
SVM
ca
n
b
e
a
b
etter
alg
o
r
ith
m
to
b
e
u
s
ed
f
o
r
s
p
ee
ch
r
ec
o
g
n
itio
n
,
s
in
ce
it h
as a
n
ac
c
u
r
ac
y
o
f
7
7
.
6
%,
wh
ich
is
4
%
b
etter
th
an
th
e
HM
M
[
9
2
]
.
3
.
3
.
Dis
cus
s
io
n:
t
a
ilo
re
d
t
o
e
a
ch
re
s
ea
rc
h ques
t
io
n
T
h
e
f
o
cu
s
o
f
th
is
r
esear
ch
is
to
ad
d
r
ess
th
e
r
esear
ch
q
u
esti
o
n
s
in
r
elatio
n
to
th
e
ev
o
lu
tio
n
a
r
y
tr
en
d
in
ASR
s
y
s
tem
s
.
B
a
s
ed
o
n
th
e
r
esu
lts
an
d
f
in
d
in
g
s
d
etailed
i
n
th
e
T
ab
les
9
to
13
,
th
e
f
in
d
in
g
s
o
f
ea
ch
q
u
esti
o
n
ar
e
h
er
eb
y
ad
d
r
ess
ed
as f
o
llo
ws:
3
.
3
.
1
.
Q
ues
t
io
n 1
W
h
en
d
id
th
e
d
ev
elo
p
m
en
t
o
f
ASR
with
AI
s
tar
t?
AS
R
h
as
b
ee
n
an
ar
ea
o
f
r
esear
ch
t
h
at
s
tar
ted
as
ea
r
ly
as
th
e
1
9
5
0
s
.
I
n
f
ac
t
m
o
s
t
r
esear
ch
er
s
[
2
]
,
[
3
4
]
,
[
9
3
]
i
n
th
e
f
ield
o
f
s
p
ee
ch
r
ec
o
g
n
iti
o
n
lis
ted
th
is
p
e
r
io
d
as
th
e
b
ir
th
o
f
s
p
ee
ch
r
ec
o
g
n
itio
n
.
I
n
[
3
]
,
[
1
7
]
,
[
2
5
]
,
it
w
as
s
tated
th
at
s
p
ee
ch
r
ec
o
g
n
i
tio
n
s
tar
ted
at
B
ell
L
ab
o
r
ato
r
y
an
d
it
was
co
i
n
ed
Au
d
r
ey
s
p
ee
c
h
r
ec
o
g
n
itio
n
[
3
]
,
[
9
1
]
.
I
n
[
9
1
]
,
s
p
ee
ch
r
ec
o
g
n
itio
n
s
tar
ted
as
ea
r
l
y
as
th
e
1
9
2
0
s
an
d
,
in
1
9
2
0
,
th
e
f
ir
s
t
m
ac
h
in
e
to
r
ec
o
g
n
ize
s
p
e
ec
h
,
R
ad
io
R
ex
to
y
was
m
an
u
f
ac
tu
r
ed
.
Ho
wev
er
,
th
e
in
co
r
p
o
r
atio
n
o
f
v
ar
io
u
s
A
I
with
ASR
v
ar
ies in
y
ea
r
s
d
ep
en
d
in
g
o
n
th
e
alg
o
r
ith
m
u
s
ed
.
Acc
o
r
d
in
g
to
An
u
s
u
y
a
a
n
d
K
atti
[
9
1
]
,
n
eu
r
al
n
etwo
r
k
s
wer
e
f
ir
s
t
in
tr
o
d
u
ce
d
i
n
th
e
1
9
5
0
s
,
b
u
t
th
ey
d
id
n
o
t
p
r
o
v
e
u
s
ef
u
l
in
itially
b
ec
au
s
e
th
ey
h
ad
m
an
y
p
r
a
ctica
l
p
r
o
b
lem
s
.
T
h
en
th
e
ex
p
lo
r
atio
n
o
f
n
eu
r
al
n
etwo
r
k
s
s
tar
ted
in
th
e
1
9
8
0
s
[
2
5
]
.
W
ith
ML
P,
wh
ich
is
a
ty
p
e
o
f
d
ee
p
lear
n
in
g
,
c
o
m
in
g
i
n
to
ex
is
ten
ce
in
th
e
1
9
9
0
s
[
1
7
]
,
r
esear
ch
er
s
b
eg
a
n
to
ex
p
lo
r
e
h
y
b
r
id
HM
M+
ANN
[
2
]
an
d
t
h
en
th
e
u
s
e
o
f
DNN
in
2
0
1
0
[
4
]
,
a
n
d
an
en
d
t
o
en
d
DNN
in
2
0
1
7
[
3
]
.
At
p
r
esen
t,
r
esear
ch
er
s
a
r
e
ex
p
lo
r
in
g
d
ee
p
lear
n
i
n
g
ar
ch
itectu
r
es
lik
e
th
e
h
y
b
r
id
L
STM
+RNN
[
2
]
an
d
tr
an
s
f
o
r
m
er
s
[
9
4
]
.
3
.
3
.
2
.
Q
ues
t
io
n 2
W
h
at
ty
p
es
o
f
AI
h
av
e
b
ee
n
ex
p
lo
r
ed
b
y
v
ar
io
u
s
r
esear
ch
er
s
to
d
ev
elo
p
ASR
?
D
if
f
er
en
t
ty
p
es
o
f
AI
tech
n
iq
u
es
h
a
v
e
b
ee
n
ex
p
lo
r
e
d
s
o
f
ar
to
m
o
d
el
s
p
ee
c
h
r
ec
o
g
n
itio
n
s
y
s
tem
s
.
Star
tin
g
with
m
ac
h
in
e
lea
r
n
in
g
ar
ch
itectu
r
e
lik
e
th
e
SVM,
k
-
n
ea
r
est
n
eig
h
b
o
r
s
(
KNN)
[
9
5
]
–
[
9
7
]
,
to
t
h
e
well
-
k
n
o
wn
d
ee
p
lear
n
in
g
ar
ch
itectu
r
es
lik
e
n
eu
r
al
n
etw
o
r
k
s
[
6
]
,
C
NN,
R
NN,
L
STM
,
SOM,
R
B
F,
T
DN
N
[
2
]
,
tr
an
s
f
o
r
m
er
s
[
2
9
]
,
SVM
-
o
n
e
ag
ain
s
t
o
n
e
tec
h
n
iq
u
es,
an
d
SVM
-
o
n
e
ag
ai
n
s
t
all
tech
n
iq
u
es
[
2
]
.
T
h
er
e
ar
e
also
h
y
b
r
id
m
ix
tu
r
es
lik
e
th
e
Evaluation Warning : The document was created with Spire.PDF for Python.