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
4
,
No
.
5
,
Octo
b
er
2
0
2
5
,
p
p
.
4
2
5
0
~
4
2
5
9
I
SS
N:
2
2
5
2
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
14
.i
5
.
p
p
4
2
5
0
-
4
2
5
9
4250
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
I
ntent
det
e
ction i
n AI chat
bo
ts:
a
c
o
mprehens
iv
e re
v
iew of
techniqu
es a
nd
th
e role of
ex
ternal
kno
wledg
e
J
em
ima
h K
.
1
,
R
a
j
k
u
m
a
r
K
a
nn
a
n
1
,
F
re
deric
An
dres
2
1
R
e
s
e
a
r
c
h
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
,
B
i
s
h
o
p
H
e
b
e
r
C
o
l
l
e
g
e
(
A
u
t
o
)
,
A
f
f
i
l
i
a
t
e
d
t
o
B
h
a
r
a
t
h
i
d
a
sa
n
U
n
i
v
e
r
si
t
y
,
Tr
i
c
h
y
,
I
n
d
i
a
2
D
i
g
i
t
a
l
C
o
n
t
e
n
t
a
n
d
M
e
d
i
a
S
c
i
e
n
c
e
s
R
e
se
a
r
c
h
D
i
v
i
si
o
n
,
N
a
t
i
o
n
a
l
I
n
st
i
t
u
t
e
o
f
I
n
f
o
r
ma
t
i
c
s
,
T
o
k
y
o
,
Ja
p
a
n
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
n
2
4
,
2
0
2
4
R
ev
is
ed
J
u
l 7
,
2
0
2
5
Acc
ep
ted
Au
g
6
,
2
0
2
5
Artifi
c
ial
in
telli
g
e
n
c
e
(AI)
c
h
a
tb
o
t
s
h
a
v
e
b
e
c
o
m
e
e
ss
e
n
ti
a
l
a
c
ro
ss
v
a
rio
u
s
in
d
u
stries
,
i
n
c
lu
d
in
g
c
u
sto
m
e
r
se
rv
ice
,
h
e
a
lt
h
c
a
re
,
e
d
u
c
a
t
io
n
,
a
n
d
e
n
terta
in
m
e
n
t
,
e
n
a
b
l
in
g
se
a
m
les
s
,
a
n
d
in
tell
ig
e
n
t
u
se
r
in
tera
c
ti
o
n
s.
A
k
e
y
c
o
m
p
o
n
e
n
t
o
f
c
h
a
tb
o
t
fu
n
c
ti
o
n
a
li
ty
is
in
ten
t
d
e
tec
ti
o
n
,
w
h
ich
d
e
term
in
e
s
th
e
u
n
d
e
rl
y
i
n
g
p
u
r
p
o
se
o
f
u
se
r
q
u
e
rie
s
to
p
ro
v
id
e
re
lev
a
n
t
re
sp
o
n
se
s.
T
ra
d
it
io
n
a
l
in
ten
t
d
e
tec
ti
o
n
m
e
th
o
d
s,
s
u
c
h
a
s
ru
le
-
b
a
se
d
a
n
d
sta
ti
stica
l
a
p
p
ro
a
c
h
e
s,
o
f
ten
stru
g
g
le
with
a
d
a
p
tab
i
li
ty
,
e
sp
e
c
ially
i
n
c
o
m
p
le
x
,
d
y
n
a
m
ic
c
o
n
v
e
rsa
ti
o
n
s.
Th
is
re
v
iew
e
x
a
m
in
e
s
t
h
e
e
v
o
lu
t
i
o
n
o
f
i
n
ten
t
d
e
tec
ti
o
n
tec
h
n
iq
u
e
s,
fro
m
e
a
rly
m
e
th
o
d
s
to
m
o
d
e
rn
d
e
e
p
lea
r
n
in
g
a
n
d
k
n
o
wle
d
g
e
-
e
n
ric
h
e
d
m
o
d
e
ls.
It
in
tro
d
u
c
e
s
th
e
d
o
m
a
in
ty
p
e
-
c
o
n
v
e
rsa
ti
o
n
t
u
rn
s
-
a
d
a
p
ti
v
i
ty
-
e
x
ter
n
a
l
k
n
o
wle
d
g
e
(
DCA
D)
c
las
sifica
ti
o
n
,
h
ig
h
li
g
h
t
in
g
it
s
sig
n
ifi
c
a
n
c
e
i
n
imp
ro
v
in
g
c
h
a
t
b
o
t
a
c
c
u
ra
c
y
a
n
d
c
o
n
tex
t
u
a
l
a
wa
re
n
e
ss
.
Th
e
p
a
p
e
r
c
a
teg
o
rize
s
e
x
isti
n
g
i
n
ten
t
d
e
tec
ti
o
n
m
o
d
e
ls,
a
n
a
ly
z
e
s
th
e
ir
a
p
p
li
c
a
ti
o
n
s
a
c
ro
ss
v
a
rio
u
s
se
c
to
rs,
a
n
d
d
isc
u
ss
e
s
k
e
y
c
h
a
ll
e
n
g
e
s,
i
n
c
lu
d
in
g
d
a
ta
in
teg
ra
ti
o
n
,
lan
g
u
a
g
e
a
m
b
i
g
u
it
y
,
a
n
d
e
th
ica
l
c
o
n
c
e
rn
s.
By
e
x
p
l
o
rin
g
e
m
e
rg
in
g
tren
d
s
a
n
d
f
u
t
u
re
d
irec
t
io
n
s,
t
h
is
re
v
iew
u
n
d
e
rsc
o
re
s
th
e
c
rit
ica
l
ro
le
o
f
e
x
tern
a
l
k
n
o
wle
d
g
e
i
n
e
n
h
a
n
c
in
g
c
h
a
tb
o
t
p
e
rfo
rm
a
n
c
e
a
n
d
u
se
r
e
x
p
e
rien
c
e
.
K
ey
w
o
r
d
s
:
Dialo
g
u
e
s
y
s
tem
E
x
ter
n
al
k
n
o
wled
g
e
I
n
ten
t d
etec
tio
n
Natu
r
al
lan
g
u
ag
e
p
r
o
ce
s
s
in
g
Natu
r
al
lan
g
u
ag
e
u
n
d
er
s
tan
d
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
:
R
ajk
u
m
ar
Kan
n
an
R
esear
ch
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
Scien
ce
,
B
is
h
o
p
Heb
e
r
C
o
lleg
e
(
Au
to
)
Af
f
iliated
to
B
h
ar
ath
id
asan
U
n
iv
er
s
ity
T
r
ich
y
,
I
n
d
ia
E
m
ail:
r
ajk
u
m
ar
@
b
h
c.
e
d
u
.
in
1.
I
NT
RO
D
UCT
I
O
N
I
n
t
o
d
a
y
’
s
d
i
g
i
t
a
l
e
r
a
,
c
h
a
t
b
o
ts
h
a
v
e
e
m
e
r
g
e
d
a
s
a
r
e
m
a
r
k
a
b
l
e
e
v
o
l
u
t
i
o
n
i
n
h
u
m
a
n
-
c
o
m
p
u
t
e
r
i
n
t
e
r
a
c
t
i
o
n
,
b
r
i
d
g
i
n
g
t
h
e
g
a
p
b
e
t
w
e
e
n
t
e
c
h
n
o
l
o
g
y
a
n
d
c
o
m
m
u
n
i
c
a
t
i
o
n
.
T
h
e
s
e
c
o
n
v
e
r
s
a
ti
o
n
a
l
a
g
e
n
t
s
,
p
o
w
e
r
e
d
b
y
a
r
t
i
f
i
c
i
a
l
i
n
t
e
l
li
g
e
n
c
e
(
A
I
)
,
h
a
v
e
t
r
a
n
s
i
t
i
o
n
e
d
f
r
o
m
s
i
m
p
l
e
r
u
le
-
b
a
s
e
d
s
y
s
t
e
m
s
t
o
s
o
p
h
is
t
i
ca
t
e
d
t
o
o
l
s
c
a
p
a
b
l
e
o
f
u
n
d
e
r
s
t
a
n
d
i
n
g
,
l
e
a
r
n
i
n
g
,
a
n
d
r
e
s
p
o
n
d
i
n
g
t
o
h
u
m
a
n
q
u
e
r
i
e
s
w
i
t
h
u
n
p
r
e
c
e
d
e
n
t
e
d
p
r
e
c
i
s
i
o
n
.
H
o
w
e
v
e
r
,
w
i
t
h
a
d
v
a
n
c
e
m
e
n
t
s
i
n
n
at
u
r
a
l
la
n
g
u
a
g
e
p
r
o
c
e
s
s
i
n
g
(
N
L
P
)
,
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
(
M
L
)
,
a
n
d
n
e
u
r
a
l
n
e
two
r
k
s
,
c
h
a
t
b
o
t
s
h
a
v
e
u
n
d
e
r
g
o
n
e
a
r
e
v
o
l
u
t
i
o
n
a
r
y
t
r
a
n
s
f
o
r
m
a
t
i
o
n
[
1
]
.
F
i
g
u
r
e
1
s
h
o
w
s
t
h
e
o
r
i
g
i
n
,
a
d
v
a
n
c
e
m
e
n
t
s
,
a
n
d
t
h
e
t
r
a
n
s
f
o
r
m
a
t
iv
e
i
m
p
a
c
t
o
f
c
h
a
t
b
o
t
s
,
t
r
a
ci
n
g
t
h
e
i
r
j
o
u
r
n
e
y
f
r
o
m
r
u
d
i
m
e
n
t
a
r
y
d
i
a
l
o
g
u
e
s
y
s
t
e
m
s
t
o
i
n
t
e
l
li
g
e
n
t
c
o
n
v
e
r
s
a
t
i
o
n
a
l
a
g
e
n
ts
r
es
h
a
p
i
n
g
t
h
e
d
i
g
i
ta
l w
o
r
l
d
.
F
r
o
m
a
i
d
i
n
g
c
u
s
t
o
m
e
r
s
e
r
v
i
c
e
t
o
e
n
h
a
n
c
i
n
g
e
d
u
c
a
t
i
o
n
a
n
d
s
t
r
e
a
m
l
i
n
i
n
g
h
e
a
l
t
h
c
a
r
e
,
c
h
a
t
b
o
t
s
a
r
e
n
o
w
a
c
o
r
n
e
r
s
t
o
n
e
o
f
u
s
e
r
e
n
g
a
g
e
m
e
n
t
,
t
r
a
n
s
f
o
r
m
i
n
g
u
s
e
r
i
n
t
e
r
a
c
ti
o
n
s
a
c
r
o
s
s
v
a
r
i
o
u
s
f
i
e
ld
s
[
2
]
.
T
h
e
e
v
o
l
u
ti
o
n
o
f
i
n
t
e
n
t
d
e
t
e
c
t
i
o
n
i
n
c
h
a
t
b
o
ts
h
a
s
m
a
r
k
e
d
a
s
i
g
n
i
f
i
c
a
n
t
d
e
p
a
r
t
u
r
e
f
r
o
m
e
a
r
l
y
r
u
l
e
-
b
a
s
e
d
s
y
s
t
e
m
s
,
s
h
i
f
t
i
n
g
t
o
w
a
r
d
s
s
o
p
h
i
s
ti
c
a
t
e
d
,
A
I
-
d
r
i
v
e
n
m
o
d
e
l
s
.
I
n
i
ti
a
ll
y
,
c
h
a
t
b
o
ts
o
p
e
r
a
t
e
d
o
n
s
i
m
p
l
e
,
f
i
x
e
d
r
u
l
e
s
a
n
d
r
e
l
i
e
d
o
n
k
e
y
w
o
r
d
m
at
ch
i
n
g
,
l
i
m
i
ti
n
g
t
h
e
i
r
f
l
e
x
i
b
i
l
it
y
a
n
d
d
e
p
t
h
o
f
u
n
d
e
r
s
ta
n
d
i
n
g
[
3
]
.
H
o
w
e
v
e
r
,
a
d
v
a
n
c
e
s
i
n
N
L
P a
n
d
M
L
h
a
v
e
r
e
v
o
l
u
tio
n
i
z
e
d
t
h
i
s
p
r
o
c
e
s
s
,
e
n
a
b
l
i
n
g
c
h
a
t
b
o
t
s
t
o
i
n
t
e
r
p
r
e
t
m
o
r
e
c
o
m
p
l
e
x
a
n
d
n
u
a
n
c
e
d
u
s
e
r
q
u
e
r
i
e
s
wi
t
h
h
i
g
h
e
r
a
c
c
u
r
a
c
y
[
4
]
.
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
I
n
ten
t d
etec
tio
n
i
n
A
I
c
h
a
tb
o
ts
:
a
co
mp
r
eh
e
n
s
ive
r
ev
iew
o
f te
ch
n
iq
u
es a
n
d
…
(
Je
mima
h
K.
)
4251
Fig
u
r
e
1
.
E
v
o
lu
tio
n
o
f
c
h
atb
o
t
tech
n
o
lo
g
y
T
h
is
r
ev
iew
e
x
am
in
es a
r
an
g
e
o
f
in
ten
t
d
etec
tio
n
tech
n
i
q
u
es,
f
r
o
m
tr
a
d
itio
n
al
ap
p
r
o
ac
h
es t
o
th
e
latest
d
ee
p
lear
n
in
g
m
eth
o
d
s
.
R
u
le
-
b
ased
s
y
s
tem
s
,
alth
o
u
g
h
s
im
p
le
an
d
ea
s
y
to
im
p
lem
en
t,
o
f
ten
s
tr
u
g
g
le
with
s
ca
lab
ilit
y
an
d
ten
d
to
f
alter
wh
en
h
an
d
lin
g
am
b
ig
u
o
u
s
q
u
er
ies.
Stati
s
tical
ap
p
r
o
ac
h
es
s
u
ch
as
b
ag
-
of
-
wo
r
d
s
(
B
o
W
)
an
d
ter
m
f
r
eq
u
e
n
cy
-
in
v
er
s
e
d
o
cu
m
en
t
f
r
e
q
u
en
c
y
(
T
F
-
I
DF)
r
ep
r
esen
ted
a
s
ig
n
if
ica
n
t
ad
v
an
ce
m
en
t
b
y
u
tili
zin
g
p
r
o
b
ab
ilis
tic
tech
n
iq
u
es
to
en
h
a
n
ce
class
if
icatio
n
ac
cu
r
ac
y
[
5
]
.
W
ith
th
e
r
is
e
o
f
ML
m
eth
o
d
s
lik
e
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
es
(
SVM)
an
d
d
ec
is
io
n
tr
ee
s
,
in
ten
t
class
if
icat
io
n
s
aw
f
u
r
th
er
en
h
an
ce
m
en
ts
.
Dee
p
lear
n
in
g
tech
n
iq
u
es
h
a
v
e
b
r
o
u
g
h
t
t
h
e
m
o
s
t
s
u
b
s
tan
tial
ad
v
an
ce
s
in
in
te
n
t
d
etec
tio
n
.
Neu
r
al
n
etwo
r
k
ar
ch
itectu
r
es,
esp
ec
ially
r
ec
u
r
r
en
t
n
eu
r
al
n
etwo
r
k
s
(
R
NNs)
an
d
th
eir
v
ar
ian
ts
lik
e
l
o
n
g
s
h
o
r
t
-
ter
m
m
em
o
r
y
(
L
STM
)
n
etwo
r
k
s
[
6
]
,
[
7
]
,
h
a
v
e
s
h
o
wn
ex
ce
p
tio
n
al
ca
p
ab
ili
ty
in
h
an
d
lin
g
s
eq
u
en
tial
d
ata
an
d
u
n
d
er
s
tan
d
i
n
g
co
n
tex
t.
T
r
an
s
f
o
r
m
e
r
-
b
ased
m
o
d
els,
s
u
ch
as
b
id
i
r
ec
tio
n
al
en
co
d
er
r
e
p
r
esen
tatio
n
s
f
r
o
m
tr
a
n
s
f
o
r
m
er
s
(
B
E
R
T
)
[
8
]
,
[
9
]
,
an
d
g
en
e
r
ativ
e
p
r
e
-
tr
ain
ed
tr
an
s
f
o
r
m
er
(
GPT)
,
r
ep
r
esen
t
th
e
lead
i
n
g
e
d
g
e
i
n
NL
P,
p
r
o
v
id
in
g
u
n
p
ar
alleled
ac
c
u
r
ac
y
a
n
d
a
d
a
p
tab
ilit
y
f
o
r
i
n
ten
t d
etec
tio
n
ta
s
k
s
[
1
0
]
.
B
y
ex
p
lo
r
i
n
g
t
h
e
ap
p
licatio
n
s
o
f
in
ten
t
d
etec
tio
n
,
th
is
r
e
v
iew
u
n
d
er
s
co
r
es
th
e
b
r
o
ad
im
p
ac
t
o
f
ch
at
b
o
ts
ac
r
o
s
s
d
if
f
er
en
t
s
ec
to
r
s
.
I
n
cu
s
to
m
er
s
er
v
ice,
ch
atb
o
ts
en
h
an
c
e
u
s
er
ex
p
er
ien
ce
b
y
o
f
f
er
in
g
t
im
ely
an
d
ac
cu
r
ate
r
esp
o
n
s
es.
I
n
h
ea
lth
ca
r
e
,
th
e
y
ass
is
t
in
p
r
elim
in
ar
y
d
iag
n
o
s
tics
an
d
p
atien
t
in
ter
ac
tio
n
[
1
1
]
.
I
n
ed
u
ca
tio
n
,
ch
atb
o
ts
f
ac
ilit
ate
p
er
s
o
n
alize
d
lear
n
in
g
ex
p
er
ien
ce
s
,
wh
ile
i
n
en
ter
tain
m
en
t,
th
e
y
en
g
a
g
e
u
s
er
s
with
in
ter
ac
tiv
e
s
to
r
y
tellin
g
an
d
cu
s
to
m
ized
r
ec
o
m
m
en
d
atio
n
s
[
1
2
]
.
Desp
ite
th
ese
ad
v
an
ce
m
en
ts
,
in
te
n
t
d
etec
tio
n
f
ac
es
p
er
s
is
ten
t
ch
allen
g
es.
I
s
s
u
es
s
u
ch
as
am
b
ig
u
it
y
in
u
s
er
q
u
er
ies,
d
iv
er
s
e
lan
g
u
ag
e
v
ar
iatio
n
s
,
an
d
th
e
n
ee
d
f
o
r
o
n
g
o
in
g
ad
a
p
tatio
n
p
r
esen
t sig
n
if
ican
t o
b
s
tacle
s
[
1
3
]
–
[
1
5
]
.
W
h
ile
v
ar
io
u
s
tech
n
iq
u
es
f
o
r
in
ten
t
d
etec
tio
n
h
av
e
b
ee
n
d
ev
elo
p
ed
,
a
u
n
if
ied
p
e
r
s
p
e
ctiv
e
th
at
ca
teg
o
r
izes
th
ese
m
eth
o
d
s
b
ased
o
n
ess
en
tial
f
ea
tu
r
es
-
s
u
ch
as
d
o
m
ain
s
p
ec
if
icity
,
co
n
v
er
s
atio
n
al
tu
r
n
s
,
ad
ap
tiv
ity
,
a
n
d
k
n
o
wled
g
e
in
t
eg
r
atio
n
-
is
s
till
lack
in
g
.
T
h
is
r
ev
iew
in
tr
o
d
u
ce
s
th
e
d
o
m
ain
ty
p
e
-
co
n
v
er
s
atio
n
tu
r
n
s
-
ad
ap
tiv
ity
-
ex
ter
n
al
k
n
o
wled
g
e
(
DC
AD)
class
if
icatio
n
m
o
d
el,
a
f
r
am
ewo
r
k
th
at
o
r
g
a
n
izes
ch
atb
o
t
m
o
d
els
b
y
th
ese
co
r
e
f
ea
t
u
r
es.
DC
AD
aim
s
to
o
f
f
er
a
s
tr
u
ct
u
r
ed
ap
p
r
o
ac
h
to
ev
alu
atin
g
an
d
ca
teg
o
r
izin
g
in
t
en
t
d
etec
tio
n
tech
n
iq
u
es
ac
r
o
s
s
d
i
v
er
s
e
ap
p
licatio
n
s
.
T
h
is
f
r
am
e
wo
r
k
n
o
t
o
n
ly
s
im
p
lifie
s
th
e
p
r
o
ce
s
s
o
f
co
m
p
ar
in
g
m
o
d
els
b
u
t
also
h
ig
h
lig
h
ts
p
atter
n
s
an
d
em
er
g
i
n
g
tr
en
d
s
,
wh
ich
ar
e
ess
en
tial
f
o
r
r
esear
ch
er
s
an
d
p
r
ac
titi
o
n
er
s
lo
o
k
in
g
to
d
e
v
elo
p
o
r
r
ef
i
n
e
c
h
atb
o
t sy
s
tem
s
.
T
h
e
r
e
s
t
o
f
t
h
e
p
ap
e
r
i
s
s
t
r
u
c
t
u
r
ed
a
s
f
o
l
l
o
w
s
.
I
n
s
e
c
t
i
o
n
2
,
th
e
r
ev
i
e
w
m
e
t
h
o
d
o
l
o
g
y
f
o
r
in
t
en
t
d
e
t
e
c
t
io
n
b
a
s
ed
D
C
A
D
c
l
a
s
s
i
f
i
c
a
t
i
o
n
o
f
ch
a
t
b
o
t
s
i
s
d
i
s
c
u
s
s
e
d
.
I
n
s
e
c
t
io
n
3
,
v
a
r
io
u
s
t
e
c
h
n
i
q
u
e
s
t
h
a
t
a
r
e
u
s
e
d
i
n
ch
a
t
b
o
t
m
o
d
e
l
s
ar
e
ex
p
l
ai
n
ed
w
i
t
h
r
e
s
ea
r
ch
p
a
p
er
s
.
I
n
s
e
c
ti
o
n
4
,
t
h
e
r
e
s
u
l
t
s
an
d
d
i
s
c
u
s
s
i
o
n
s
e
c
t
io
n
c
o
m
p
ar
e
t
h
e
r
e
s
u
l
t
s
o
f
th
e
e
x
i
s
t
i
n
g
m
e
t
h
o
d
o
lo
g
i
e
s
a
n
d
a
l
s
o
d
i
s
c
u
s
s
e
s
t
h
e
c
h
a
l
le
n
g
e
s
,
a
d
v
a
n
ta
g
e
s
,
l
i
m
i
t
a
t
io
n
s
a
n
d
i
m
p
l
i
c
a
t
io
n
s
in
i
n
t
en
t
d
e
t
e
c
t
io
n
,
f
o
l
lo
w
ed
b
y
t
h
e
d
a
ta
s
e
t
t
h
a
t
i
s
u
s
e
d
in
t
h
e
r
e
s
e
ar
ch
p
ap
e
r
s
w
e
r
e
d
i
s
c
u
s
s
e
d
.
F
i
n
a
l
l
y
,
i
n
s
e
c
t
i
o
n
5
,
th
e
co
n
cl
u
s
i
o
n
i
s
m
ad
e
.
2.
RE
VI
E
W
M
E
T
H
O
DO
L
O
G
Y
I
n
te
n
t
d
et
ec
t
io
n
i
n
c
h
at
b
o
ts
c
a
n
b
e
an
al
y
z
e
d
ac
r
o
s
s
f
o
u
r
k
e
y
asp
ec
ts
:
d
o
m
ai
n
ty
p
e,
c
o
n
v
e
r
s
a
tio
n
a
l
ty
p
e,
ad
ap
ti
v
it
y
,
a
n
d
k
n
o
wl
ed
g
e
ty
p
e.
C
o
ll
ec
ti
v
el
y
r
ef
e
r
r
ed
to
as
t
h
e
DC
AD
c
lass
i
f
i
ca
t
io
n
,
t
h
es
e
a
s
p
e
cts
a
r
e
ess
e
n
ti
al
to
t
h
e
i
n
t
en
t
d
ete
cti
o
n
p
r
o
ce
s
s
wit
h
i
n
ch
at
b
o
t
te
ch
n
o
l
o
g
y
.
As
d
e
p
i
cte
d
i
n
Fi
g
u
r
e
2
,
t
h
e
cl
ass
if
icat
io
n
t
r
e
e
o
u
tl
in
es
1
6
ca
t
eg
o
r
i
es
b
as
e
d
o
n
th
ese
f
e
atu
r
e
co
m
b
i
n
at
io
n
s
,
f
a
cil
ita
ti
n
g
ea
s
y
c
ate
g
o
r
iza
ti
o
n
o
f
ch
at
b
o
t
m
o
d
e
ls
.
P
r
i
m
a
r
il
y
T
ur
ni
ng
T
e
st
E
l
i
z
a
P
ar
r
y
J
ab
be
r
wac
k
y
D
r
.S
B
A
IT
S
O
A
LIC
E
S
m
ar
t
e
r
Chi
l
d
IB
M
's W
at
so
n
S
i
r
i
G
o
o
g
l
e
N
o
w
C
o
r
t
an
a(
M
i
c
r
o
so
f
t
)
A
l
exa
(
A
m
az
on)
M
e
s
s
e
n
g
e
r
B
o
t
s
J
as
pe
r
A
I
C
ha
t
G
P
T
B
ar
d
G
o
o
g
l
e
G
e
m
i
ni
M
i
c
r
o
so
f
t
C
o
pi
l
o
t
195
0
197
2
199
2
200
1
201
1
201
4
201
6
202
2
202
3
Var
i
o
us
S
t
ag
e
s
o
f
C
ha
t
bo
t
Adv
an
c
e
m
e
nt
s
Y
e
ar
s
o
f
A
dv
an
c
e
m
e
nt
E
v
ol
ut
i
on
of
C
hat
bot
T
e
chnol
ogy
(1950
-
2023)
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
4
,
No
.
5
,
Octo
b
er
2
0
2
5
:
4
2
5
0
-
4
2
5
9
4252
u
s
e
d
as
a
r
e
v
i
ew
ap
p
r
o
a
ch
,
th
e
DC
AD
cl
ass
if
ic
ati
o
n
p
r
o
v
i
d
es
r
ea
d
e
r
s
w
it
h
i
n
s
i
g
h
t
i
n
t
o
h
o
w
ex
is
ti
n
g
m
o
d
e
ls
ar
e
s
tr
u
ct
u
r
e
d
a
r
o
u
n
d
s
p
e
ci
f
ic
f
ea
t
u
r
es.
E
ac
h
o
f
t
h
es
e
f
e
at
u
r
es
wi
ll
b
e
e
x
a
m
i
n
e
d
i
n
d
eta
il
in
t
h
e
f
o
ll
o
wi
n
g
s
ec
t
io
n
s
.
Fig
u
r
e
2
.
I
n
ten
t
d
etec
tio
n
b
ase
d
DC
AD
class
if
icatio
n
tr
ee
2
.
1
.
Do
m
a
in t
y
pe
in DCA
D
cla
s
s
if
ica
t
io
n
C
h
atb
o
ts
co
m
e
u
n
d
er
two
t
y
p
es
o
f
d
o
m
ain
s
,
o
n
e
is
o
p
en
d
o
m
ain
an
d
an
o
th
er
is
clo
s
ed
d
o
m
ain
o
r
d
o
m
ain
s
p
ec
i
f
ic.
Mo
s
t
o
f
th
e
e
ar
ly
r
esear
ch
was
m
ain
ly
b
ase
d
o
n
clo
s
ed
d
o
m
ai
n
wh
ich
b
ec
am
e
a
d
r
awb
ac
k
f
o
r
th
e
ch
atb
o
t
co
m
m
u
n
ity
.
T
h
is
f
ea
tu
r
e
will
h
elp
to
f
in
d
ch
atb
o
ts
wh
ich
f
alls
in
to
th
e
o
p
en
d
o
m
ain
an
d
d
o
m
ain
s
p
ec
if
ic
ca
teg
o
r
ies.
2
.
2
.
Co
nv
er
s
a
t
io
n t
y
pe
in DCA
D
cla
s
s
if
ica
t
io
n
Ma
n
y
p
er
s
o
n
alize
d
ch
atb
o
ts
p
r
o
v
id
e
a
m
u
lti
-
tu
r
n
co
n
v
er
s
ati
o
n
al
ex
p
er
ien
ce
to
th
e
u
s
er
.
W
h
er
ea
s
,
a
m
in
o
r
ity
o
f
b
o
ts
ar
e
u
s
ed
as
in
f
o
r
m
atio
n
r
et
r
iev
al
ag
en
ts
wh
ich
g
iv
e
a
s
in
g
le
-
tu
r
n
c
o
n
v
e
r
s
atio
n
al
ex
p
er
ien
ce
to
th
e
u
s
er
.
R
ec
en
t a
d
v
a
n
ce
m
en
t
in
ch
atb
o
t te
ch
n
o
lo
g
y
h
as f
o
cu
ed
o
n
t
h
is
f
ea
tu
r
e
o
f
m
u
lti
-
tu
r
n
co
n
v
er
s
atio
n
.
2
.
3
.
Ada
ptiv
it
y
t
y
pe
in DCA
D
cla
s
s
if
ica
t
io
n
An
ad
ap
tiv
e
b
o
t
is
s
o
m
eth
in
g
th
at
u
n
d
er
s
tan
d
s
a
n
d
wo
r
k
s
ac
co
r
d
in
g
to
th
e
u
s
er
s
’
n
ee
d
s
an
d
k
ee
p
s
tr
ac
k
o
f
th
e
h
is
to
r
y
co
n
tex
t
o
f
t
h
e
u
s
er
s
’
r
eq
u
est.
On
ly
p
er
s
o
n
alize
d
b
o
ts
ca
n
in
h
er
it
ad
ap
tiv
i
ty
.
A
n
o
n
-
ad
a
p
tiv
e
b
o
t
wo
r
k
s
o
n
its
o
wn
p
ath
a
n
d
d
o
es
n
o
t
d
e
v
iate
ac
co
r
d
in
g
to
u
s
er
s
’
r
e
q
u
est
wh
ich
is
a
m
ajo
r
f
law
in
a
p
er
s
o
n
alize
d
ch
at
b
o
t.
User
ex
p
er
ien
ce
is
a
v
er
y
cr
u
cial
r
o
le
in
wh
ich
ad
a
p
tiv
ity
p
lay
s
an
i
m
p
o
r
tan
t
r
o
le.
2
.
4
.
K
no
wledg
e
t
y
pe
in DCA
D
cla
s
s
if
ica
t
io
n
So
m
e
ch
atb
o
ts
ar
e
s
u
p
p
o
r
ted
with
an
in
f
o
r
m
atio
n
b
ase
o
r
k
n
o
wled
g
e
b
ase
wh
ich
co
n
tain
s
in
f
o
r
m
atio
n
b
ased
o
n
s
p
ec
if
ic
to
p
ics.
T
h
u
s
,
th
e
in
f
o
r
m
atio
n
r
etr
iev
al
h
ap
p
en
s
b
ased
o
n
th
e
k
n
o
wled
g
e
p
r
o
v
id
ed
.
I
n
co
n
tr
ast,
s
o
m
e
b
o
ts
d
o
n
o
t
u
s
e
a
s
ep
ar
ate
k
n
o
wled
g
e
b
ase;
in
s
tead
,
th
ey
wo
r
k
with
a
p
r
e
-
tr
ain
e
d
d
ata
m
o
d
el
wh
ich
class
if
ies th
e
r
eq
u
est with
in
th
e
tr
ain
ed
f
ea
tu
r
es.
3.
T
E
CH
N
I
Q
UE
S
I
N
I
NT
E
N
T
DE
T
E
C
T
I
O
N
I
n
ten
t
d
etec
tio
n
is
tr
ad
itio
n
al
ly
ap
p
r
o
ac
h
ed
as
a
s
en
ten
ce
-
lev
el
class
if
icatio
n
p
r
o
b
lem
,
aim
in
g
to
id
en
tify
th
e
u
n
d
er
ly
in
g
in
ten
t
b
eh
in
d
a
u
s
er
'
s
u
tter
an
ce
with
in
th
e
r
ea
lm
o
f
NL
P
.
A
to
tal
o
f
5
6
r
esear
ch
ar
ticles
o
n
in
ten
t
d
etec
tio
n
wer
e
s
tu
d
i
ed
an
d
r
ev
iewe
d
.
T
h
is
s
ec
tio
n
p
r
esen
ts
a
s
u
m
m
ar
y
o
f
s
elec
ted
wo
r
k
s
,
co
v
e
r
in
g
b
o
th
tr
ad
itio
n
al
ap
p
r
o
ac
h
es a
n
d
r
ec
en
t a
d
v
an
ce
m
e
n
ts
in
in
te
n
t d
etec
tio
n
.
3
.
1
.
O
v
er
v
iew
o
f
inte
nt
det
ec
t
io
n a
pp
ro
a
ches
B
ef
o
r
e
2
0
2
0
,
in
ten
t
d
etec
tio
n
r
esear
ch
p
r
im
ar
ily
r
elied
o
n
tr
ad
itio
n
al
ML
tech
n
iq
u
es,
in
clu
d
in
g
k
-
n
ea
r
est
n
eig
h
b
o
r
s
(
KNN)
,
SVM,
an
d
r
an
d
o
m
f
o
r
est
(
R
F)
.
I
n
co
n
tr
ast,
th
is
ar
ticle
em
p
h
asizes
r
ec
en
t
ad
v
an
ce
m
e
n
ts
th
at
u
tili
ze
d
e
ep
lear
n
in
g
an
d
tr
a
n
s
f
o
r
m
er
-
b
ased
m
o
d
els
f
o
r
i
m
p
r
o
v
ed
p
er
f
o
r
m
an
ce
.
I
t
also
ad
d
r
ess
es
m
u
ltil
in
g
u
al
in
ten
t
d
etec
tio
n
wh
ich
u
s
es
m
u
ltil
in
g
u
al
tex
t
d
ata
[
1
6
]
.
No
tab
ly
,
c
ateg
o
r
y
C
1
4
,
wh
ich
in
clu
d
es
d
o
m
ain
-
s
p
ec
if
ic
co
m
b
in
atio
n
al
f
ea
tu
r
es
f
o
r
in
te
n
t
d
etec
tio
n
,
s
h
o
ws
a
h
ig
h
n
u
m
b
er
o
f
s
u
cc
ess
f
u
l
m
o
d
els,
as
s
h
o
wn
in
Fig
u
r
e
3
.
T
h
is
in
d
icate
s
th
at
d
o
m
ain
-
s
p
e
cif
ic
f
ea
tu
r
es
p
la
y
a
cr
u
cial
r
o
l
e
in
en
h
an
cin
g
u
s
er
ex
p
er
ien
ce
f
o
r
in
te
n
t d
etec
tio
n
task
s
in
m
an
y
o
f
th
e
latest ch
atb
o
t m
o
d
els.
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
I
n
ten
t d
etec
tio
n
i
n
A
I
c
h
a
tb
o
ts
:
a
co
mp
r
eh
e
n
s
ive
r
ev
iew
o
f te
ch
n
iq
u
es a
n
d
…
(
Je
mima
h
K.
)
4253
3
.
1
.
1
.
At
t
ent
io
n ba
s
ed
inte
nt
det
ec
t
io
n
A
u
n
if
ied
C
NN
-
b
ased
p
a
r
allel
ar
ch
itectu
r
e
with
c
r
o
s
s
-
f
u
s
io
n
an
d
m
ask
in
g
im
p
r
o
v
e
d
jo
i
n
t
i
n
ten
t
an
d
s
lo
t
p
r
ed
ictio
n
.
E
ar
ly
ap
p
r
o
a
ch
es
to
in
ten
t
class
if
icatio
n
f
o
r
d
ialo
g
u
e
u
tter
an
ce
s
e
m
p
lo
y
ed
a
B
o
W
m
o
d
el
co
m
b
in
ed
with
n
aïv
e
B
ay
es,
f
o
llo
wed
b
y
co
n
tin
u
o
u
s
-
B
o
W
in
teg
r
ated
with
SVM.
T
h
e
f
in
a
l
class
if
icatio
n
was
p
er
f
o
r
m
ed
u
s
in
g
L
STM
n
etwo
r
k
s
[
1
7
]
.
T
o
en
h
a
n
ce
n
atu
r
a
l
lan
g
u
ag
e
in
ter
p
r
etatio
n
f
o
r
cu
s
to
m
ized
d
ig
ital
ass
is
tan
ts
,
ad
ap
tiv
e
in
ten
t
d
etec
tio
n
in
m
u
lti
-
d
o
m
ain
e
n
v
ir
o
n
m
en
ts
(
Aid
Me
)
was
in
tr
o
d
u
ce
d
.
T
h
is
s
y
s
tem
em
p
lo
y
ed
a
s
em
an
tic
s
im
ilar
ity
ev
alu
atio
n
m
o
d
el
b
u
ilt
o
n
in
ten
t
d
ata
to
d
if
f
er
en
tiate
b
etwe
en
k
n
o
wn
in
ten
ts
an
d
au
to
n
o
m
o
u
s
ly
id
en
tif
y
n
e
w
v
ar
iatio
n
s
o
f
e
x
is
tin
g
in
ten
t
s
[
1
8
]
.
A
m
u
ltimo
d
al
m
o
d
el
f
o
r
in
ten
t
d
etec
tio
n
an
d
s
lo
t
f
illi
n
g
was
p
r
o
p
o
s
ed
,
lev
er
ag
in
g
h
is
to
r
ic
al
co
n
tex
t
an
d
e
x
ter
n
al
k
n
o
wled
g
e
with
i
n
th
e
f
r
am
ewo
r
k
o
f
s
p
o
k
e
n
la
n
g
u
ag
e
u
n
d
er
s
tan
d
in
g
(
SLU
)
.
T
h
is
m
o
d
el
ca
ter
ed
to
s
p
ec
if
ic
task
ca
teg
o
r
ies
[
1
9
]
.
An
o
p
en
-
d
o
m
ai
n
ch
atb
o
t
was
tr
ain
ed
en
d
-
to
-
en
d
,
m
a
p
p
in
g
i
n
p
u
t
ta
g
s
to
o
p
tim
ized
o
u
tp
u
t
s
eq
u
e
n
ce
ta
g
s
.
T
h
is
ch
at
b
o
t
d
em
o
n
s
tr
ated
th
e
a
b
ilit
y
to
p
er
f
o
r
m
co
m
m
o
n
-
s
en
s
e
r
ea
s
o
n
i
n
g
d
u
r
in
g
h
u
m
a
n
in
ter
ac
tio
n
s
[
2
0
]
.
I
n
th
e
b
an
k
in
g
s
ec
to
r
,
a
r
ea
l
-
ti
m
e
ch
atb
o
t
was
im
p
lem
en
ted
to
p
r
o
m
o
te
s
u
s
tain
ab
le
d
ev
elo
p
m
en
t
b
y
en
h
an
cin
g
cu
s
to
m
er
s
atis
f
ac
ti
o
n
.
T
h
is
s
y
s
tem
id
en
tifie
d
u
s
er
s
’
co
n
tin
u
an
ce
in
ten
tio
n
s
b
ased
o
n
s
atis
f
ac
tio
n
,
tr
u
s
t,
an
d
p
e
r
ce
iv
ed
u
s
ef
u
ln
es
s
,
with
tr
u
s
t
ex
er
ti
n
g
th
e
s
tr
o
n
g
est
in
f
lu
en
ce
[
2
1
]
.
A
co
n
tr
as
tiv
e
lear
n
in
g
-
b
ased
task
ad
ap
tatio
n
m
o
d
el
was
p
r
o
p
o
s
ed
f
o
r
f
ew
-
s
h
o
t
in
ten
t
r
ec
o
g
n
itio
n
,
u
s
in
g
a
co
n
tr
asti
v
e
lo
s
s
to
en
h
an
ce
ca
teg
o
r
y
s
ep
ar
atio
n
an
d
ex
tr
ac
t ta
s
k
-
s
p
ec
if
ic
f
ea
tu
r
es with
in
m
eta
-
task
s
[
2
2
]
.
T
h
e
ch
u
n
k
-
le
v
e
l
i
n
t
e
n
t
d
ete
c
tio
n
(
C
L
I
D)
f
r
am
ew
o
r
k
u
s
es
a
s
li
d
i
n
g
wi
n
d
o
w
-
b
as
ed
s
e
l
f
-
att
en
ti
o
n
(
SW
SA)
m
e
c
h
a
n
is
m
to
d
e
te
ct
ch
u
n
k
-
l
e
v
el
i
n
te
n
ts
a
n
d
i
n
t
en
t
tr
a
n
s
i
ti
o
n
s
,
s
e
g
m
e
n
ti
n
g
u
tt
er
an
ce
s
a
n
d
p
r
ed
ict
in
g
s
u
b
-
u
tt
e
r
a
n
c
e
i
n
te
n
ts
ef
f
e
cti
v
e
l
y
[
2
3
]
.
A
m
u
lti
-
t
u
r
n
d
ial
o
g
i
n
t
en
t
d
ete
cti
o
n
m
o
d
el
was
d
e
v
e
l
o
p
e
d
,
i
n
co
r
p
o
r
ati
n
g
h
is
t
o
r
ic
al
i
n
f
o
r
m
a
ti
o
n
.
S
em
an
ti
c
d
ata
f
r
o
m
b
o
t
h
u
s
e
r
s
an
d
s
y
s
t
em
s
w
er
e
co
m
b
i
n
e
d
u
s
i
n
g
a
s
el
f
-
att
en
ti
o
n
n
etw
o
r
k
,
wh
i
le
a
n
L
S
T
M
c
ap
t
u
r
ed
h
is
t
o
r
ic
al
c
o
n
te
x
t
[
2
4
]
.
A
m
eta
-
l
ea
r
n
i
n
g
f
r
a
m
ew
o
r
k
f
o
r
ze
r
o
-
s
h
o
t
i
n
te
n
t
class
if
ica
ti
o
n
was
p
r
ese
n
te
d
,
u
ti
lizi
n
g
a
h
y
b
r
id
att
en
ti
o
n
m
e
c
h
a
n
is
m
t
h
a
t m
er
g
es
d
is
tr
ib
u
t
io
n
al
s
i
g
n
at
u
r
e
a
tte
n
ti
o
n
wi
th
m
u
lt
i
-
lay
e
r
p
e
r
c
ep
tr
o
n
(
M
L
P
)
-
b
as
e
d
atte
n
t
io
n
t
o
ef
f
e
cti
v
e
ly
c
a
p
t
u
r
e
p
att
er
n
s
o
f
k
ey
w
o
r
d
o
cc
u
r
r
e
n
ce
s
[
2
5
]
.
F
in
all
y
,
a
m
u
ltit
as
k
lea
r
n
i
n
g
a
p
p
r
o
a
ch
w
as
p
r
o
p
o
s
e
d
f
o
r
m
u
lt
ili
n
g
u
al
i
n
te
n
t
d
e
te
cti
o
n
an
d
s
l
o
t
f
ill
in
g
,
h
an
d
l
in
g
i
n
p
u
ts
in
E
n
g
l
is
h
,
B
en
g
ali
,
an
d
Hi
n
d
i
u
s
in
g
m
u
lti
li
n
g
u
al
wo
r
d
em
b
ed
d
i
n
g
s
(
MWE
)
t
o
j
o
i
n
t
ly
m
o
d
e
l
b
o
th
t
ask
s
a
cr
o
s
s
lan
g
u
a
g
es
[
1
6
]
.
Fig
u
r
e
3
.
C
h
atb
o
t
ca
teg
o
r
ies
with
co
m
b
in
atio
n
al
f
ea
tu
r
es
3
.
1
.
2
.
Bi
-
direct
io
na
l m
o
dels
us
ed
in i
nte
nt
det
ec
t
io
n
A
co
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
-
b
id
ir
ec
tio
n
al
g
ated
r
ec
u
r
r
e
n
t
u
n
it
(
C
NN
-
B
GR
U
)
m
o
d
el
lev
er
ag
in
g
wo
r
d
em
b
e
d
d
in
g
s
was
p
r
o
p
o
s
ed
f
o
r
ef
f
ec
tiv
e
in
ten
t
class
if
icatio
n
in
d
ialo
g
u
e
s
y
s
tem
s
[
3
0
]
.
E
m
o
tio
n
an
d
in
ten
tio
n
ar
e
two
m
ajo
r
f
ac
to
r
s
wh
ich
in
f
lu
en
ce
p
eo
p
les’
co
m
m
u
n
icatio
n
in
th
ei
r
ev
er
y
d
a
y
life
.
A
h
ier
ar
ch
ical
v
ar
iatio
n
al
m
o
d
el
was
d
ev
elo
p
ed
to
p
r
ed
ict
th
e
e
m
o
tio
n
s
a
n
d
in
ten
ts
in
s
eq
u
e
n
ce
an
d
is
u
s
ed
to
co
n
v
ey
th
e
r
esp
o
n
s
e
ac
co
r
d
in
g
to
t
h
e
p
r
ed
icted
em
o
tio
n
in
an
o
p
e
n
d
o
m
ain
s
y
s
tem
[
3
1
]
.
A
two
-
s
tag
e
tr
ip
let
tr
ain
i
n
g
f
r
am
ewo
r
k
u
s
in
g
r
atio
-
m
em
o
r
y
ce
llu
lar
n
o
n
lin
ea
r
n
etwo
r
k
(
R
MCN
N
)
f
o
r
s
tr
u
ctu
r
al
f
ea
tu
r
es
an
d
a
Siam
ese
B
E
R
T
en
co
d
er
f
o
r
in
te
n
t
d
et
ec
tio
n
im
p
r
o
v
ed
m
u
lticlas
s
class
if
icatio
n
[
3
2
]
.
A
m
u
lti
-
task
SLU
m
o
d
el
with
b
id
ir
ec
tio
n
al
lo
n
g
s
h
o
r
t
-
ter
m
m
em
o
r
y
n
etwo
r
k
(
B
iLST
M
)
/GR
U
an
d
atten
tio
n
m
ec
h
an
is
m
s
en
h
an
ce
d
in
ten
t
d
etec
tio
n
,
s
lo
t f
illi
n
g
,
a
n
d
d
ial
o
g
u
e
ac
t c
lass
if
icatio
n
with
s
tatis
tically
v
alid
ated
r
esu
lts
[
3
3
]
.
3
.
1
.
3
.
M
em
o
ry
net
wo
r
k
ba
s
e
d inte
nt
det
ec
t
io
n
T
h
e
m
em
o
r
y
n
etwo
r
k
with
p
a
tter
n
s
(
MN
P)
m
o
d
el
im
p
r
o
v
e
s
u
p
o
n
tr
ad
itio
n
al
m
em
o
r
y
n
e
two
r
k
s
b
y
in
co
r
p
o
r
atin
g
two
d
is
tin
ct
m
e
m
o
r
y
c
o
m
p
o
n
en
ts
.
O
n
e
d
ed
ica
ted
to
ca
p
t
u
r
in
g
co
n
tex
t
u
al
in
f
o
r
m
atio
n
f
r
o
m
ea
r
lier
d
ialo
g
u
e
t
u
r
n
s
,
a
n
d
a
n
o
th
er
d
esig
n
ed
to
s
to
r
e
p
atter
n
s
i
d
en
tifie
d
b
y
r
eg
u
la
r
ex
p
r
ess
io
n
s
.
T
h
is
d
u
al
-
m
em
o
r
y
ar
ch
itectu
r
e
allo
ws
th
e
m
o
d
e
l
to
r
ep
r
esen
t
in
ten
t
p
atter
n
s
m
o
r
e
ef
f
ec
tiv
ely
th
r
o
u
g
h
its
s
p
ec
ialized
p
atter
n
m
em
o
r
y
,
en
h
a
n
cin
g
its
o
v
e
r
all
ca
p
ab
ilit
y
to
d
ete
r
m
in
e
u
s
er
i
n
ten
t
[
3
4
]
.
3
.
1
.
4
.
G
ra
ph
net
wo
rk
ba
s
ed
inte
nt
det
ec
t
io
n
A
d
ee
p
lear
n
in
g
m
o
d
el
u
s
in
g
h
ier
ar
ch
ical
B
iLST
M
wi
th
atten
tio
n
was
p
r
o
p
o
s
ed
f
o
r
tex
t
r
e
p
r
esen
tatio
n
an
d
s
im
ilar
ity
,
en
h
a
n
ce
d
b
y
a
n
ex
ter
n
al
k
n
o
wled
g
e
g
r
a
p
h
t
o
r
etr
iev
e
in
f
o
r
m
atio
n
r
elev
a
n
t
to
u
s
er
q
u
e
r
ies
[
8
]
,
0
5
10
Num
b
e
r
o
f
P
a
pe
r
s
DC
A
D
c
l
a
s
s
i
f
i
c
a
t
i
o
n
c
a
t
e
go
r
i
e
s
Cha
t
bot
c
a
t
e
g
ori
e
s
unde
r D
CA
D
c
l
a
s
s
i
fi
c
a
t
i
on
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
4
,
No
.
5
,
Octo
b
er
2
0
2
5
:
4
2
5
0
-
4
2
5
9
4254
[
3
5
]
.
A
r
ei
n
f
o
r
ce
m
en
t
lear
n
i
n
g
-
b
ased
ap
p
r
o
ac
h
was
in
tr
o
d
u
ce
d
to
in
ter
p
r
et
i
n
ten
t
in
m
u
lt
i
-
tu
r
n
d
ialo
g
u
es
b
y
id
en
tify
in
g
p
ath
s
in
a
k
n
o
wled
g
e
g
r
a
p
h
,
u
s
in
g
a
Ma
r
k
o
v
d
ec
is
io
n
p
r
o
ce
s
s
to
g
en
e
r
ate
in
ter
p
r
etab
le
in
ten
t
tr
ajec
to
r
ies
[
3
6
]
.
A
co
-
o
cc
u
r
r
e
n
ce
-
en
h
a
n
ce
d
E
R
NI
E
n
etwo
r
k
was
p
r
o
p
o
s
ed
f
o
r
co
m
p
lex
i
n
ten
t
r
ec
o
g
n
itio
n
in
m
ed
ical
q
u
er
ies,
lev
er
ag
in
g
c
o
n
ce
p
t
co
-
o
c
cu
r
r
e
n
ce
p
atter
n
s
m
in
ed
u
s
in
g
th
e
Ap
r
io
r
i
alg
o
r
ith
m
to
im
p
r
o
v
e
co
n
tex
tu
al
u
n
d
e
r
s
tan
d
in
g
[
3
7
]
.
A
s
y
m
p
to
m
atic
ass
ess
m
en
t
ch
atb
o
t
(
SAC
)
was
d
ev
el
o
p
ed
to
r
aise
awa
r
en
ess
o
f
b
r
ea
s
t
ca
n
ce
r
s
y
m
p
to
m
s
in
w
o
m
en
an
d
to
f
ac
ilit
ate
f
o
r
m
al
m
ed
ical
co
n
s
u
ltatio
n
s
,
u
tili
zin
g
a
h
o
s
p
ital
-
b
ased
k
n
o
wled
g
e
b
ase
f
r
o
m
T
aiwa
n
[
3
8
]
.
Mu
lti
-
s
ca
le
d
y
n
am
ic
co
n
v
o
l
u
tio
n
al
n
etwo
r
k
(
M
-
D
C
N
)
,
a
m
u
lti
-
s
ca
le
d
y
n
am
ic
c
o
n
v
o
lu
tio
n
al
n
etwo
r
k
,
en
h
a
n
ce
s
k
n
o
wled
g
e
g
r
ap
h
em
b
ed
d
in
g
b
y
m
o
d
elin
g
c
o
m
p
lex
r
elatio
n
s
u
s
in
g
d
y
n
am
ic
f
ilter
s
an
d
alter
n
atin
g
en
tity
-
r
elatio
n
co
m
p
o
s
itio
n
s
,
o
u
tp
e
r
f
o
r
m
in
g
b
aselin
e
m
o
d
els
ac
r
o
s
s
f
iv
e
lin
k
p
r
ed
ictio
n
b
en
ch
m
a
r
k
d
atasets
[
3
9
]
.
3
.
1
.
5
.
L
o
w
-
re
s
o
urce
ba
s
ed
m
et
ho
ds
us
e
d in int
ent
det
ec
t
i
o
n
A
two
-
s
tag
e
in
ten
t
d
etec
tio
n
m
o
d
el
was
d
esig
n
e
d
f
o
r
h
an
d
l
in
g
co
m
p
lex
u
s
er
u
tter
a
n
ce
s
,
i
n
teg
r
atin
g
s
en
ten
ce
co
m
p
r
ess
io
n
an
d
in
te
n
t
class
if
icatio
n
with
in
a
m
u
lti
-
task
lear
n
in
g
f
r
am
ewo
r
k
[
4
0
]
.
An
in
ten
t
d
etec
tio
n
an
d
en
tity
ex
tr
ac
tio
n
s
y
s
tem
was
p
r
o
p
o
s
ed
f
o
r
h
ea
lth
ca
r
e
q
u
e
r
ies
in
I
n
d
ic
lan
g
u
a
g
es.
T
h
e
p
r
im
ar
y
co
n
tr
ib
u
tio
n
in
clu
d
es
th
e
c
r
ea
tio
n
o
f
a
m
u
lti
lin
g
u
al
h
ea
lth
ca
r
e
d
ataset
co
m
p
r
is
in
g
r
ea
l
-
wo
r
ld
h
o
s
p
ital
q
u
e
r
ies
in
E
n
g
lis
h
a
n
d
s
ix
I
n
d
ic
l
an
g
u
ag
es
(
Hin
d
i,
B
e
n
g
ali,
T
am
il,
T
elu
g
u
,
Ma
r
at
h
i,
an
d
Gu
jar
ati)
,
a
n
n
o
tated
with
b
o
th
q
u
er
y
i
n
ten
ts
an
d
en
titi
es
[
4
1
]
.
A
ze
r
o
a
n
d
f
ew
-
s
h
o
t
in
ten
t
i
d
en
tific
atio
n
was
p
r
o
p
o
s
ed
u
s
in
g
lar
g
e
lan
g
u
ag
e
m
o
d
els
(
L
L
Ms)
.
I
t
d
em
o
n
s
tr
ated
f
o
u
r
ap
p
r
o
ac
h
e
s
n
am
ely
d
o
m
ain
a
d
ap
tatio
n
,
d
ata
au
g
m
en
tatio
n
,
ze
r
o
-
s
h
o
t
p
r
ed
ictio
n
with
p
r
o
m
p
tin
g
an
d
p
a
r
am
eter
-
ef
f
icien
t
f
in
e
-
tu
n
in
g
.
Z
er
o
-
s
h
o
t
an
d
f
ew
-
s
h
o
t
in
ten
t
p
r
ed
ictio
n
co
m
p
letely
r
em
o
v
e
o
r
s
u
b
s
tan
tially
r
ed
u
ce
th
e
wo
r
k
to
p
r
o
v
id
e
in
ten
t
-
u
tter
an
ce
s
,
r
esp
ec
tiv
ely
[
4
2
]
.
An
in
ten
t
a
n
d
en
tity
d
etec
tio
n
ap
p
r
o
ac
h
wer
e
d
ev
el
o
p
ed
f
o
r
a
m
en
tal
h
ea
lth
v
ir
tu
al
ass
is
tan
t
ch
atb
o
t,
in
c
o
r
p
o
r
atin
g
d
a
ta
au
g
m
en
tatio
n
t
o
au
to
m
atica
lly
g
e
n
er
ate
tr
ai
n
in
g
d
ata.
T
h
e
m
o
d
el
jo
in
tly
ad
d
r
ess
ed
in
ten
t
a
n
d
en
tity
d
ete
ctio
n
b
y
lev
e
r
ag
in
g
th
eir
in
t
er
d
ep
e
n
d
en
cies,
r
at
h
er
th
an
tr
ea
tin
g
t
h
em
as sep
ar
ate
class
if
icatio
n
task
s
[
4
3
]
.
A
C
K
-
Ker
as
m
o
d
el
u
s
in
g
W
o
r
d
2
Vec
em
b
ed
d
in
g
s
ac
h
iev
e
d
th
e
b
est
F1
-
s
co
r
e
f
o
r
wo
r
d
-
lev
e
l
lan
g
u
ag
e
id
en
tific
atio
n
in
Kan
n
ad
a
-
E
n
g
lis
h
co
d
e
-
m
ix
e
d
Yo
u
T
u
b
e
c
o
m
m
en
ts
.
T
h
is
s
tu
d
y
ex
p
lo
r
e
s
Kan
n
ad
a
-
E
n
g
lis
h
wo
r
d
-
lev
el
lan
g
u
ag
e
id
en
tific
a
tio
n
u
s
in
g
ch
ar
ac
ter
-
a
n
d
wo
r
d
-
b
ased
f
ea
tu
r
es
in
ML
an
d
d
e
ep
lear
n
in
g
m
o
d
els
[
4
4
]
.
T
o
ad
d
r
ess
b
r
o
ad
e
r
co
d
e
-
m
ix
in
g
is
s
u
es,
a
h
ier
ar
ch
ical
tr
an
s
f
o
r
m
er
(
HI
T
)
m
o
d
el
with
m
u
lti
-
h
ea
d
e
d
s
elf
-
atten
tio
n
an
d
o
u
ter
p
r
o
d
u
ct
atten
tio
n
was
in
tr
o
d
u
ce
d
,
o
u
tp
er
f
o
r
m
in
g
ex
is
tin
g
m
u
ltil
in
g
u
al
m
o
d
els
ac
r
o
s
s
m
u
ltip
le
lan
g
u
ag
es
a
n
d
task
s
[
4
5
]
,
[
4
6
]
.
Fu
r
th
er
m
o
r
e,
a
Hin
d
i
-
E
n
g
lis
h
p
a
r
allel
co
r
p
u
s
a
n
d
e
n
h
an
ce
d
tr
an
s
latio
n
p
ip
elin
e
d
e
m
o
n
s
tr
ated
im
p
r
o
v
ed
m
ac
h
i
n
e
tr
an
s
latio
n
p
er
f
o
r
m
an
ce
f
o
r
c
o
d
e
-
m
i
x
ed
te
x
ts
,
o
f
f
er
in
g
s
ca
lab
ilit
y
to
o
th
er
lan
g
u
ag
e
p
air
s
[
4
7
]
.
L
in
g
u
is
tic
-
ad
ap
tiv
e
r
etr
iev
al
-
au
g
m
en
tatio
n
(
L
AR
A
)
a
f
r
am
ewo
r
k
ad
d
r
ess
es
m
u
lti
-
tu
r
n
in
ten
t
class
if
icatio
n
b
y
in
teg
r
atin
g
a
f
in
e
-
tu
n
ed
m
o
d
el
with
r
etr
iev
al
-
au
g
m
en
te
d
L
L
Ms,
en
ab
lin
g
co
n
tex
t
-
awa
r
e
p
r
ed
ictio
n
s
ac
r
o
s
s
s
ix
lan
g
u
ag
es with
o
u
t e
x
te
n
s
iv
e
r
etr
ai
n
in
g
[
4
8
]
.
T
o
o
v
er
co
m
e
d
ata
s
ca
r
city
in
m
en
tal
h
ea
lth
d
ialo
g
u
e
s
y
s
tem
s
,
th
e
s
in
g
le
-
t
u
r
n
to
m
u
lti
-
tu
r
n
in
cl
u
s
iv
e
lan
g
u
ag
e
e
x
p
an
s
io
n
(
SMI
L
E
)
t
ec
h
n
iq
u
e
g
en
er
ates
m
u
lti
-
tu
r
n
co
n
v
er
s
atio
n
s
f
r
o
m
s
in
g
le
-
tu
r
n
d
ata
u
s
in
g
C
h
atGPT
.
T
h
is
led
to
th
e
cr
ea
tio
n
o
f
th
e
SMI
L
E
C
HAT
d
ataset
an
d
th
e
d
e
v
elo
p
m
e
n
t o
f
th
e
Me
C
h
at
ch
atb
o
t,
v
alid
ate
d
with
a
r
ea
l
-
life
co
u
n
s
elin
g
d
ataset
[
4
9
]
.
3.
2
.
I
s
s
ues
a
dd
re
s
s
ed
a
nd
s
o
lutio
ns
pro
po
s
e
d in int
ent
de
t
ec
t
io
n
3
.
2
.
1
.
Da
t
a
s
ca
rc
it
y
A
g
e
n
e
r
a
l
i
z
e
d
z
e
r
o
-
s
h
o
t
(
G
Z
S)
c
l
a
s
s
i
f
i
c
at
i
o
n
m
o
d
e
l
w
as
i
n
tr
o
d
u
c
e
d
t
o
a
d
d
r
e
s
s
t
h
e
p
r
o
b
l
em
o
f
d
a
t
a
s
c
a
r
c
it
y
f
o
r
u
n
s
e
e
n
i
n
t
e
n
t
s
b
y
in
c
o
r
p
o
r
a
t
i
n
g
c
o
m
m
o
n
s
e
n
s
e
k
n
o
w
l
e
d
g
e
i
n
t
o
t
h
e
i
n
t
e
n
t
d
e
t
ec
t
io
n
f
r
a
m
e
w
o
r
k
[
1
3
]
.
T
h
i
s
a
p
p
r
o
a
c
h
e
n
a
b
l
es
t
h
e
m
o
d
e
l
t
o
b
e
t
te
r
u
n
d
e
r
s
t
a
n
d
a
n
d
c
la
s
s
i
f
y
u
s
e
r
i
n
t
e
n
ts
,
e
v
e
n
w
h
e
n
i
t
h
a
s
n
o
t
e
n
c
o
u
n
t
e
r
e
d
s
p
e
c
i
f
i
c
e
x
a
m
p
l
e
s
d
u
r
i
n
g
t
r
ai
n
i
n
g
.
A
d
d
i
t
i
o
n
a
l
l
y
,
a
n
a
u
t
o
m
at
ic
i
n
t
e
n
t
-
s
l
o
t
i
n
d
u
c
t
i
o
n
m
e
t
h
o
d
w
a
s
p
r
o
p
o
s
e
d
a
s
a
d
o
m
a
i
n
-
i
n
d
e
p
e
n
d
e
n
t
t
e
c
h
n
i
q
u
e
t
h
a
t
u
s
es
r
o
l
e
l
a
b
e
l
i
n
g
,
c
o
n
c
e
p
t
m
i
n
i
n
g
,
a
n
d
p
a
t
t
e
r
n
m
i
n
i
n
g
t
o
i
d
e
n
t
i
f
y
b
o
t
h
b
r
o
a
d
i
n
t
e
n
t
r
o
l
es
a
n
d
m
o
r
e
s
p
e
c
i
f
i
c
a
b
s
t
r
a
c
t
c
o
n
c
e
p
ts
,
m
a
k
i
n
g
it
a
d
a
p
t
a
b
l
e
t
o
n
ew
d
o
m
a
i
n
s
[
1
4
]
.
T
h
e
s
e
a
d
v
a
n
c
e
m
e
n
t
s
h
e
l
p
i
m
p
r
o
v
e
t
h
e
f
l
e
x
i
b
i
li
t
y
a
n
d
r
o
b
u
s
t
n
e
s
s
o
f
i
n
t
e
n
t
d
e
t
ec
t
i
o
n
s
y
s
te
m
s
a
c
r
o
s
s
a
v
a
r
i
e
t
y
o
f
a
p
p
l
i
c
a
t
i
o
n
a
r
e
as
.
3
.
2
.
2
.
E
m
er
g
ing
inte
nt
C
o
n
t
e
x
t
u
a
li
z
e
d
D
e
e
p
S
e
m
S
p
a
ce
w
a
s
i
n
t
r
o
d
u
c
e
d
f
o
r
i
n
t
e
n
t
d
e
t
e
c
t
i
o
n
b
y
g
e
n
e
r
a
t
i
n
g
s
y
n
s
e
t
v
ec
t
o
r
s
w
i
t
h
W
o
r
d
N
e
t
3
.
1
a
n
d
c
o
n
v
e
r
t
i
n
g
w
o
r
d
s
f
r
o
m
i
n
t
e
n
t
d
a
t
as
e
ts
i
n
to
c
o
n
t
e
x
t
u
a
l
i
z
e
d
s
e
m
a
n
t
i
c
v
e
ct
o
r
s
.
T
h
i
s
a
p
p
r
o
a
ch
f
u
n
c
t
i
o
n
s
i
n
a
m
a
n
n
e
r
s
i
m
i
l
a
r
to
d
e
e
p
c
o
n
t
e
x
t
u
a
l
e
m
b
e
d
d
i
n
g
s
s
u
c
h
a
s
B
E
R
T
,
e
m
b
e
d
d
i
n
g
s
f
r
o
m
l
a
n
g
u
a
g
e
m
o
d
e
l
(
ELMo
)
,
a
n
d
G
P
T
-
3
,
a
l
l
o
wi
n
g
f
o
r
a
r
i
c
h
e
r
r
e
p
r
e
s
e
n
t
a
t
i
o
n
o
f
w
o
r
d
m
e
a
n
i
n
g
s
i
n
c
o
n
t
e
x
t
[
1
0
]
.
A
d
d
i
t
i
o
n
a
l
l
y
,
an
i
n
c
r
e
m
e
n
t
a
l
i
n
t
e
n
t
d
e
t
e
c
t
i
o
n
m
e
t
h
o
d
w
a
s
p
r
o
p
o
s
e
d
f
o
r
t
h
e
m
e
d
i
c
a
l
d
o
m
a
i
n
u
s
i
n
g
C
o
n
t
r
as
t
i
v
e
R
e
p
l
a
y
N
e
tw
o
r
k
s
,
w
h
i
c
h
e
n
a
b
l
e
t
h
e
m
o
d
e
l
t
o
l
e
a
r
n
n
e
w
i
n
t
e
n
t
s
f
r
o
m
u
s
e
r
q
u
e
r
i
e
s
w
h
i
l
e
e
f
f
e
c
ti
v
e
l
y
h
a
n
d
l
i
n
g
d
at
a
i
m
b
a
l
a
n
c
e
a
n
d
r
a
r
e
m
e
d
i
c
a
l
te
r
m
i
n
o
l
o
g
y
[
1
5
]
.
T
h
e
s
e
i
n
n
o
v
a
t
i
o
n
s
e
n
h
a
n
c
e
t
h
e
a
d
a
p
t
a
b
i
l
it
y
t
o
e
m
e
r
g
i
n
g
i
n
t
e
n
ts
a
n
d
p
e
r
f
o
r
m
a
n
c
e
o
f
i
n
t
e
n
t
d
e
t
e
ct
i
o
n
s
y
s
t
e
m
s
,
p
a
r
t
icu
l
a
r
l
y
i
n
s
p
e
ci
a
l
iz
e
d
d
o
m
a
i
n
s
.
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
I
n
ten
t d
etec
tio
n
i
n
A
I
c
h
a
tb
o
ts
:
a
co
mp
r
eh
e
n
s
ive
r
ev
iew
o
f te
ch
n
iq
u
es a
n
d
…
(
Je
mima
h
K.
)
4255
3
.
2
.
3
.
I
m
ba
la
nced
da
t
a
C
las
s
life
lo
n
g
lear
n
in
g
-
b
ased
in
ten
t
d
etec
tio
n
(
C
L
L
-
I
D)
en
ab
les
co
n
tin
u
al
lear
n
in
g
o
f
n
ew
in
ten
t
class
es
wh
ile
m
itig
atin
g
ca
tast
r
o
p
h
ic
f
o
r
g
ettin
g
,
em
p
lo
y
in
g
s
tr
u
ctu
r
e
-
b
ased
r
etr
o
s
p
ec
ti
o
n
a
n
d
c
o
n
tr
asti
v
e
k
n
o
wled
g
e
d
is
till
atio
n
to
ad
d
r
ess
cla
s
s
im
b
alan
ce
[
5
0
]
.
Sp
ec
ialized
B
E
R
T
m
o
d
els
tr
ain
ed
o
n
m
ed
ical
co
r
p
o
r
a
o
u
tp
er
f
o
r
m
ed
g
en
er
al
B
E
R
T
in
b
io
m
ed
ical
task
s
.
T
h
e
m
o
d
el
s
wer
e
ev
alu
ated
o
n
b
io
m
ed
ica
l
d
atasets
,
in
clu
d
in
g
in
f
o
r
m
atics f
o
r
in
teg
r
atin
g
b
io
lo
g
y
an
d
th
e
b
ed
s
id
e,
an
d
d
r
u
g
-
d
r
u
g
in
ter
ac
tio
n
d
atasets
[
5
1
]
.
T
ab
le
1
d
escr
ib
es
th
e
is
s
u
es
ad
d
r
ess
ed
an
d
th
e
m
eth
o
d
o
l
o
g
y
u
s
ed
to
o
v
er
co
m
e
th
e
is
s
u
es
f
r
o
m
v
ar
io
u
s
co
n
tr
ib
u
tio
n
s
u
n
d
e
r
ea
ch
DC
AD
class
if
icat
io
n
with
r
esp
ec
t
to
v
ar
io
u
s
ca
teg
o
r
ies
f
r
o
m
1
to
1
6
.
T
h
e
ta
b
le
h
ig
h
lig
h
ts
o
n
ly
th
e
latest
co
n
tr
ib
u
tio
n
s
m
a
d
e
in
ea
c
h
ca
teg
o
r
iey
,
wh
ich
r
ef
lects
th
e
m
o
s
t
r
e
liab
le
m
o
d
el
b
ased
o
n
th
e
f
ea
tu
r
es
in
DC
AD
c
las
s
if
icatio
n
to
d
esig
n
an
d
d
ep
lo
y
a
s
u
cc
ess
f
u
l
ch
atb
o
t.
T
h
e
m
o
s
t
r
ec
en
t
wo
r
k
s
ar
e
u
n
d
er
ca
teg
o
r
y
1
,
1
0
an
d
1
3
wh
ich
d
ir
ec
tly
s
h
o
ws
th
e
ch
atb
o
t
m
o
d
els’
r
ec
en
t
wo
r
k
s
ar
e
b
as
ed
o
n
ad
ap
tab
ilit
y
as o
n
e
o
f
its
im
p
o
r
tan
t f
ea
tu
r
e
s
.
T
ab
le
1
.
Me
th
o
d
o
lo
g
y
a
n
d
ad
d
r
ess
ed
is
s
u
es o
n
ea
ch
ca
teg
o
r
y
P
a
p
e
r
A
d
d
r
e
ss
e
d
i
ss
u
e
s
A
p
p
r
o
a
c
h
e
s
C
1
-
o
p
e
n
d
o
m
a
i
n
,
s
i
n
g
l
e
-
t
u
r
n
,
a
d
a
p
t
i
v
e
a
n
d
p
r
e
-
t
r
a
i
n
e
d
i
n
f
o
r
m
a
t
i
o
n
O
r
h
a
n
e
t
a
l
.
[
1
0
]
i
n
c
r
e
me
n
t
a
l
l
y
i
n
f
e
r
r
i
n
g
n
e
w
i
n
t
e
n
t
f
r
o
m
u
t
t
e
r
a
n
c
e
s
o
u
t
si
d
e
t
h
e
e
x
i
s
t
i
n
g
s
e
t
D
e
e
p
c
o
n
t
r
a
s
t
i
v
e
c
l
u
s
t
e
r
i
n
g
a
l
g
o
r
i
t
h
m
C
2
-
o
p
e
n
d
o
m
a
i
n
,
s
i
n
g
l
e
-
t
u
r
n
,
a
d
a
p
t
i
v
e
a
n
d
k
n
o
w
l
e
d
g
e
g
r
a
p
h
S
i
d
d
i
q
u
e
e
t
a
l
.
[
1
3
]
t
r
a
i
n
i
n
g
d
a
t
a
s
c
a
r
c
i
t
y
o
f
u
n
s
e
e
n
i
n
t
e
n
t
s
R
e
l
a
t
i
o
n
s
h
i
p
me
t
a
-
f
e
a
t
u
r
e
c
o
mp
u
t
a
t
i
o
n
a
n
d
i
n
t
e
g
r
a
t
e
s
c
o
m
mo
n
se
n
se
k
n
o
w
l
e
d
g
e
C
3
-
o
p
e
n
d
o
m
a
i
n
,
s
i
n
g
l
e
-
t
u
r
n
,
n
o
n
-
a
d
a
p
t
i
v
e
a
n
d
p
r
e
-
t
r
a
i
n
e
d
i
n
f
o
A
b
r
o
e
t
a
l
.
[
9
]
r
e
d
u
c
e
t
h
e
n
e
e
d
f
o
r
mass
i
v
e
s
u
p
e
r
v
i
se
d
d
a
t
a
W
e
i
g
h
t
e
d
f
i
n
i
t
e
st
a
t
e
t
r
a
n
sd
u
c
e
r
(
WFST
)
-
B
ER
T
C
5
-
o
p
e
n
d
o
m
a
i
n
,
mu
l
t
i
-
t
u
r
n
,
a
d
a
p
t
i
v
e
a
n
d
p
r
e
-
t
r
a
i
n
e
d
i
n
f
o
La
i
r
e
t
a
l
.
[
1
8
]
l
e
a
r
n
i
n
g
n
e
w
i
n
t
e
n
t
s
i
n
c
u
s
t
o
mi
z
e
d
a
s
si
st
a
n
t
s
S
e
n
t
e
n
c
e
si
m
i
l
a
r
i
t
y
met
h
o
d
C6
-
o
p
e
n
d
o
m
a
i
n
,
mu
l
t
i
-
t
u
r
n
,
a
d
a
p
t
i
v
e
a
n
d
k
n
o
w
l
e
d
g
e
g
r
a
p
h
W
a
n
g
e
t
a
l
.
[
1
9
]
Li
mi
t
e
d
a
b
i
l
i
t
y
o
f
S
LU
mo
d
e
l
s
t
r
a
i
n
e
d
o
n
l
a
r
g
e
-
sca
l
e
c
o
r
p
u
s
t
o
u
n
d
e
r
st
a
n
d
m
u
l
t
i
-
t
u
r
n
d
i
a
l
o
g
u
e
H
i
st
o
r
y
d
i
a
l
o
g
u
e
e
n
c
o
d
e
r
a
n
d
k
n
o
w
l
e
d
g
e
a
t
t
e
n
t
i
o
n
C7
-
o
p
e
n
d
o
m
a
i
n
,
mu
l
t
i
-
t
u
r
n
,
n
o
n
-
a
d
a
p
t
i
v
e
a
n
d
p
r
e
-
t
r
a
i
n
e
d
i
n
f
o
R
e
n
a
n
d
X
u
[
3
2
]
En
h
a
n
c
e
t
h
e
a
b
i
l
i
t
y
o
f
d
e
t
e
c
t
i
n
g
i
n
t
e
n
t
f
r
o
m
S
LU
u
si
n
g
Tr
i
p
l
e
t
t
r
a
i
n
i
n
g
a
n
d
f
u
s
i
o
n
st
r
a
t
e
g
y
R
M
C
N
N
a
s
S
i
a
m
e
se
e
n
c
o
d
e
r
s
a
n
d
B
ER
T,
ELM
o
a
n
d
G
P
T
f
o
r
u
t
t
e
r
a
n
c
e
r
e
p
r
e
s
e
n
t
a
t
i
o
n
C8
-
o
p
e
n
d
o
m
a
i
n
,
mu
l
t
i
-
t
u
r
n
,
n
o
n
-
a
d
a
p
t
i
v
e
a
n
d
k
n
o
w
l
e
d
g
e
g
r
a
p
h
Y
a
n
g
e
t
a
l
.
[
3
6
]
La
c
k
s
t
r
a
n
s
p
a
r
e
n
c
y
i
n
h
o
w
i
n
t
e
n
t
s
a
r
e
p
r
e
d
i
c
t
e
d
i
n
mu
l
t
i
-
t
u
r
n
d
i
a
l
o
g
u
e
s
P
o
l
i
c
y
-
g
u
i
d
e
d
r
e
i
n
f
o
r
c
e
me
n
t
l
e
a
r
n
i
n
g
a
n
d
M
a
r
k
o
v
D
e
c
i
s
i
o
n
P
r
o
c
e
ss
C
10
-
d
o
m
a
i
n
s
p
e
c
i
f
i
c
,
si
n
g
l
e
-
t
u
r
n
,
a
d
a
p
t
i
v
e
a
n
d
k
n
o
w
l
e
d
g
e
g
r
a
p
h
Zh
a
n
g
e
t
a
l
.
[
3
7
]
D
e
t
e
c
t
i
n
g
si
n
g
l
e
-
i
n
t
e
n
t
,
m
u
l
t
i
-
i
n
t
e
n
t
a
n
d
i
m
p
l
i
c
i
t
i
n
t
e
n
t
i
n
p
a
t
i
e
n
t
s’
u
t
t
e
r
a
n
c
e
ER
N
I
E
w
i
t
h
A
p
r
i
o
r
i
a
l
g
o
r
i
t
h
m
a
n
d
N
o
d
e
2
V
e
c
C
1
1
-
d
o
m
a
i
n
s
p
e
c
i
f
i
c
,
s
i
n
g
l
e
-
t
u
r
n
,
n
o
n
-
a
d
a
p
t
i
v
e
a
n
d
p
r
e
-
t
r
a
i
n
e
d
i
n
f
o
B
a
o
e
t
a
l
.
[
3
5
]
A
n
sw
e
r
c
o
m
p
l
e
x
me
d
i
c
a
l
q
u
e
st
i
o
n
s
H
i
e
r
a
r
c
h
i
c
a
l
B
i
LST
M
a
t
t
e
n
t
i
o
n
mo
d
e
l
(
H
B
A
M
)
w
i
t
h
k
n
o
w
l
e
d
g
e
g
r
a
p
h
C
1
3
-
d
o
m
a
i
n
s
p
e
c
i
f
i
c
,
mu
l
t
i
-
t
u
r
n
,
a
d
a
p
t
i
v
e
a
n
d
p
r
e
-
t
r
a
i
n
e
d
i
n
f
o
P
a
r
i
k
h
e
t
a
l
.
[
4
2
]
I
n
t
e
n
t
d
e
t
e
c
t
i
o
n
i
n
l
o
w
r
e
s
o
u
r
c
e
s
e
t
t
i
n
g
Ze
r
o
-
s
h
o
t
l
e
a
r
n
i
n
g
w
i
t
h
f
i
n
e
-
t
u
n
i
n
g
.
U
ses
d
o
m
a
i
n
a
d
a
p
t
a
t
i
o
n
a
n
d
d
a
t
a
a
u
g
m
e
n
t
a
t
i
o
n
C
14
-
d
o
m
a
i
n
s
p
e
c
i
f
i
c
,
mu
l
t
i
-
t
u
r
n
,
a
d
a
p
t
i
v
e
a
n
d
k
n
o
w
l
e
d
g
e
g
r
a
p
h
X
u
e
t
a
l
.
[
6
]
a
u
t
o
m
a
t
i
c
me
d
i
c
a
l
d
i
a
g
n
o
s
i
s
i
n
c
o
r
p
o
r
a
t
i
n
g
m
e
d
i
c
a
l
k
n
o
w
l
e
d
g
e
K
n
o
w
l
e
d
g
e
r
o
u
t
e
d
D
e
e
p
Q
-
n
e
t
w
o
r
k
4.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
e
DC
AD
class
if
icatio
n
f
r
am
ewo
r
k
o
f
f
er
s
a
s
tr
u
ctu
r
ed
a
p
p
r
o
ac
h
to
an
al
y
zin
g
in
ten
t
d
etec
tio
n
ac
r
o
s
s
d
iv
er
s
e
ch
atb
o
t
m
o
d
els.
I
t
ca
teg
o
r
izes
m
o
d
els
b
ased
o
n
f
o
u
r
k
ey
d
im
en
s
io
n
s
:
d
o
m
ain
s
p
ec
if
i
city
,
co
n
v
er
s
atio
n
al
tu
r
n
s
,
ad
a
p
tiv
ity
,
an
d
k
n
o
wle
d
g
e
in
te
g
r
atio
n
.
T
h
is
f
r
am
ew
o
r
k
f
ac
ilit
ates
a
n
u
an
ce
d
ev
al
u
atio
n
o
f
h
o
w
th
ese
f
ea
tu
r
es
im
p
ac
t
ch
atb
o
t
p
er
f
o
r
m
an
ce
an
d
ef
f
ec
tiv
en
ess
.
T
h
e
class
if
icat
io
n
r
esu
lts
ar
e
v
is
u
ally
r
ep
r
esen
ted
in
Fig
u
r
e
2
,
illu
s
tr
atin
g
1
6
d
is
tin
ct
ca
teg
o
r
ies,
ea
ch
c
o
r
r
esp
o
n
d
in
g
to
a
u
n
iq
u
e
co
m
b
i
n
atio
n
o
f
t
h
e
DC
AD
f
ea
tu
r
es.
4
.
1
.
F
re
qu
ency
a
na
ly
s
is
o
f
f
ea
t
ure
co
m
bin
a
t
io
ns
T
h
e
g
r
a
p
h
i
c
a
l
r
e
p
r
e
s
e
n
t
a
ti
o
n
i
n
d
i
c
a
t
e
s
t
h
at
c
e
r
t
a
i
n
f
e
a
t
u
r
e
co
m
b
i
n
a
t
i
o
n
s
,
p
a
r
t
i
c
u
la
r
l
y
d
o
m
a
i
n
-
s
p
e
c
i
f
i
c
m
o
d
e
l
s
w
i
t
h
i
n
t
e
g
r
a
te
d
k
n
o
w
le
d
g
e
a
n
d
a
d
a
p
t
i
v
e
c
a
p
a
b
il
i
ti
e
s
,
a
r
e
m
o
r
e
f
r
e
q
u
e
n
t
l
y
u
s
e
d
i
n
r
e
c
e
n
t
r
es
e
a
r
c
h
.
F
o
r
i
n
s
t
a
n
c
e
,
ca
t
e
g
o
r
y
C
1
4
,
w
h
i
c
h
i
n
c
l
u
d
e
s
d
o
m
a
i
n
-
s
p
ec
i
f
i
c
f
ea
t
u
r
e
s
w
it
h
e
n
h
a
n
c
e
d
k
n
o
w
l
e
d
g
e
in
t
e
g
r
a
t
i
o
n
,
a
c
c
o
u
n
t
s
f
o
r
a
s
i
g
n
i
f
i
c
a
n
t
p
o
r
t
i
o
n
o
f
s
u
c
c
e
s
s
f
u
l
m
o
d
e
l
s
.
T
h
i
s
h
i
g
h
l
i
g
h
ts
t
h
e
i
m
p
o
r
t
a
n
c
e
o
f
d
o
m
a
i
n
r
e
l
e
v
a
n
c
e
a
n
d
e
n
r
i
c
h
e
d
k
n
o
w
l
e
d
g
e
s
o
u
r
c
e
s
i
n
d
e
l
i
v
e
r
i
n
g
a
c
c
u
r
a
t
e
a
n
d
c
o
n
t
e
x
t
-
s
e
n
s
it
i
v
e
r
e
s
p
o
n
s
es
f
o
r
i
n
t
e
n
t
d
e
te
c
t
i
o
n
t
a
s
k
s
.
4.
2.
L
im
it
a
t
io
ns
a
nd
f
uture
direct
io
ns
W
h
i
l
e
t
h
e
D
C
A
D
f
r
a
m
ew
o
r
k
o
f
f
e
r
s
a
c
o
m
p
r
e
h
e
n
s
i
v
e
c
l
ass
i
f
i
c
a
t
i
o
n
a
p
p
r
o
a
c
h
,
t
h
e
r
e
a
r
e
l
i
m
i
t
a
t
i
o
n
s
t
o
c
o
n
s
i
d
e
r
.
S
o
m
e
f
e
a
t
u
r
e
c
o
m
b
in
a
t
i
o
n
s
,
p
a
r
ti
c
u
l
a
r
l
y
t
h
o
s
e
i
n
v
o
l
v
i
n
g
h
i
g
h
a
d
a
p
t
a
b
i
l
it
y
o
r
e
x
t
e
n
s
i
v
e
k
n
o
wl
e
d
g
e
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
4
,
No
.
5
,
Octo
b
er
2
0
2
5
:
4
2
5
0
-
4
2
5
9
4256
i
n
t
e
g
r
a
t
i
o
n
,
m
a
y
r
e
q
u
i
r
e
s
i
g
n
if
i
c
a
n
t
c
o
m
p
u
t
at
i
o
n
a
l
r
es
o
u
r
c
es
a
n
d
a
d
v
a
n
c
e
d
i
n
f
r
a
s
t
r
u
c
t
u
r
e.
F
u
r
t
h
e
r
m
o
r
e
,
t
h
e
f
r
a
m
e
w
o
r
k
c
u
r
r
e
n
t
l
y
r
e
l
i
es
o
n
a
s
t
a
ti
c
c
a
t
e
g
o
r
i
za
t
i
o
n
,
w
h
i
c
h
m
ay
n
o
t
c
a
p
t
u
r
e
t
h
e
e
v
o
l
v
i
n
g
n
a
t
u
r
e
o
f
c
o
n
v
e
r
s
a
t
i
o
n
a
l
A
I
a
s
n
e
w
t
e
c
h
n
i
q
u
e
s
a
n
d
h
y
b
r
i
d
m
o
d
e
l
s
e
m
e
r
g
e
.
F
u
t
u
r
e
r
e
s
ea
r
c
h
s
h
o
u
l
d
e
x
p
l
o
r
e
d
y
n
a
m
i
c
cl
a
s
s
i
f
i
ca
t
i
o
n
m
o
d
e
ls
t
h
a
t
a
c
c
o
u
n
t
f
o
r
t
h
e
s
e
e
v
o
l
v
i
n
g
f
a
c
t
o
r
s
a
n
d
i
n
v
e
s
ti
g
a
t
e
t
h
e
a
p
p
l
i
c
a
b
i
li
t
y
o
f
u
n
d
e
r
u
t
i
l
i
z
e
d
f
e
at
u
r
e
c
o
m
b
i
n
a
t
i
o
n
s
i
n
d
i
f
f
e
r
e
n
t
c
o
n
t
e
x
t
s
.
F
r
o
m
t
h
e
T
ab
l
e
2
,
B
E
R
T
a
n
d
v
a
r
i
a
n
t
s
o
f
B
E
R
T
p
e
r
f
o
r
m
s
b
e
t
t
e
r
t
h
a
n
t
h
e
p
r
e
v
i
o
u
s
m
o
d
e
l
s
b
u
t
t
h
e
c
o
m
b
i
n
a
t
i
o
n
o
f
f
e
a
t
u
r
e
s
m
a
t
t
e
r
s
w
h
i
l
e
p
e
r
f
o
r
m
i
n
g
N
L
P
t
as
k
.
A
s
w
e
c
a
te
g
o
r
i
z
e
d
v
ar
i
o
u
s
c
o
m
b
i
n
a
t
i
o
n
o
f
f
e
a
t
u
r
e
s
c
at
e
g
o
r
y
C
1
0
a
n
d
C
1
3
h
a
s
r
e
c
e
n
t
m
o
d
e
l
s
i
n
d
o
m
a
i
n
s
p
e
c
i
f
i
c
a
n
d
C
1
i
n
o
p
e
n
d
o
m
a
i
n
w
h
i
c
h
s
h
o
ws
t
h
e
i
n
t
e
r
e
s
t
s
o
f
b
u
i
l
d
i
n
g
c
h
at
b
o
t
w
it
h
t
h
e
s
e
f
e
at
u
r
e
s
p
la
y
s
m
a
j
o
r
r
o
l
e
i
n
s
u
c
ce
s
s
f
u
l
u
s
e
r
i
n
t
e
r
a
ct
i
o
n
.
T
ab
le
2
.
Per
f
o
r
m
an
ce
co
m
p
a
r
is
o
n
tab
le
b
ased
o
n
DC
AD
f
r
am
ewo
r
k
ca
teg
o
r
ies
P
a
p
e
r
s
Y
e
a
r
M
o
d
e
l
n
a
m
e
A
c
c
u
r
a
c
y
S
i
d
d
i
q
u
e
e
t
a
l
.
[
1
3
]
2
0
2
1
LSTM
+
I
n
t
e
n
t
En
c
o
d
e
r
s
+
C
o
n
c
e
p
t
N
e
t
9
2
.
5
4
X
u
e
a
n
d
R
e
n
[
5
2
]
2
0
2
1
IE
-
B
ER
TC
a
p
sN
e
t
9
7
.
5
2
K
u
mar
a
n
d
B
a
g
h
e
l
[
5
3
]
2
0
2
1
B
ER
T+
E
LM
o
+
B
i
LST
M
9
7
.
7
7
S
u
n
e
t
a
l
.
[
2
4
]
2
0
2
2
S
e
l
f
-
a
t
t
e
n
t
i
o
n
+
LST
M
9
6
.
7
A
b
r
o
e
t
a
l
.
[
9
]
2
0
2
2
W
F
S
T+
B
E
R
T
9
8
.
1
2
Zh
a
n
g
e
t
a
l
.
[
2
2
]
2
0
2
2
B
ER
T+
S
e
l
f
A
t
t
n
+
P
r
o
t
o
N
e
t
9
3
.
3
H
u
a
n
g
e
t
a
l
.
[
2
3
]
2
0
2
2
S
l
i
d
i
n
g
W
i
n
d
o
w
+
S
e
l
f
-
A
t
t
n
9
4
.
7
O
r
h
a
n
e
t
a
l
.
[
1
0
]
2
0
2
3
G
e
n
e
r
a
l
i
z
e
d
S
e
mS
p
a
c
e
+
B
i
LST
M
9
9
.
7
8
F
i
r
d
a
u
s
e
t
a
l
.
[
1
6
]
2
0
2
3
mBE
R
T+
A
t
t
n
+
C
R
F
9
9
.
1
8
Zh
a
n
g
e
t
a
l
.
[
3
7
]
2
0
2
3
ER
N
I
E+
C
R
F
+
A
t
t
n
9
8
.
1
3
B
a
b
u
a
n
d
B
o
d
d
u
[
1
1
]
2
0
2
4
B
ER
T
F
i
n
e
-
t
u
n
e
d
98
H
u
a
n
g
e
t
a
l
.
[
5
4
]
2
0
2
4
D
i
st
i
l
l
B
E
R
T+
F
A
N
9
7
.
9
W
u
e
t
a
l
.
[
5
5
]
2
0
2
4
C
B
L
M
A
-
B
(
C
N
N
+
B
i
LST
M
+
M
u
l
t
i
-
h
e
a
d
A
t
t
+
B
ER
T)
9
4
.
9
4.
3
.
Da
t
a
s
et
T
ab
le
3
p
r
esen
ts
a
ca
teg
o
r
ized
o
v
er
v
iew
o
f
th
e
d
atasets
u
s
ed
in
th
is
s
tu
d
y
,
o
f
f
er
in
g
v
alu
ab
l
e
in
s
ig
h
ts
in
to
th
eir
d
is
tr
ib
u
tio
n
an
d
d
iv
er
s
ity
ac
r
o
s
s
d
if
f
er
en
t
in
te
n
t
d
etec
tio
n
s
ce
n
ar
io
s
.
T
h
is
ca
teg
o
r
izatio
n
s
u
p
p
o
r
ts
a
co
m
p
r
eh
e
n
s
iv
e
ev
alu
atio
n
o
f
th
e
m
o
d
els,
en
s
u
r
in
g
r
o
b
u
s
t b
e
n
ch
m
ar
k
i
n
g
.
B
y
d
etailin
g
th
ese
d
atasets
,
th
e
tab
le
u
n
d
er
s
co
r
es th
e
r
elev
an
ce
a
n
d
ap
p
licab
ilit
y
o
f
t
h
e
s
elec
ted
d
a
ta
to
r
ea
l
-
wo
r
ld
i
n
ten
t d
etec
tio
n
task
s
.
T
ab
le
3
.
L
is
t o
f
d
atasets
u
s
ed
in
v
ar
io
u
s
ca
teg
o
r
ies
S
.
N
o
D
a
t
a
s
e
t
s
D
o
ma
i
n
La
n
g
u
a
g
e
s
C
a
t
e
g
o
r
i
e
s
1
A
TI
S
,
TR
A
I
N
S
,
a
n
d
F
R
A
M
ES
Tr
a
v
e
l
a
n
d
H
o
t
e
l
En
g
l
i
sh
C
3
a
n
d
C
7
.
2
A
sk
u
b
u
n
t
u
a
n
d
W
e
b
A
p
p
l
i
c
a
t
i
o
n
s
(
Q
&A
)
S
o
f
t
w
a
r
e
S
u
p
p
o
r
t
En
g
l
i
sh
C3
3
S
N
I
P
S
R
e
st
a
u
r
a
n
t
a
n
d
En
t
e
r
t
a
i
n
me
n
t
En
g
l
i
sh
C
3
,
C
6
4
C
o
n
c
e
p
t
N
e
t
C
o
mm
o
n
se
n
c
e
M
u
l
t
i
l
i
n
g
u
a
l
C6
5
C
o
r
n
e
l
l
M
o
v
i
e
D
i
a
l
o
g
u
e
D
a
t
a
se
t
M
o
v
i
e
D
i
a
l
o
g
u
e
En
g
l
i
sh
C6
6
D
a
i
l
y
D
i
a
l
o
g
u
e
s Da
t
a
se
t
D
a
i
l
y
D
i
a
l
o
g
u
e
En
g
l
i
sh
C
5
a
n
d
C7
7
D
X
M
e
d
i
c
a
l
D
i
a
l
o
g
u
e
a
n
d
M
Z
D
a
t
a
s
e
t
M
e
d
i
c
a
l
En
g
l
i
sh
C
1
4
8
F
a
c
e
b
o
o
k
M
u
l
t
i
l
i
n
g
u
a
l
D
i
a
l
o
g
u
e
M
u
l
t
i
-
D
o
m
a
i
n
s
M
u
l
t
i
l
i
n
g
u
a
l
C
3
a
n
d
C
7
.
9
K
V
R
ET
(
K
e
y
V
a
l
u
e
R
e
t
r
i
e
v
a
l
)
In
-
C
a
r
A
ssi
st
a
n
t
En
g
l
i
sh
C6
10
M
I
T
D
a
t
a
s
e
t
M
o
v
i
e
a
n
d
R
e
s
t
a
u
r
a
n
t
En
g
l
i
sh
C6
11
M
S
D
i
a
l
o
g
u
e
a
n
d
U
b
u
n
t
u
D
i
a
l
o
g
u
e
C
o
r
p
u
s
M
i
c
r
o
s
o
f
t
P
r
o
d
u
c
t
s
En
g
l
i
sh
C8
12
M
t
u
r
k
(
A
maz
o
n
)
M
u
l
t
i
-
D
o
m
a
i
n
En
l
i
s
h
C5
13
M
u
l
t
i
W
O
Z
D
a
t
a
set
M
u
l
t
i
-
D
o
m
a
i
n
En
g
l
i
sh
C
6
,
C
8
14
S
M
P
S
o
c
i
a
l
M
e
d
i
a
En
g
l
i
sh
C
3
&
C
7
.
15
S
TS SemE
v
a
l
D
a
t
a
s
e
t
N
e
w
s
En
g
l
i
sh
C5
16
M
o
v
i
e
Ti
c
k
e
t
B
o
o
k
i
n
g
M
o
v
i
e
B
o
o
k
i
n
g
En
g
l
i
sh
C
1
3
17
R
e
st
a
u
r
a
n
t
R
e
s
e
r
v
a
t
i
o
n
R
e
st
a
u
r
a
n
t
En
g
l
i
sh
C
1
3
18
Ta
x
i
O
r
d
e
r
i
n
g
Tr
a
v
e
l
En
g
l
i
sh
C
1
3
19
C
M
e
d
Q
M
e
d
i
c
a
l
Q
u
e
r
y
C
h
i
n
e
se
C
1
0
20
M
e
d
D
i
a
l
o
g
u
e
M
e
d
i
c
a
l
D
i
a
l
o
g
u
e
C
h
i
n
e
se
& E
n
g
l
i
sh
C
1
0
5.
CO
NCLU
SI
O
N
C
h
atb
o
ts
ar
e
a
g
r
ea
t
co
m
p
an
io
n
f
o
r
p
e
o
p
le,
p
r
o
v
id
in
g
s
atis
f
y
in
g
r
esp
o
n
s
es
to
th
e
u
s
er
wit
h
a
wea
lth
o
f
in
f
o
r
m
atio
n
.
I
n
ten
t
d
etec
tio
n
an
d
en
tity
r
ec
o
g
n
itio
n
is
th
e
co
r
e
task
wh
ich
s
h
o
u
ld
b
e
p
e
r
f
o
r
m
e
d
ac
cu
r
ately
.
Acc
o
r
d
in
g
to
th
e
in
ten
t
t
h
e
in
f
o
r
m
atio
n
is
s
e
ar
ch
ed
f
r
o
m
th
e
ex
ter
n
al
k
n
o
wled
g
e
a
n
d
th
e
a
p
p
r
o
p
r
iate
r
esp
o
n
s
e
is
s
en
t.
I
n
th
is
p
ap
er
,
s
o
m
e
i
m
p
o
r
tan
t
in
ten
t
d
etec
tio
n
task
s
wer
e
d
is
cu
s
s
ed
,
f
o
llo
wed
b
y
a
b
r
ief
h
is
to
r
y
o
n
ch
atb
o
ts
’
d
ev
el
o
p
m
en
t.
Fo
u
r
im
p
o
r
tan
t
f
ea
tu
r
es
wer
e
s
tu
d
ied
wh
ich
class
if
ies
th
e
ch
atb
o
t
in
to
v
a
r
io
u
s
ca
teg
o
r
ies.
I
n
ten
t
d
etec
tio
n
in
ch
atb
o
t
u
s
ec
ases
wer
e
also
d
is
cu
s
s
ed
wh
ich
g
av
e
an
id
ea
o
n
h
o
w
in
ten
t
d
etec
tio
n
p
lay
s
a
m
ajo
r
r
o
le
in
r
ea
l
-
tim
e
ap
p
licatio
n
s
.
DC
AD
class
if
i
ca
tio
n
s
h
o
w
a
clea
r
p
ictu
r
e
o
f
h
o
w
a
s
u
cc
ess
f
u
l
ch
atb
o
t
ca
n
b
e
b
u
ilt
with
ad
d
ed
f
ea
tu
r
es
an
d
also
s
h
o
ws
h
o
w
r
ec
en
t
m
o
d
els
p
er
f
o
r
m
b
ett
er
wh
ith
r
esp
ec
tiv
e
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
I
n
ten
t d
etec
tio
n
i
n
A
I
c
h
a
tb
o
ts
:
a
co
mp
r
eh
e
n
s
ive
r
ev
iew
o
f te
ch
n
iq
u
es a
n
d
…
(
Je
mima
h
K.
)
4257
f
ea
tu
r
es.
T
h
is
p
ap
er
will
h
elp
th
e
r
ea
d
er
s
to
a
n
aly
ze
th
e
r
ec
e
n
t
tr
en
d
s
an
d
ar
ea
s
in
wh
ich
i
n
ten
t
d
etec
tio
n
ca
n
b
e
m
ad
e
p
o
s
s
ib
le
in
r
ea
l
-
tim
e
ap
p
licatio
n
s
.
I
t
also
h
elp
s
th
e
r
ea
d
er
s
to
k
n
o
w
m
o
r
e
ab
o
u
t
f
e
atu
r
e
co
m
b
i
n
atio
n
s
in
a
ch
atb
o
t
d
esig
n
.
I
n
f
u
tu
r
e,
u
s
in
g
th
e
DC
AD
class
if
icatio
n
an
aly
s
is
,
n
ew
in
ten
t
d
etec
tio
n
m
o
d
els
will
b
e
b
u
ilt
in
ar
ea
s
wh
ich
ar
e
y
et
to
co
m
e.
Mo
r
eo
v
er
,
f
u
r
th
e
r
f
ea
tu
r
es
will
also
b
e
ad
d
ed
in
th
e
DC
AD
class
if
icatio
n
wh
ich
will g
iv
e
s
o
m
e
clar
ity
f
o
r
th
e
b
u
d
d
in
g
r
esear
ch
e
r
s
.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
Au
th
o
r
s
s
tate
n
o
f
u
n
d
in
g
in
v
o
lv
ed
.
AUTHO
R
CO
NT
RI
B
UT
I
O
NS ST
A
T
E
M
E
N
T
T
h
is
jo
u
r
n
al
u
s
es
th
e
C
o
n
t
r
ib
u
to
r
R
o
les
T
a
x
o
n
o
m
y
(
C
R
ed
iT)
to
r
ec
o
g
n
ize
in
d
iv
i
d
u
al
au
th
o
r
co
n
tr
ib
u
tio
n
s
,
r
ed
u
ce
au
th
o
r
s
h
ip
d
is
p
u
tes,
an
d
f
ac
ilit
ate
co
llab
o
r
atio
n
.
Na
m
e
o
f
Aut
ho
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
J
em
im
ah
K.
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
R
ajk
u
m
ar
Kan
n
an
✓
✓
✓
✓
✓
✓
✓
Fre
d
er
ic
An
d
r
es
✓
✓
✓
✓
✓
✓
✓
C
:
C
o
n
c
e
p
t
u
a
l
i
z
a
t
i
o
n
M
:
M
e
t
h
o
d
o
l
o
g
y
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
d
a
t
i
o
n
Fo
:
Fo
r
mal
a
n
a
l
y
s
i
s
I
:
I
n
v
e
s
t
i
g
a
t
i
o
n
R
:
R
e
so
u
r
c
e
s
D
:
D
a
t
a
C
u
r
a
t
i
o
n
O
:
W
r
i
t
i
n
g
-
O
r
i
g
i
n
a
l
D
r
a
f
t
E
:
W
r
i
t
i
n
g
-
R
e
v
i
e
w
&
E
d
i
t
i
n
g
Vi
:
Vi
su
a
l
i
z
a
t
i
o
n
Su
:
Su
p
e
r
v
i
s
i
o
n
P
:
P
r
o
j
e
c
t
a
d
mi
n
i
st
r
a
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
si
t
i
o
n
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
Au
th
o
r
s
s
tate
n
o
co
n
f
lict o
f
in
t
er
est
.
DATA AV
AI
L
AB
I
L
I
T
Y
Data
av
ailab
ilit
y
is
n
o
t
ap
p
lica
b
le
to
th
is
p
ap
er
as n
o
n
ew
d
at
a
wer
e
cr
ea
ted
o
r
an
aly
ze
d
in
t
h
is
s
tu
d
y
.
RE
F
E
R
E
NC
E
S
[
1
]
H
.
A
b
d
u
l
l
a
,
A
.
M
.
El
t
a
h
i
r
,
S
.
A
l
w
a
h
a
i
s
h
i
,
K
.
S
a
g
h
a
i
r
,
J.
P
l
a
t
o
s,
a
n
d
V
.
S
n
a
s
e
l
,
“
C
h
a
t
b
o
t
s
d
e
v
e
l
o
p
m
e
n
t
u
si
n
g
n
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
r
o
c
e
ss
i
n
g
:
a
r
e
v
i
e
w
,
”
i
n
2
0
2
2
2
6
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
C
i
r
c
u
i
t
s
,
S
y
st
e
m
s,
C
o
m
m
u
n
i
c
a
t
i
o
n
s
a
n
d
C
o
m
p
u
t
e
rs
(
C
S
C
C
)
,
2
0
2
2
,
p
p
.
1
2
2
–
128
,
d
o
i
:
1
0
.
1
1
0
9
/
C
S
C
C
5
5
9
3
1
.
2
0
2
2
.
0
0
0
3
0
.
[
2
]
V
.
G
u
p
t
a
,
V
.
J
o
s
h
i
,
A
.
J
a
i
n
,
a
n
d
I
.
G
a
r
g
,
“
C
h
a
t
b
o
t
f
o
r
me
n
t
a
l
h
e
a
l
t
h
s
u
p
p
o
r
t
u
si
n
g
N
LP
,
”
i
n
2
0
2
3
4
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
f
o
r
Em
e
rg
i
n
g
T
e
c
h
n
o
l
o
g
y
,
2
0
2
3
,
p
p
.
1
–
6
,
d
o
i
:
1
0
.
1
1
0
9
/
I
N
C
ET5
7
9
7
2
.
2
0
2
3
.
1
0
1
7
0
5
7
3
.
[
3
]
A
.
M
c
c
a
l
l
u
m
a
n
d
K
.
N
i
g
a
m
,
“
A
c
o
m
p
a
r
i
s
o
n
o
f
e
v
e
n
t
m
o
d
e
l
s
f
o
r
n
a
i
v
e
B
a
y
e
s
t
e
x
t
c
l
a
s
si
f
i
c
a
t
i
o
n
,
”
i
n
Pro
c
e
e
d
i
n
g
s
i
n
Wo
r
k
sh
o
p
o
n
L
e
a
r
n
i
n
g
f
o
r T
e
x
t
C
a
t
e
g
o
r
i
z
a
t
i
o
n
,
AA
AI
’
9
8
,
AA
AI
,
1
9
9
8
,
p
p
.
4
1
–
4
8
.
[
4
]
J.
W
e
i
z
e
n
b
a
u
m,
“
ELI
ZA
—
a
c
o
m
p
u
t
e
r
p
r
o
g
r
a
m
f
o
r
t
h
e
st
u
d
y
o
f
n
a
t
u
r
a
l
l
a
n
g
u
a
g
e
c
o
mm
u
n
i
c
a
t
i
o
n
b
e
t
w
e
e
n
ma
n
a
n
d
ma
c
h
i
n
e
,
”
C
o
m
m
u
n
i
c
a
t
i
o
n
s
o
f
t
h
e
A
C
M
,
v
o
l
.
9
,
n
o
.
1
,
p
p
.
3
6
–
4
5
,
J
a
n
.
1
9
6
6
,
d
o
i
:
1
0
.
1
1
4
5
/
3
6
5
1
5
3
.
3
6
5
1
6
8
.
[
5
]
J.
S
c
h
u
u
r
ma
n
s a
n
d
F
.
F
r
a
s
i
n
c
a
r
,
“
I
n
t
e
n
t
c
l
a
ssi
f
i
c
a
t
i
o
n
f
o
r
d
i
a
l
o
g
u
e
u
t
t
e
r
a
n
c
e
s,”
I
EEE
I
n
t
e
l
l
i
g
e
n
t
S
y
st
e
m
s
,
v
o
l
.
3
5
,
n
o
.
1
,
p
p
.
8
2
–
8
8
,
Jan
.
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
M
I
S
.
2
0
1
9
.
2
9
5
4
9
6
6
.
[
6
]
L.
X
u
,
Q
.
Z
h
o
u
,
K
.
G
o
n
g
,
X
.
Li
a
n
g
,
J
.
Ta
n
g
,
a
n
d
L.
Li
n
,
“
E
n
d
-
to
-
e
n
d
k
n
o
w
l
e
d
g
e
-
r
o
u
t
e
d
r
e
l
a
t
i
o
n
a
l
d
i
a
l
o
g
u
e
s
y
s
t
e
m
f
o
r
a
u
t
o
ma
t
i
c
d
i
a
g
n
o
si
s,
”
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
A
A
AI
C
o
n
f
e
r
e
n
c
e
o
n
Art
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
,
v
o
l
.
3
3
,
n
o
.
0
1
,
p
p
.
7
3
4
6
–
7
3
5
3
,
J
u
l
.
2
0
1
9
,
d
o
i
:
1
0
.
1
6
0
9
/
a
a
a
i
.
v
3
3
i
0
1
.
3
3
0
1
7
3
4
6
.
[
7
]
N
.
V
e
d
u
l
a
,
N
.
Li
p
k
a
,
P
.
M
a
n
e
r
i
k
e
r
,
a
n
d
S
.
P
a
r
t
h
a
s
a
r
a
t
h
y
,
“
T
o
w
a
r
d
s
o
p
e
n
i
n
t
e
n
t
d
i
s
c
o
v
e
r
y
f
o
r
c
o
n
v
e
r
sat
i
o
n
a
l
t
e
x
t
,
”
a
rXi
v
,
A
p
r
.
2
0
1
9
.
[
8
]
Z.
Z
h
a
n
g
,
Z
.
Z
h
a
n
g
,
H
.
C
h
e
n
,
a
n
d
Z
.
Zh
a
n
g
,
“
A
j
o
i
n
t
l
e
a
r
n
i
n
g
f
r
a
m
e
w
o
r
k
w
i
t
h
B
E
R
T
f
o
r
s
p
o
k
e
n
l
a
n
g
u
a
g
e
u
n
d
e
r
st
a
n
d
i
n
g
,
”
I
EEE
Ac
c
e
ss
,
v
o
l
.
7
,
p
p
.
1
6
8
8
4
9
–
1
6
8
8
5
8
,
2
0
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
1
9
.
2
9
5
4
7
6
6
.
[
9
]
W
.
A
.
A
b
r
o
,
G
.
Q
i
,
M
.
A
a
m
i
r
,
a
n
d
Z.
A
l
i
,
“
J
o
i
n
t
i
n
t
e
n
t
d
e
t
e
c
t
i
o
n
a
n
d
s
l
o
t
f
i
l
l
i
n
g
u
s
i
n
g
w
e
i
g
h
t
e
d
f
i
n
i
t
e
s
t
a
t
e
t
r
a
n
sd
u
c
e
r
a
n
d
B
ER
T
,
”
Ap
p
l
i
e
d
I
n
t
e
l
l
i
g
e
n
c
e
,
v
o
l
.
5
2
,
n
o
.
1
5
,
p
p
.
1
7
3
5
6
–
1
7
3
7
0
,
D
e
c
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
0
4
8
9
-
0
2
2
-
0
3
2
9
5
-
9.
[
1
0
]
U
.
O
r
h
a
n
,
E.
G
.
To
s
u
n
,
a
n
d
O
.
O
z
k
a
y
a
,
“
I
n
t
e
n
t
d
e
t
e
c
t
i
o
n
u
s
i
n
g
c
o
n
t
e
x
t
u
a
l
i
z
e
d
d
e
e
p
S
e
mS
p
a
c
e
,
”
Ara
b
i
a
n
J
o
u
r
n
a
l
f
o
r
S
c
i
e
n
c
e
a
n
d
En
g
i
n
e
e
ri
n
g
,
v
o
l
.
4
8
,
n
o
.
2
,
p
p
.
2
0
0
9
–
2
0
2
0
,
F
e
b
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
3
3
6
9
-
0
2
2
-
0
7
0
1
6
-
9.
[
1
1
]
A
.
B
a
b
u
a
n
d
S
.
B
.
B
o
d
d
u
,
“
B
E
R
T
-
b
a
se
d
me
d
i
c
a
l
c
h
a
t
b
o
t
:
e
n
h
a
n
c
i
n
g
h
e
a
l
t
h
c
a
r
e
c
o
mm
u
n
i
c
a
t
i
o
n
t
h
r
o
u
g
h
n
a
t
u
r
a
l
l
a
n
g
u
a
g
e
u
n
d
e
r
s
t
a
n
d
i
n
g
,
”
Ex
p
l
o
r
a
t
o
ry
Re
s
e
a
r
c
h
i
n
C
l
i
n
i
c
a
l
a
n
d
S
o
c
i
a
l
P
h
a
rm
a
c
y
,
v
o
l
.
1
3
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
r
c
so
p
.
2
0
2
4
.
1
0
0
4
1
9
.
[
1
2
]
T.
A
.
I
.
T.
B
a
h
a
,
M
.
E
l
H
a
j
j
i
,
Y
.
Es
-
S
a
a
d
y
,
a
n
d
H
.
F
a
d
i
l
i
,
“
T
o
w
a
r
d
s
h
i
g
h
l
y
a
d
a
p
t
i
v
e
Ed
u
-
C
h
a
t
b
o
t
,
”
Pro
c
e
d
i
a
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
1
9
8
,
p
p
.
3
9
7
–
4
0
3
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
p
r
o
c
s.
2
0
2
1
.
1
2
.
2
6
0
.
[
1
3
]
A
.
B
.
S
i
d
d
i
q
u
e
,
F
.
Jam
o
u
r
,
L.
X
u
,
a
n
d
V
.
H
r
i
s
t
i
d
i
s
,
“
G
e
n
e
r
a
l
i
z
e
d
z
e
r
o
-
s
h
o
t
i
n
t
e
n
t
d
e
t
e
c
t
i
o
n
v
i
a
c
o
mm
o
n
se
n
s
e
k
n
o
w
l
e
d
g
e
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
4
4
t
h
I
n
t
e
rn
a
t
i
o
n
a
l
A
C
M
S
I
G
I
R
C
o
n
f
e
re
n
c
e
o
n
R
e
se
a
r
c
h
a
n
d
D
e
v
e
l
o
p
m
e
n
t
i
n
I
n
f
o
rm
a
t
i
o
n
Re
t
r
i
e
v
a
l
,
2
0
2
1
,
p
p
.
1
9
2
5
–
1
9
2
9
,
d
o
i
:
1
0
.
1
1
4
5
/
3
4
0
4
8
3
5
.
3
4
6
2
9
8
5
.
[
1
4
]
Z.
Ze
n
g
,
D
.
M
a
,
H
.
Y
a
n
g
,
Z.
G
o
u
,
a
n
d
J
.
S
h
e
n
,
“
A
u
t
o
m
a
t
i
c
i
n
t
e
n
t
-
sl
o
t
i
n
d
u
c
t
i
o
n
f
o
r
d
i
a
l
o
g
u
e
s
y
s
t
e
ms,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
We
b
C
o
n
f
e
re
n
c
e
2
0
2
1
,
A
p
r
.
2
0
2
1
,
p
p
.
2
5
7
8
–
2
5
8
9
,
d
o
i
:
1
0
.
1
1
4
5
/
3
4
4
2
3
8
1
.
3
4
5
0
0
2
6
.
[
1
5
]
G
.
B
a
i
,
S
.
H
e
,
K
.
L
i
u
,
a
n
d
J.
Z
h
a
o
,
“
I
n
c
r
e
me
n
t
a
l
i
n
t
e
n
t
d
e
t
e
c
t
i
o
n
f
o
r
me
d
i
c
a
l
d
o
mai
n
w
i
t
h
c
o
n
t
r
a
s
t
r
e
p
l
a
y
n
e
t
w
o
r
k
s
,
”
i
n
F
i
n
d
i
n
g
s
o
f
t
h
e
Ass
o
c
i
a
t
i
o
n
f
o
r
C
o
m
p
u
t
a
t
i
o
n
a
l
L
i
n
g
u
i
st
i
c
s:
A
C
L
2
0
2
2
,
2
0
2
2
,
p
p
.
3
5
4
9
–
3
5
5
6
,
d
o
i
:
1
0
.
1
8
6
5
3
/
v
1
/
2
0
2
2
.
f
i
n
d
i
n
g
s
-
a
c
l
.
2
8
0
.
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
4
,
No
.
5
,
Octo
b
er
2
0
2
5
:
4
2
5
0
-
4
2
5
9
4258
[
1
6
]
M
.
F
i
r
d
a
u
s
,
A
.
E
k
b
a
l
,
a
n
d
E
.
C
a
mb
r
i
a
,
“
M
u
l
t
i
t
a
s
k
l
e
a
r
n
i
n
g
f
o
r
m
u
l
t
i
l
i
n
g
u
a
l
i
n
t
e
n
t
d
e
t
e
c
t
i
o
n
a
n
d
sl
o
t
f
i
l
l
i
n
g
i
n
d
i
a
l
o
g
u
e
sy
s
t
e
ms
,
”
I
n
f
o
rm
a
t
i
o
n
F
u
si
o
n
,
v
o
l
.
9
1
,
p
p
.
2
9
9
–
3
1
5
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
n
f
f
u
s.
2
0
2
2
.
0
9
.
0
2
9
.
[
1
7
]
A
.
B
h
a
s
i
n
,
B
.
N
a
t
a
r
a
j
a
n
,
G
.
M
a
t
h
u
r
,
J.
H
.
Je
o
n
,
a
n
d
J.
-
S
.
K
i
m,
“
U
n
i
f
i
e
d
p
a
r
a
l
l
e
l
i
n
t
e
n
t
a
n
d
sl
o
t
p
r
e
d
i
c
t
i
o
n
w
i
t
h
c
r
o
s
s
f
u
si
o
n
a
n
d
s
l
o
t
mask
i
n
g
,
”
N
a
t
u
ra
l
L
a
n
g
u
a
g
e
Pr
o
c
e
ssi
n
g
a
n
d
I
n
f
o
rm
a
t
i
o
n
S
y
st
e
m
s
(
N
L
D
B
2
0
1
9
)
,
C
h
a
m
,
S
w
i
t
z
e
r
l
a
n
d
:
S
p
r
i
n
g
e
r
,
2
0
1
9
,
p
p
.
2
7
7
–
285
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
030
-
2
3
2
8
1
-
8
_
2
2
.
[
1
8
]
N
.
La
i
r
,
C
.
D
e
l
g
r
a
n
g
e
,
D
.
M
u
g
i
sh
a
,
J.
-
M
.
D
u
ss
o
u
x
,
P
.
-
Y
.
O
u
d
e
y
e
r
,
a
n
d
P
.
F
.
D
o
mi
n
e
y
,
“
U
ser
-
in
-
t
h
e
-
l
o
o
p
a
d
a
p
t
i
v
e
i
n
t
e
n
t
d
e
t
e
c
t
i
o
n
f
o
r
i
n
s
t
r
u
c
t
a
b
l
e
d
i
g
i
t
a
l
a
ss
i
st
a
n
t
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
2
5
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
I
n
t
e
l
l
i
g
e
n
t
U
ser
I
n
t
e
r
f
a
c
e
s
,
M
a
r
.
2
0
2
0
,
p
p
.
1
1
6
–
127
,
d
o
i
:
1
0
.
1
1
4
5
/
3
3
7
7
3
2
5
.
3
3
7
7
4
9
0
.
[
1
9
]
Y
.
W
a
n
g
,
T
.
H
e
,
R
.
F
a
n
,
W
.
Zh
o
u
,
a
n
d
X
.
Tu
,
“
Ef
f
e
c
t
i
v
e
u
t
i
l
i
z
a
t
i
o
n
o
f
e
x
t
e
r
n
a
l
k
n
o
w
l
e
d
g
e
a
n
d
h
i
st
o
r
y
c
o
n
t
e
x
t
i
n
mu
l
t
i
-
t
u
r
n
s
p
o
k
e
n
l
a
n
g
u
a
g
e
u
n
d
e
r
st
a
n
d
i
n
g
m
o
d
e
l
,
”
i
n
2
0
1
9
I
EEE
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
B
i
g
D
a
t
a
(
Bi
g
D
a
t
a
)
,
D
e
c
.
2
0
1
9
,
p
p
.
9
6
0
–
9
6
7
,
d
o
i
:
1
0
.
1
1
0
9
/
B
i
g
D
a
t
a
4
7
0
9
0
.
2
0
1
9
.
9
0
0
6
1
6
2
.
[
2
0
]
S
.
S
.
A
b
d
u
l
l
a
h
i
,
S
.
Y
i
m
i
n
g
,
A
.
A
b
d
u
l
l
a
h
i
,
a
n
d
U
.
A
l
i
y
u
,
“
O
p
e
n
d
o
ma
i
n
c
h
a
t
b
o
t
b
a
se
d
o
n
a
t
t
e
n
t
i
v
e
e
n
d
-
to
-
e
n
d
S
e
q
2
S
e
q
me
c
h
a
n
i
sm,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
2
0
1
9
2
n
d
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
A
l
g
o
r
i
t
h
m
s,
C
o
m
p
u
t
i
n
g
a
n
d
Ar
t
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
,
D
e
c
.
2
0
1
9
,
p
p
.
3
3
9
–
344
,
d
o
i
:
1
0
.
1
1
4
5
/
3
3
7
7
7
1
3
.
3
3
7
7
7
7
3
.
[
2
1
]
D
.
M
.
N
g
u
y
e
n
,
Y
.
-
T.
H
.
C
h
i
u
,
a
n
d
H
.
D
.
L
e
,
“
D
e
t
e
r
m
i
n
a
n
t
s o
f
c
o
n
t
i
n
u
a
n
c
e
i
n
t
e
n
t
i
o
n
t
o
w
a
r
d
s
B
a
n
k
s’
c
h
a
t
b
o
t
ser
v
i
c
e
s
i
n
V
i
e
t
n
a
m:
a
n
e
c
e
ssi
t
y
f
o
r
s
u
s
t
a
i
n
a
b
l
e
d
e
v
e
l
o
p
m
e
n
t
,
”
S
u
st
a
i
n
a
b
i
l
i
t
y
,
v
o
l
.
1
3
,
n
o
.
1
4
,
p
.
7
6
2
5
,
J
u
l
.
2
0
2
1
,
d
o
i
:
1
0
.
3
3
9
0
/
s
u
1
3
1
4
7
6
2
5
.
[
2
2
]
X
.
Z
h
a
n
g
,
F
.
C
a
i
,
X
.
H
u
,
J.
Zh
e
n
g
,
a
n
d
H
.
C
h
e
n
,
“
A
c
o
n
t
r
a
s
t
i
v
e
l
e
a
r
n
i
n
g
-
b
a
se
d
t
a
s
k
a
d
a
p
t
a
t
i
o
n
m
o
d
e
l
f
o
r
f
e
w
-
sh
o
t
i
n
t
e
n
t
r
e
c
o
g
n
i
t
i
o
n
,
”
I
n
f
o
rm
a
t
i
o
n
P
ro
c
e
ss
i
n
g
& M
a
n
a
g
e
m
e
n
t
,
v
o
l
.
5
9
,
n
o
.
3
,
p
.
1
0
2
8
6
3
,
M
a
y
2
0
2
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
p
m.
2
0
2
1
.
1
0
2
8
6
3
.
[
2
3
]
H
.
H
u
a
n
g
,
P
.
H
u
a
n
g
,
Z.
Z
h
u
,
J.
Li
,
a
n
d
P
.
Li
n
,
“
C
LI
D
:
a
c
h
u
n
k
-
l
e
v
e
l
i
n
t
e
n
t
d
e
t
e
c
t
i
o
n
f
r
a
mew
o
r
k
f
o
r
m
u
l
t
i
p
l
e
i
n
t
e
n
t
sp
o
k
e
n
l
a
n
g
u
a
g
e
u
n
d
e
r
s
t
a
n
d
i
n
g
,
”
I
E
EE
S
i
g
n
a
l
Pro
c
e
ss
i
n
g
L
e
t
t
e
rs
,
v
o
l
.
2
9
,
p
p
.
2
1
2
3
–
2
1
2
7
,
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
LSP
.
2
0
2
2
.
3
2
1
1
1
5
6
.
[
2
4
]
X
.
S
u
n
,
H
.
Zh
e
n
g
,
a
n
d
Z.
T
a
n
g
,
“
H
i
s
t
o
r
i
c
a
l
i
n
f
o
r
m
a
t
i
o
n
-
b
a
se
d
i
n
t
e
n
t
d
e
t
e
c
t
i
o
n
f
o
r
m
u
l
t
i
t
u
r
n
d
i
a
l
o
g
u
e
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
8
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
C
o
m
p
u
t
i
n
g
a
n
d
Ar
t
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
,
M
a
r
.
2
0
2
2
,
p
p
.
5
6
6
–
572
,
d
o
i
:
1
0
.
1
1
4
5
/
3
5
3
2
2
1
3
.
3
5
3
2
2
9
9
.
[
2
5
]
H
.
Li
u
e
t
a
l
.
,
“
A
s
i
m
p
l
e
me
t
a
-
l
e
a
r
n
i
n
g
p
a
r
a
d
i
g
m
f
o
r
z
e
r
o
-
sh
o
t
i
n
t
e
n
t
c
l
a
ss
i
f
i
c
a
t
i
o
n
w
i
t
h
m
i
x
t
u
r
e
a
t
t
e
n
t
i
o
n
mec
h
a
n
i
sm
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
4
5
t
h
I
n
t
e
rn
a
t
i
o
n
a
l
AC
M
S
I
G
I
R
C
o
n
f
e
re
n
c
e
o
n
R
e
se
a
rc
h
a
n
d
D
e
v
e
l
o
p
m
e
n
t
i
n
I
n
f
o
rm
a
t
i
o
n
R
e
t
ri
e
v
a
l
,
Ju
l
.
2
0
2
2
,
p
p
.
2
0
4
7
–
2
0
5
2
,
d
o
i
:
1
0
.
1
1
4
5
/
3
4
7
7
4
9
5
.
3
5
3
1
8
0
3
.
[
2
6
]
M
.
F
i
r
d
a
u
s
,
A
.
K
u
m
a
r
,
A
.
E
k
b
a
l
,
a
n
d
P
.
B
h
a
t
t
a
c
h
a
r
y
y
a
,
“
A
m
u
l
t
i
-
t
a
s
k
h
i
e
r
a
r
c
h
i
c
a
l
a
p
p
r
o
a
c
h
f
o
r
i
n
t
e
n
t
d
e
t
e
c
t
i
o
n
a
n
d
s
l
o
t
f
i
l
l
i
n
g
,
”
K
n
o
w
l
e
d
g
e
-
B
a
se
d
S
y
s
t
e
m
s
,
v
o
l
.
1
8
3
,
2
0
1
9
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
k
n
o
sy
s
.
2
0
1
9
.
0
7
.
0
1
7
.
[
2
7
]
T.
-
E.
Li
n
a
n
d
H
.
X
u
,
“
A
p
o
st
-
p
r
o
c
e
s
si
n
g
met
h
o
d
f
o
r
d
e
t
e
c
t
i
n
g
u
n
k
n
o
w
n
i
n
t
e
n
t
o
f
d
i
a
l
o
g
u
e
s
y
st
e
m
v
i
a
p
r
e
-
t
r
a
i
n
e
d
d
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
c
l
a
ss
i
f
i
e
r
,
”
K
n
o
w
l
e
d
g
e
-
B
a
se
d
S
y
s
t
e
m
s
,
v
o
l
.
1
8
6
,
p
.
1
0
4
9
7
9
,
D
e
c
.
2
0
1
9
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
k
n
o
s
y
s.
2
0
1
9
.
1
0
4
9
7
9
.
[
2
8
]
H.
-
M
.
C
h
u
a
n
g
a
n
d
D
.
-
W
.
C
h
e
n
g
,
“
C
o
n
v
e
r
sa
t
i
o
n
a
l
A
I
o
v
e
r
m
i
l
i
t
a
r
y
sc
e
n
a
r
i
o
s
u
s
i
n
g
i
n
t
e
n
t
d
e
t
e
c
t
i
o
n
a
n
d
r
e
s
p
o
n
se
g
e
n
e
r
a
t
i
o
n
,
”
Ap
p
l
i
e
d
S
c
i
e
n
c
e
s
,
v
o
l
.
1
2
,
n
o
.
5
,
p
.
2
4
9
4
,
F
e
b
.
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
a
p
p
1
2
0
5
2
4
9
4
.
[
2
9
]
M
.
H
.
M
a
q
b
o
o
l
,
M
.
F
e
r
e
i
d
o
u
n
i
,
A
.
B
.
S
i
d
d
i
q
u
e
,
a
n
d
H
.
F
o
r
o
o
s
h
,
“
M
o
d
e
l
-
a
g
n
o
s
t
i
c
z
e
r
o
-
sh
o
t
i
n
t
e
n
t
d
e
t
e
c
t
i
o
n
v
i
a
c
o
n
t
r
a
s
t
i
v
e
t
r
a
n
sf
e
r
l
e
a
r
n
i
n
g
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
S
e
m
a
n
t
i
c
C
o
m
p
u
t
i
n
g
,
v
o
l
.
1
8
,
n
o
.
0
1
,
p
p
.
5
–
2
4
,
M
a
r
.
2
0
2
4
,
d
o
i
:
1
0
.
1
1
4
2
/
S
1
7
9
3
3
5
1
X
2
4
4
1
0
0
1
0
.
[
3
0
]
Y
.
W
a
n
g
,
J
.
H
u
a
n
g
,
T.
H
e
,
a
n
d
X
.
T
u
,
“
D
i
a
l
o
g
u
e
i
n
t
e
n
t
c
l
a
ssi
f
i
c
a
t
i
o
n
w
i
t
h
c
h
a
r
a
c
t
e
r
-
C
N
N
-
B
G
R
U
n
e
t
w
o
r
k
s
,
”
M
u
l
t
i
m
e
d
i
a
T
o
o
l
s
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
7
9
,
n
o
.
7
–
8
,
p
p
.
4
5
5
3
–
4
5
7
2
,
F
e
b
.
2
0
2
0
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
1
0
4
2
-
0
1
9
-
7
6
7
8
-
1.
[
3
1
]
Y
.
Li
,
R
.
Z
h
a
n
g
,
W
.
Li
,
a
n
d
Z
.
C
a
o
,
“
H
i
e
r
a
r
c
h
i
c
a
l
p
r
e
d
i
c
t
i
o
n
a
n
d
a
d
v
e
r
sar
i
a
l
l
e
a
r
n
i
n
g
f
o
r
c
o
n
d
i
t
i
o
n
a
l
r
e
sp
o
n
se
g
e
n
e
r
a
t
i
o
n
,
”
I
EEE
T
ra
n
s
a
c
t
i
o
n
s
o
n
K
n
o
w
l
e
d
g
e
a
n
d
D
a
t
a
En
g
i
n
e
e
ri
n
g
,
v
o
l
.
3
4
,
n
o
.
1
,
p
p
.
3
1
4
–
3
2
7
,
Ja
n
.
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
T
K
D
E
.
2
0
2
0
.
2
9
7
7
6
3
7
.
[
3
2
]
F
.
R
e
n
a
n
d
S
.
X
u
e
,
“
I
n
t
e
n
t
i
o
n
d
e
t
e
c
t
i
o
n
b
a
se
d
o
n
S
i
a
mes
e
n
e
u
r
a
l
n
e
t
w
o
r
k
w
i
t
h
t
r
i
p
l
e
t
l
o
s
s,”
I
EEE
Ac
c
e
ss
,
v
o
l
.
8
,
p
p
.
8
2
2
4
2
–
8
2
2
5
4
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
0
.
2
9
9
1
4
8
4
.
[
3
3
]
M
.
F
i
r
d
a
u
s,
H
.
G
o
l
c
h
h
a
,
A
.
Ek
b
a
l
,
a
n
d
P
.
B
h
a
t
t
a
c
h
a
r
y
y
a
,
“
A
d
e
e
p
m
u
l
t
i
-
t
a
s
k
mo
d
e
l
f
o
r
d
i
a
l
o
g
u
e
a
c
t
c
l
a
s
si
f
i
c
a
t
i
o
n
,
i
n
t
e
n
t
d
e
t
e
c
t
i
o
n
a
n
d
s
l
o
t
f
i
l
l
i
n
g
,
”
C
o
g
n
i
t
i
v
e
C
o
m
p
u
t
a
t
i
o
n
,
v
o
l
.
1
3
,
n
o
.
3
,
p
p
.
6
2
6
–
6
4
5
,
M
a
y
2
0
2
1
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
2
5
5
9
-
0
2
0
-
0
9
7
1
8
-
4.
[
3
4
]
W
.
A
.
A
b
r
o
,
G
.
Q
i
,
H
.
G
a
o
,
M
.
A
.
K
h
a
n
,
a
n
d
Z.
A
l
i
,
“
M
u
l
t
i
-
t
u
r
n
i
n
t
e
n
t
d
e
t
e
r
mi
n
a
t
i
o
n
f
o
r
g
o
a
l
-
o
r
i
e
n
t
e
d
d
i
a
l
o
g
u
e
s
y
st
e
ms,
”
i
n
2
0
1
9
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
i
n
t
C
o
n
f
e
r
e
n
c
e
o
n
N
e
u
ra
l
N
e
t
w
o
rks
,
J
u
l
.
2
0
1
9
,
p
p
.
1
–
8
,
d
o
i
:
1
0
.
1
1
0
9
/
I
JC
N
N
.
2
0
1
9
.
8
8
5
2
2
4
6
.
[
3
5
]
Q
.
B
a
o
,
L.
N
i
,
a
n
d
J
.
Li
u
,
“
H
H
H
:
a
n
o
n
l
i
n
e
me
d
i
c
a
l
c
h
a
t
b
o
t
s
y
st
e
m
b
a
s
e
d
o
n
k
n
o
w
l
e
d
g
e
g
r
a
p
h
a
n
d
h
i
e
r
a
r
c
h
i
c
a
l
b
i
-
d
i
r
e
c
t
i
o
n
a
l
a
t
t
e
n
t
i
o
n
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
A
u
st
r
a
l
a
si
a
n
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
We
e
k
Mu
l
t
i
c
o
n
f
e
r
e
n
c
e
,
F
e
b
.
2
0
2
0
,
p
p
.
1
–
10
,
d
o
i
:
1
0
.
1
1
4
5
/
3
3
7
3
0
1
7
.
3
3
7
3
0
4
9
.
[
3
6
]
K
.
Y
a
n
g
,
X
.
K
o
n
g
,
Y
.
W
a
n
g
,
J
.
Z
h
a
n
g
,
a
n
d
G
.
D
e
M
e
l
o
,
“
R
e
i
n
f
o
r
c
e
me
n
t
l
e
a
r
n
i
n
g
o
v
e
r
k
n
o
w
l
e
d
g
e
g
r
a
p
h
s
f
o
r
e
x
p
l
a
i
n
a
b
l
e
d
i
a
l
o
g
u
e
i
n
t
e
n
t
mi
n
i
n
g
,
”
I
EEE
Ac
c
e
ss
,
v
o
l
.
8
,
p
p
.
8
5
3
4
8
–
8
5
3
5
8
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
0
.
2
9
9
1
2
5
7
.
[
3
7
]
B
.
Z
h
a
n
g
,
Z.
Tu
,
S
.
H
a
n
g
,
D
.
C
h
u
,
a
n
d
X
.
X
u
,
“
C
o
n
c
o
-
ER
N
I
E:
c
o
m
p
l
e
x
u
ser
i
n
t
e
n
t
d
e
t
e
c
t
m
o
d
e
l
f
o
r
smar
t
h
e
a
l
t
h
c
a
r
e
c
o
g
n
i
t
i
v
e
b
o
t
,
”
A
C
M
T
r
a
n
sa
c
t
i
o
n
s
o
n
I
n
t
e
rn
e
t
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
2
3
,
n
o
.
1
,
p
p
.
1
–
2
4
,
F
e
b
.
2
0
2
3
,
d
o
i
:
1
0
.
1
1
4
5
/
3
5
7
4
1
3
5
.
[
3
8
]
M
.
D
.
I
.
-
M
a
n
z
a
n
o
,
N
.
V
.
L
ó
p
e
z
,
N
.
A
.
G
o
n
z
á
l
e
z
,
a
n
d
C
.
C
.
R
o
d
r
í
g
u
e
z
,
“
I
mp
l
e
me
n
t
a
t
i
o
n
o
f
c
h
a
t
b
o
t
i
n
o
n
l
i
n
e
c
o
m
merce
,
a
n
d
o
p
e
n
i
n
n
o
v
a
t
i
o
n
,
”
J
o
u
r
n
a
l
o
f
O
p
e
n
I
n
n
o
v
a
t
i
o
n
:
T
e
c
h
n
o
l
o
g
y
,
M
a
rk
e
t
,
a
n
d
C
o
m
p
l
e
x
i
t
y
,
v
o
l
.
7
,
n
o
.
2
,
p
.
1
2
5
,
J
u
n
.
2
0
2
1
,
d
o
i
:
1
0
.
3
3
9
0
/
j
o
i
t
mc
7
0
2
0
1
2
5
.
[
3
9
]
Z.
Zh
a
n
g
,
Z.
L
i
,
H
.
Li
u
,
a
n
d
N
.
N
.
X
i
o
n
g
,
“
M
u
l
t
i
-
sca
l
e
d
y
n
a
m
i
c
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
t
w
o
r
k
f
o
r
k
n
o
w
l
e
d
g
e
g
r
a
p
h
e
mb
e
d
d
i
n
g
,
”
I
EEE
T
ra
n
s
a
c
t
i
o
n
s
o
n
K
n
o
w
l
e
d
g
e
a
n
d
D
a
t
a
En
g
i
n
e
e
ri
n
g
,
v
o
l
.
3
4
,
n
o
.
5
,
p
p
.
2
3
3
5
–
2
3
4
7
,
M
a
y
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
T
K
D
E.
2
0
2
0
.
3
0
0
5
9
5
2
.
[
4
0
]
S
.
S
o
n
g
,
X
.
C
h
e
n
,
C
.
W
a
n
g
,
X
.
Y
u
,
J.
W
a
n
g
,
a
n
d
X
.
H
e
,
“
A
t
w
o
-
st
a
g
e
u
ser i
n
t
e
n
t
d
e
t
e
c
t
i
o
n
mo
d
e
l
o
n
c
o
m
p
l
i
c
a
t
e
d
u
t
t
e
r
a
n
c
e
s
w
i
t
h
mu
l
t
i
-
t
a
s
k
l
e
a
r
n
i
n
g
,
”
i
n
C
o
m
p
a
n
i
o
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
We
b
C
o
n
f
e
r
e
n
c
e
2
0
2
2
,
A
p
r
.
2
0
2
2
,
p
p
.
1
9
7
–
200
,
d
o
i
:
1
0
.
1
1
4
5
/
3
4
8
7
5
5
3
.
3
5
2
4
2
3
2
.
[
4
1
]
A
.
M
u
l
l
i
c
k
,
I
.
M
o
n
d
a
l
,
S
.
R
a
y
,
R
.
R
,
G
.
C
h
a
i
t
a
n
y
a
,
a
n
d
P
.
G
o
y
a
l
,
“
I
n
t
e
n
t
i
d
e
n
t
i
f
i
c
a
t
i
o
n
a
n
d
e
n
t
i
t
y
e
x
t
r
a
c
t
i
o
n
f
o
r
h
e
a
l
t
h
c
a
r
e
q
u
e
r
i
e
s
i
n
I
n
d
i
c
l
a
n
g
u
a
g
e
s
,
”
i
n
F
i
n
d
i
n
g
s
o
f
t
h
e
Ass
o
c
i
a
t
i
o
n
f
o
r
C
o
m
p
u
t
a
t
i
o
n
a
l
L
i
n
g
u
i
s
t
i
c
s:
E
AC
L
2
0
2
3
,
2
0
2
3
,
p
p
.
1
8
7
0
–
1
8
8
1
,
d
o
i
:
1
0
.
1
8
6
5
3
/
v
1
/
2
0
2
3
.
f
i
n
d
i
n
g
s
-
e
a
c
l
.
1
4
0
.
[
4
2
]
S
.
P
a
r
i
k
h
,
M
.
Ti
w
a
r
i
,
P
.
T
u
m
b
a
d
e
,
a
n
d
Q
.
V
o
h
r
a
,
“
E
x
p
l
o
r
i
n
g
z
e
r
o
a
n
d
f
e
w
-
sh
o
t
t
e
c
h
n
i
q
u
e
s
f
o
r
i
n
t
e
n
t
c
l
a
ssi
f
i
c
a
t
i
o
n
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
6
1
s
t
A
n
n
u
a
l
Me
e
t
i
n
g
o
f
t
h
e
A
sso
c
i
a
t
i
o
n
f
o
r
C
o
m
p
u
t
a
t
i
o
n
a
l
L
i
n
g
u
i
st
i
c
s
(
V
o
l
u
m
e
5
:
I
n
d
u
st
r
y
T
r
a
c
k
)
,
2
0
2
3
,
p
p
.
7
4
4
–
7
5
1
,
d
o
i
:
1
0
.
1
8
6
5
3
/
v
1
/
2
0
2
3
.
a
c
l
-
i
n
d
u
s
t
r
y
.
7
1
.
[
4
3
]
A
.
Za
m
a
n
i
e
t
a
l
.
,
“
I
n
t
e
n
t
a
n
d
e
n
t
i
t
y
d
e
t
e
c
t
i
o
n
w
i
t
h
d
a
t
a
a
u
g
m
e
n
t
a
t
i
o
n
f
o
r
a
men
t
a
l
h
e
a
l
t
h
v
i
r
t
u
a
l
a
s
si
s
t
a
n
t
c
h
a
t
b
o
t
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
2
3
rd
A
C
M
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
I
n
t
e
l
l
i
g
e
n
t
V
i
rt
u
a
l
A
g
e
n
t
s
,
S
e
p
.
2
0
2
3
,
p
p
.
1
–
4
,
d
o
i
:
1
0
.
1
1
4
5
/
3
5
7
0
9
4
5
.
3
6
0
7
3
2
4
.
[
4
4
]
O
.
E
.
O
j
o
,
A
.
G
e
l
b
u
k
h
,
H
.
C
a
l
v
o
,
a
n
d
A
.
F
e
l
d
m
a
n
,
“
L
a
n
g
u
a
g
e
i
d
e
n
t
i
f
i
c
a
t
i
o
n
a
t
t
h
e
w
o
r
d
l
e
v
e
l
i
n
c
o
d
e
-
m
i
x
e
d
t
e
x
t
s
u
s
i
n
g
c
h
a
r
a
c
t
e
r
s
e
q
u
e
n
c
e
a
n
d
w
o
r
d
e
m
b
e
d
d
i
n
g
,
”
i
n
P
r
o
c
e
e
d
i
n
g
s
o
f
t
h
e
1
9
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
N
a
t
u
r
a
l
L
a
n
g
u
a
g
e
P
r
o
c
e
s
s
i
n
g
,
2
0
2
2
,
p
p
.
1
–
6.
[
4
5
]
T.
S
u
r
e
sh
,
A
.
S
e
n
g
u
p
t
a
,
M
.
S
.
A
k
h
t
a
r
,
a
n
d
T.
C
h
a
k
r
a
b
o
r
t
y
,
“
A
c
o
m
p
r
e
h
e
n
si
v
e
u
n
d
e
r
s
t
a
n
d
i
n
g
o
f
c
o
d
e
-
m
i
x
e
d
l
a
n
g
u
a
g
e
s
e
ma
n
t
i
c
s
u
si
n
g
h
i
e
r
a
r
c
h
i
c
a
l
t
r
a
n
sf
o
r
mer,
”
I
EE
E
T
ra
n
sa
c
t
i
o
n
s
o
n
C
o
m
p
u
t
a
t
i
o
n
a
l
S
o
c
i
a
l
S
y
s
t
e
m
s
,
v
o
l
.
1
1
,
n
o
.
3
,
p
p
.
4
1
3
9
–
4
1
4
8
,
Ju
n
.
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
TC
S
S
.
2
0
2
4
.
3
3
6
0
3
7
8
.
[
4
6
]
A
.
S
e
n
g
u
p
t
a
,
S
.
K
.
B
h
a
t
t
a
c
h
a
r
j
e
e
,
T.
C
h
a
k
r
a
b
o
r
t
y
,
a
n
d
M
.
S
.
A
k
h
t
a
r
,
“
H
I
T
-
a
h
i
e
r
a
r
c
h
i
c
a
l
l
y
f
u
s
e
d
d
e
e
p
a
t
t
e
n
t
i
o
n
n
e
t
w
o
r
k
f
o
r
r
o
b
u
st
c
o
d
e
-
m
i
x
e
d
l
a
n
g
u
a
g
e
r
e
p
r
e
s
e
n
t
a
t
i
o
n
,
”
i
n
F
i
n
d
i
n
g
s
o
f
t
h
e
Ass
o
c
i
a
t
i
o
n
f
o
r
C
o
m
p
u
t
a
t
i
o
n
a
l
L
i
n
g
u
i
st
i
c
s:
AC
L
-
I
J
C
N
L
P
2
0
2
1
,
2
0
2
1
,
p
p
.
4
6
2
5
–
4
6
3
9
,
d
o
i
:
1
0
.
1
8
6
5
3
/
v
1
/
2
0
2
1
.
f
i
n
d
i
n
g
s
-
a
c
l
.
4
0
7
.
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
I
n
ten
t d
etec
tio
n
i
n
A
I
c
h
a
tb
o
ts
:
a
co
mp
r
eh
e
n
s
ive
r
ev
iew
o
f te
ch
n
iq
u
es a
n
d
…
(
Je
mima
h
K.
)
4259
[
4
7
]
V
.
S
r
i
v
a
s
t
a
v
a
a
n
d
M
.
S
i
n
g
h
,
“
P
H
I
N
C
:
a
p
a
r
a
l
l
e
l
H
i
n
g
l
i
sh
s
o
c
i
a
l
m
e
d
i
a
c
o
d
e
-
mi
x
e
d
c
o
r
p
u
s
f
o
r
mac
h
i
n
e
t
r
a
n
sl
a
t
i
o
n
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
S
i
x
t
h
Wo
rks
h
o
p
o
n
N
o
i
sy
U
s
e
r
-
g
e
n
e
r
a
t
e
d
T
e
x
t
(
W
-
N
U
T
2
0
2
0
)
,
2
0
2
0
,
p
p
.
4
1
–
49
,
d
o
i
:
1
0
.
1
8
6
5
3
/
v
1
/
2
0
2
0
.
w
n
u
t
-
1
.
7
.
[
4
8
]
J.
Li
u
,
T.
Y
.
K
e
a
t
,
B
.
F
u
,
a
n
d
K
.
H
.
L
i
m,
“
LA
R
A
:
l
i
n
g
u
i
st
i
c
-
a
d
a
p
t
i
v
e
r
e
t
r
i
e
v
a
l
-
a
u
g
me
n
t
a
t
i
o
n
f
o
r
mu
l
t
i
-
t
u
r
n
i
n
t
e
n
t
c
l
a
ss
i
f
i
c
a
t
i
o
n
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
2
0
2
4
C
o
n
f
e
re
n
c
e
o
n
Em
p
i
ri
c
a
l
M
e
t
h
o
d
s
i
n
N
a
t
u
ra
l
L
a
n
g
u
a
g
e
Pro
c
e
ss
i
n
g
:
I
n
d
u
st
r
y
T
ra
c
k
,
2
0
2
4
,
p
p
.
1
0
9
6
–
1
1
0
6
,
d
o
i
:
1
0
.
1
8
6
5
3
/
v
1
/
2
0
2
4
.
e
m
n
l
p
-
i
n
d
u
st
r
y
.
8
2
.
[
4
9
]
H
.
Q
i
u
,
H
.
H
e
,
S
.
Z
h
a
n
g
,
A
.
L
i
,
a
n
d
Z.
La
n
,
“
S
M
I
LE:
s
i
n
g
l
e
-
t
u
r
n
t
o
mu
l
t
i
-
t
u
r
n
i
n
c
l
u
si
v
e
l
a
n
g
u
a
g
e
e
x
p
a
n
s
i
o
n
v
i
a
C
h
a
t
G
P
T
f
o
r
m
e
n
t
a
l
h
e
a
l
t
h
s
u
p
p
o
r
t
,
”
i
n
Fi
n
d
i
n
g
s
o
f
t
h
e
Asso
c
i
a
t
i
o
n
f
o
r
C
o
m
p
u
t
a
t
i
o
n
a
l
L
i
n
g
u
i
s
t
i
c
s:
EM
N
L
P
2
0
2
4
,
2
0
2
4
,
p
p
.
6
1
5
–
6
3
6
,
d
o
i
:
1
0
.
1
8
6
5
3
/
v
1
/
2
0
2
4
.
f
i
n
d
i
n
g
s
-
e
mn
l
p
.
3
4
.
[
5
0
]
Q
.
Li
u
e
t
a
l
.
,
“
C
l
a
ss
l
i
f
e
l
o
n
g
l
e
a
r
n
i
n
g
f
o
r
i
n
t
e
n
t
d
e
t
e
c
t
i
o
n
v
i
a
s
t
r
u
c
t
u
r
e
c
o
n
s
o
l
i
d
a
t
i
o
n
n
e
t
w
o
r
k
s
,
”
i
n
F
i
n
d
i
n
g
s o
f
t
h
e
Asso
c
i
a
t
i
o
n
f
o
r
C
o
m
p
u
t
a
t
i
o
n
a
l
L
i
n
g
u
i
s
t
i
c
s:
A
C
L
2
0
2
3
,
2
0
2
3
,
p
p
.
2
9
3
–
306
,
d
o
i
:
1
0
.
1
8
6
5
3
/
v
1
/
2
0
2
3
.
f
i
n
d
i
n
g
s
-
a
c
l
.
2
0
.
[
5
1
]
S
.
A
y
a
n
o
u
z
,
B
.
A
.
A
b
d
e
l
h
a
k
i
m,
a
n
d
M
.
B
.
A
h
me
d
,
“
U
si
n
g
n
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
r
o
c
e
ss
i
n
g
t
o
e
v
a
l
u
a
t
e
t
h
e
i
mp
a
c
t
o
f
s
p
e
c
i
a
l
i
z
e
d
t
r
a
n
sf
o
r
mers
mo
d
e
l
s
o
n
me
d
i
c
a
l
d
o
m
a
i
n
t
a
s
k
s,
”
I
AE
S
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
Ar
t
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
,
v
o
l
.
1
3
,
n
o
.
2
,
p
.
1
7
3
2
,
Ju
n
.
2
0
2
4
,
d
o
i
:
1
0
.
1
1
5
9
1
/
i
j
a
i
.
v
1
3
.
i
2
.
p
p
1
7
3
2
-
1
7
4
0
.
[
5
2
]
S
.
X
u
e
a
n
d
F
.
R
e
n
,
“
I
n
t
e
n
t
-
e
n
h
a
n
c
e
d
a
t
t
e
n
t
i
v
e
B
E
R
T
c
a
p
s
u
l
e
n
e
t
w
o
r
k
f
o
r
z
e
r
o
-
sh
o
t
i
n
t
e
n
t
i
o
n
d
e
t
e
c
t
i
o
n
,
”
N
e
u
r
o
c
o
m
p
u
t
i
n
g
,
v
o
l
.
4
5
8
,
p
p
.
1
–
1
3
,
O
c
t
.
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
n
e
u
c
o
m.
2
0
2
1
.
0
5
.
0
8
5
.
[
5
3
]
N
.
K
u
m
a
r
a
n
d
B
.
K
.
B
a
g
h
e
l
,
“
S
mart
st
a
c
k
i
n
g
o
f
d
e
e
p
l
e
a
r
n
i
n
g
mo
d
e
l
s
f
o
r
g
r
a
n
u
l
a
r
j
o
i
n
t
i
n
t
e
n
t
-
sl
o
t
e
x
t
r
a
c
t
i
o
n
f
o
r
m
u
l
t
i
-
i
n
t
e
n
t
S
LU
,
”
I
EEE
A
c
c
e
ss
,
v
o
l
.
9
,
p
p
.
9
7
5
8
2
–
9
7
5
9
0
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
1
.
3
0
9
5
4
1
6
.
[
5
4
]
L.
H
u
a
n
g
,
S
.
L
i
a
n
g
,
F
.
Y
e
,
a
n
d
N
.
G
a
o
,
“
A
f
a
st
a
t
t
e
n
t
i
o
n
n
e
t
w
o
r
k
f
o
r
j
o
i
n
t
i
n
t
e
n
t
d
e
t
e
c
t
i
o
n
a
n
d
sl
o
t
f
i
l
l
i
n
g
o
n
e
d
g
e
d
e
v
i
c
e
s,
”
I
EEE
T
ra
n
s
a
c
t
i
o
n
s
o
n
Ar
t
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
,
v
o
l
.
5
,
n
o
.
2
,
p
p
.
5
3
0
–
5
4
0
,
F
e
b
.
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
TA
I
.
2
0
2
3
.
3
3
0
9
2
7
2
.
[
5
5
]
T.
W
u
,
M
.
W
a
n
g
,
Y
.
X
i
,
a
n
d
Z.
Z
h
a
o
,
“
I
n
t
e
n
t
r
e
c
o
g
n
i
t
i
o
n
m
o
d
e
l
b
a
s
e
d
o
n
s
e
q
u
e
n
t
i
a
l
i
n
f
o
r
ma
t
i
o
n
a
n
d
se
n
t
e
n
c
e
f
e
a
t
u
r
e
s,”
N
e
u
ro
c
o
m
p
u
t
i
n
g
,
v
o
l
.
5
6
6
,
p
.
1
2
7
0
5
4
,
Jan
.
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
n
e
u
c
o
m.
2
0
2
3
.
1
2
7
0
5
4
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
J
e
m
im
a
h
K.
p
u
rsu
i
n
g
P
h
.
D.
in
C
o
m
p
u
ter
S
c
ien
c
e
a
t
Bish
o
p
He
b
e
r
Co
ll
e
g
e
(Au
to
),
Affili
a
ted
t
o
B
h
a
ra
th
id
a
sa
n
Un
iv
e
rsit
y
,
Tr
ich
y
,
In
d
ia.
S
h
e
h
o
ld
s
a
M
a
ste
r
o
f
P
h
il
o
so
p
h
y
De
g
re
e
fro
m
B
h
a
ra
th
id
a
sa
n
Un
i
v
e
rsity
,
Tri
c
h
y
.
S
h
e
re
c
e
iv
e
d
h
e
r
M
.
S
c
.
a
n
d
B.
S
c
.
in
C
o
m
p
u
ter
S
c
ie
n
c
e
d
e
g
re
e
s
fro
m
Bish
o
p
He
b
e
r
Co
ll
e
g
e
(Au
t
o
).
He
r
re
se
a
rc
h
in
tere
st
in
c
lu
d
e
s
n
a
tu
ra
l
lan
g
u
a
g
e
p
r
o
c
e
ss
in
g
,
c
h
a
tb
o
t
tec
h
n
o
l
o
g
ies
a
n
d
lo
w
re
so
u
rc
e
lan
g
u
a
g
e
s
.
S
h
e
h
a
s
a
tt
e
n
d
e
d
m
a
n
y
in
tern
a
ti
o
n
a
l
c
o
n
fe
re
n
c
e
s
a
n
d
wo
rk
s
h
o
p
s
o
rg
a
n
i
z
e
d
b
y
ACM
In
d
ia.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
jem
ima
h
8
@g
m
a
il
.
c
o
m
.
Ra
jk
u
m
a
r
K
a
n
n
a
n
h
o
l
d
s
a
P
h
.
D.
in
Co
m
p
u
ter
Ap
p
l
ica
ti
o
n
s
fr
o
m
th
e
Na
ti
o
n
a
l
In
stit
u
te
o
f
Tec
h
n
o
lo
g
y
,
Ti
r
u
c
h
i
ra
p
p
a
ll
i,
In
d
ia
(2
0
0
7
).
He
is
c
u
rre
n
tl
y
t
h
e
Co
n
tr
o
ll
e
r
o
f
Ex
a
m
in
a
ti
o
n
s
a
n
d
P
ro
fe
ss
o
r
in
t
h
e
De
p
a
rtme
n
t
o
f
C
o
m
p
u
ter
S
c
i
e
n
c
e
(S
h
ift
-
I)
a
t
Bis
h
o
p
He
b
e
r
Co
ll
e
g
e
(Au
t
o
n
o
m
o
u
s),
I
n
d
ia.
Ad
d
it
i
o
n
a
ll
y
,
h
e
lea
d
s
t
h
e
Artifi
c
ial
In
telli
g
e
n
c
e
Re
se
a
rc
h
a
n
d
Tec
h
n
o
l
o
g
y
De
v
e
lo
p
m
e
n
t
Lab
a
t
t
h
e
i
n
stit
u
ti
o
n
.
His
re
se
a
rc
h
fo
c
u
se
s
o
n
in
fo
rm
a
ti
o
n
re
tri
e
v
a
l,
m
u
lt
ime
d
ia
se
m
a
n
ti
c
s,
d
a
ta
sc
i
e
n
c
e
,
a
n
d
c
o
ll
a
b
o
ra
ti
v
e
a
n
d
c
o
ll
e
c
ti
v
e
in
telli
g
e
n
c
e
.
P
re
v
io
u
sly
,
His
wa
s a
ss
o
c
iate
d
with
Kin
g
F
a
isa
l
Un
i
v
e
rsity
,
Kin
g
d
o
m
o
f
S
a
u
d
i
Ara
b
ia,
wh
e
re
h
e
tau
g
h
t
ABET
-
a
c
c
re
d
it
e
d
u
n
d
e
r
g
ra
d
u
a
te
a
n
d
p
o
stg
ra
d
u
a
te
p
r
o
g
ra
m
s
in
Co
m
p
u
ter
S
c
ien
c
e
a
n
d
I
n
fo
rm
a
ti
o
n
S
y
ste
m
s.
He
is
t
h
e
a
u
th
o
r
o
f
t
h
e
wid
e
l
y
u
se
d
tex
tb
o
o
k
Ja
v
a
P
ro
g
ra
m
m
in
g
(P
e
a
rso
n
)
a
n
d
h
a
s
e
d
it
e
d
v
o
l
u
m
e
s
fo
r
S
p
ri
n
g
e
r
LNCS
a
n
d
IG
I
G
lo
b
a
l
(USA).
His
is
th
e
P
ro
g
ra
m
Ch
a
ir
o
f
th
e
In
ter
n
a
ti
o
n
a
l
Co
n
fe
re
n
c
e
o
n
Da
ta
En
g
i
n
e
e
rin
g
a
n
d
M
a
n
a
g
e
m
e
n
t
se
ries
a
n
d
se
rv
e
s
o
n
e
d
it
o
rial
b
o
a
r
d
s
o
f
se
v
e
ra
l
i
n
tern
a
ti
o
n
a
l
j
o
u
r
n
a
ls.
He
is
a
se
n
io
r
m
e
m
b
e
r
o
f
ACM
(U
S
A),
l
ife
m
e
m
b
e
r
o
f
CS
I,
IS
T
E,
a
n
d
IF
I
P
WG
8
.
2
,
a
n
d
a
fo
rm
e
r
mem
b
e
r
o
f
IEE
E
a
n
d
BCS
(UK
).
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
ra
j
k
u
m
a
r@b
h
c
.
e
d
u
.
in
.
Fre
d
e
r
ic
And
r
e
s
re
c
e
iv
e
d
h
is
P
h
.
D.
i
n
In
f
o
rm
a
ti
o
n
S
y
ste
m
s
fro
m
th
e
Un
iv
e
rsit
y
o
f
P
a
ris
VI
"
P
ierre
e
t
M
a
rie
Cu
rie"
in
1
9
9
3
,
fo
ll
o
we
d
b
y
h
is
Dr.
Ha
b
il
it
a
ti
o
n
i
n
In
fo
rm
a
ti
c
s
fro
m
th
e
Un
i
v
e
rsity
o
f
Na
n
tes
i
n
2
0
0
0
.
S
in
c
e
2
0
0
0
,
h
e
h
a
s
b
e
e
n
a
n
a
ss
o
c
iate
p
ro
fe
ss
o
r
a
t
th
e
Dig
it
a
l
C
o
n
te
n
t
a
n
d
M
e
d
ia
S
c
ien
c
e
s
Re
se
a
rc
h
Div
isio
n
,
NII,
T
o
k
y
o
,
Ja
p
a
n
,
a
n
d
si
n
c
e
2
0
0
2
,
with
S
o
k
e
n
d
a
i
(G
ra
d
u
a
te
Un
iv
e
rs
it
y
f
o
r
Ad
v
a
n
c
e
d
S
t
u
d
ies
),
T
o
k
y
o
.
He
h
a
s
a
u
t
h
o
re
d
o
v
e
r
1
7
0
p
u
b
li
c
a
ti
o
n
s
a
c
ro
ss
i
n
tern
a
ti
o
n
a
l
jo
u
rn
a
ls,
b
o
o
k
s,
a
n
d
c
o
n
fe
re
n
c
e
s.
Hi
s
re
se
a
rc
h
in
tere
sts
c
o
ll
e
c
ti
v
e
in
telli
g
e
n
c
e
-
b
a
se
d
c
o
m
m
u
n
it
y
m
a
n
a
g
e
m
e
n
t,
e
a
rly
wa
rn
i
n
g
s
y
ste
m
s,
m
u
lse
m
e
d
ia,
c
o
o
k
i
n
g
re
c
ip
e
c
o
m
p
u
t
in
g
,
c
a
rb
o
n
fo
o
tp
r
in
t
o
p
t
imiz
a
ti
o
n
i
n
m
u
l
ti
-
d
ish
p
lan
n
in
g
,
a
n
d
wa
ter
c
ry
sta
l
c
o
m
p
u
t
in
g
.
He
a
n
d
h
is
t
e
a
m
e
a
r
n
e
d
th
e
1
st
p
rize
in
th
e
IEE
E
Bra
i
n
Da
ta
Ba
n
k
Ch
a
ll
e
n
g
e
s
a
t
COMP
S
AC
2
0
1
8
.
His
se
rv
e
s
a
s
co
-
e
d
it
o
r
-
in
-
c
h
ief
o
f
th
e
In
ter
n
a
ti
o
n
a
l
Jo
u
rn
a
l
o
f
Ag
ric
u
lt
u
ra
l
a
n
d
En
v
iro
n
m
e
n
ta
l
In
f
o
rm
a
ti
o
n
S
y
ste
m
s
a
lo
n
g
si
d
e
c
o
ll
e
a
g
u
e
s
fr
o
m
AIT
a
n
d
JA
IS
T.
He
c
o
-
o
r
g
a
n
ize
s
t
h
e
De
c
o
r
W
o
rk
s
h
o
p
se
ries
a
t
ICDE
c
o
n
fe
re
n
c
e
s
g
l
o
b
a
ll
y
.
As
a
m
e
m
b
e
r
o
f
Ja
p
a
n
’s
IS
O/IEC
JTC
1
/S
C
3
6
Co
m
m
it
tee
,
h
e
h
a
s
c
o
n
t
rib
u
te
d
a
s
e
d
it
o
r
t
o
k
e
y
sta
n
d
a
rd
s
o
n
st
u
d
e
n
t
d
a
ta
p
riv
a
c
y
a
n
d
se
m
a
n
ti
c
in
tero
p
e
ra
b
i
li
ty
i
n
e
-
lea
rn
i
n
g
.
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
a
n
d
re
s@
n
ii
.
a
c
.
jp
.
Evaluation Warning : The document was created with Spire.PDF for Python.