I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
pu
t
er
E
ng
ineering
(
I
J
E
CE
)
Vo
l.
15
,
No
.
2
,
A
p
r
il
20
25
,
p
p
.
2
3
2
5
~
2
3
3
3
I
SS
N:
2088
-
8
7
0
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijece.
v
15
i
2
.
pp
2
3
2
5
-
2
3
3
3
2325
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m
The
a
dv
a
nces
in
na
tural
la
ng
ua
g
e
pro
cess
ing
techno
lo
g
y
a
nd
its
impa
ct
on
mo
der
n
so
ciety
Na
dezhda
B
o
riso
v
a
,
E
lena
K
a
ra
s
htr
a
no
v
a
,
I
re
na
At
a
na
s
o
v
a
D
e
p
a
r
t
me
n
t
of
I
n
f
o
r
m
a
t
i
c
s,
F
a
c
u
l
t
y
of
M
a
t
h
e
m
a
t
i
c
s
a
n
d
N
a
t
u
r
a
l
S
c
i
e
n
c
e
s,
S
o
u
t
h
-
W
e
st
U
n
i
v
e
r
si
t
y
,
B
l
a
g
o
e
v
g
r
a
d
,
B
u
l
g
a
r
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
n
2
0
,
2
0
2
4
R
ev
is
ed
Sep
1
7
,
2
0
2
4
Acc
ep
ted
Oct
1
,
2
0
2
4
A
wid
e
ra
n
g
e
of
in
f
o
rm
a
ti
o
n
,
su
c
h
as
b
o
o
k
s,
n
e
ws
,
re
p
o
rts,
a
n
d
o
t
h
e
r
c
o
n
ten
t,
is
c
o
n
sta
n
tl
y
b
e
i
n
g
p
r
o
d
u
c
e
d
,
m
u
c
h
of
wh
ich
is
a
v
a
il
a
b
le
o
n
li
n
e
.
M
a
c
h
in
e
tran
sla
ti
o
n
,
sp
a
m
d
e
t
e
c
ti
o
n
,
n
a
tu
ra
l
la
n
g
u
a
g
e
in
terfa
c
e
s,
a
n
d
q
u
e
stio
n
-
a
n
sw
e
rin
g
a
p
p
li
c
a
ti
o
n
s
h
a
v
e
b
e
c
o
m
e
in
c
re
a
sin
g
l
y
p
o
p
u
lar.
Na
tu
ra
l
lan
g
u
a
g
e
p
r
o
c
e
ss
in
g
(NL
P
)
is
at
th
e
c
o
re
of
th
e
a
u
to
m
a
ti
c
r
e
tri
e
v
a
l
of
in
fo
rm
a
ti
o
n
sto
re
d
on
c
o
m
p
u
te
rs.
Th
is
a
rti
c
le
d
isc
u
ss
e
s
NLP
a
n
d
its
a
p
p
li
c
a
ti
o
n
s
in
d
a
il
y
a
c
ti
v
it
ies
.
It
c
o
v
e
rs
t
h
e
m
a
in
sta
g
e
s
of
NLP
a
n
d
p
ro
v
id
e
s
e
x
a
m
p
les
of
its
a
d
v
a
n
c
e
s
in
v
a
ri
o
u
s
h
ig
h
e
r
-
lev
e
l
t
a
sk
s.
Th
e
o
b
jec
ti
v
e
is
to
h
i
g
h
l
ig
h
t
t
h
e
sig
n
if
ica
n
c
e
of
NLP
in
p
ro
c
e
ss
in
g
o
n
li
n
e
c
o
n
te
n
t
a
n
d
in
e
fficie
n
t
i
n
tera
c
ti
o
n
s
b
e
tw
e
e
n
h
u
m
a
n
s
a
n
d
c
o
m
p
u
ters
a
c
ro
s
s
v
a
rio
u
s
a
p
p
li
c
a
ti
o
n
s.
As
an
e
ss
e
n
ti
a
l
e
le
m
e
n
t
of
a
rti
ficia
l
in
telli
g
e
n
c
e
,
N
LP
p
ro
v
id
e
s
so
lu
ti
o
n
s
fo
r
re
a
l
-
wo
r
ld
p
ro
b
lem
s
a
n
d
h
a
s
t
h
e
p
o
ten
ti
a
l
to
tran
sfo
r
m
th
e
wa
y
c
o
m
p
a
n
ies
o
p
e
ra
te
.
K
ey
w
o
r
d
s
:
C
o
n
ten
t
p
r
o
ce
s
s
in
g
L
an
g
u
ag
e
-
r
elate
d
task
s
Natu
r
al
lan
g
u
ag
e
p
r
o
ce
s
s
in
g
NL
P
ap
p
licatio
n
s
Un
s
tr
u
ctu
r
ed
tex
t
Th
is
is
an
o
p
e
n
a
c
c
e
ss
a
rti
c
le
u
n
d
e
r
th
e
CC
BY
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Nad
ez
h
d
a
B
o
r
is
o
v
a
Dep
ar
tm
en
t
of
I
n
f
o
r
m
atics,
Facu
lty
of
Ma
th
e
m
atics
an
d
Natu
r
al
Scien
ce
s
,
So
u
th
-
W
est
Un
iv
er
s
ity
66
I
v
a
n
Mic
h
ailo
v
B
u
lv
.
,
2
7
0
0
B
lag
o
ev
g
r
a
d
,
B
u
lg
ar
ia
E
m
ail:
n
b
o
r
is
o
v
a@
s
wu
.
b
g
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
in
ter
n
et
is
an
u
n
d
en
iab
l
e
f
ac
to
r
in
th
e
ev
e
r
y
d
a
y
life
of
m
o
d
er
n
s
o
ciety
.
T
h
is
g
lo
b
al
s
y
s
tem
p
r
o
v
id
es
i
n
f
o
r
m
atio
n
r
eso
u
r
ce
s
an
d
s
er
v
ices
to
b
illi
o
n
s
of
u
s
er
s
,
but
m
u
ch
of
t
h
e
av
aila
b
le
in
f
o
r
m
atio
n
is
in
tex
t
f
o
r
m
at
an
d
is
not
au
to
m
a
tically
p
r
o
ce
s
s
ed
an
d
in
ter
p
r
et
ed
.
T
h
is
m
ea
n
s
th
at
u
s
er
s
h
av
e
to
d
eter
m
in
e
th
e
m
ea
n
in
g
an
d
class
if
icatio
n
of
th
e
in
f
o
r
m
atio
n
th
e
y
r
ec
ei
v
e.
T
ec
h
n
iq
u
es
of
au
to
m
atic
n
atu
r
al
lan
g
u
ag
e
p
r
o
ce
s
s
in
g
(
NL
P)
h
av
e
b
ee
n
u
n
d
er
d
ev
elo
p
m
en
t
s
in
ce
th
e
1
9
5
0
s
,
an
d
in
r
ec
e
n
t
y
ea
r
s
,
p
r
o
g
r
ess
in
th
e
tech
n
o
lo
g
y
h
as
im
p
ac
ted
th
e
e
f
f
ec
tiv
en
ess
of
NL
P
ap
p
licatio
n
s
in
m
an
y
a
r
ea
s
,
in
clu
d
in
g
m
ac
h
in
e
tr
an
s
latio
n
,
s
p
am
r
ec
o
g
n
itio
n
,
n
at
u
r
al
lan
g
u
ag
e
in
ter
f
ac
es,
a
n
d
s
p
ee
ch
r
e
co
g
n
itio
n
.
Acc
o
r
d
in
g
to
L
id
d
y
[
1
]
,
“
NL
P
is
a
th
eo
r
etica
lly
m
o
tiv
ated
r
an
g
e
of
co
m
p
u
tatio
n
al
tech
n
iq
u
es
f
o
r
an
aly
zin
g
an
d
r
ep
r
esen
tin
g
n
atu
r
ally
o
cc
u
r
r
in
g
tex
ts
at
o
n
e
or
more
le
v
els
of
lin
g
u
is
tic
an
aly
s
is
to
ac
h
iev
e
h
u
m
an
-
lik
e
la
n
g
u
a
g
e
p
r
o
ce
s
s
in
g
f
o
r
a
r
a
n
g
e
of
task
s
or
ap
p
licatio
n
s
.
”
Fo
r
th
e
p
r
o
p
o
s
ed
d
ef
in
itio
n
,
s
ev
er
al
elem
en
ts
ar
e
f
u
r
th
e
r
d
etailed
:
−
"
R
an
g
e
of
co
m
p
u
tatio
n
al
tech
n
iq
u
es"
r
ef
er
s
to
th
e
v
ar
i
ety
of
m
eth
o
d
s
or
tech
n
i
q
u
es
th
at
can
be
im
p
lem
en
ted
in
d
if
f
er
e
n
t
ty
p
e
s
of
lin
g
u
is
tic
an
aly
s
is
.
−
"Na
tu
r
ally
o
cc
u
r
r
in
g
tex
ts
"
r
e
f
er
to
th
e
lan
g
u
ag
es
u
s
ed
by
h
u
m
an
s
to
co
m
m
u
n
icate
with
ea
ch
o
th
er
.
−
T
h
e
"lev
els
of
lin
g
u
is
tic
an
aly
s
is
"
r
ef
er
to
th
e
co
m
b
in
atio
n
s
of
d
if
f
er
e
n
t
lev
els
of
lin
g
u
is
tic
an
aly
s
is
u
s
e
d
by
v
ar
i
o
u
s
NL
P
s
y
s
tem
s
.
−
"Fo
r
a
r
a
n
g
e
of
task
s
or
ap
p
li
ca
tio
n
s
"
m
ea
n
s
th
at
NL
P
is
in
ten
d
ed
f
o
r
p
er
f
o
r
m
in
g
s
p
ec
if
i
c
task
s
,
s
u
ch
as
th
o
s
e
u
s
ed
in
s
y
s
tem
s
f
o
r
in
f
o
r
m
atio
n
r
etr
iev
al,
m
ac
h
in
e
tr
a
n
s
latio
n
,
an
d
q
u
esti
o
n
an
s
wer
in
g
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
2
,
Ap
r
il
20
25
:
2
3
2
5
-
2
3
3
3
2326
T
h
er
e
ar
e
two
m
ai
n
f
o
cu
s
es
in
NL
P
,
n
atu
r
al
lan
g
u
ag
e
u
n
d
er
s
tan
d
in
g
(
NL
U)
a
n
d
n
atu
r
al
lan
g
u
ag
e
g
en
er
atio
n
(
NL
G)
.
NL
U
is
s
i
m
ilar
to
th
e
r
o
le
of
a
r
ea
d
er
or
lis
ten
er
,
an
d
its
p
u
r
p
o
s
e
is
to
u
n
d
er
s
tan
d
an
d
cr
ea
te
a
m
ea
n
in
g
f
u
l
r
ep
r
esen
ta
tio
n
of
in
p
u
t
d
ata
in
n
atu
r
al
la
n
g
u
ag
e
.
NL
G,
also
k
n
o
wn
as
t
ex
t
g
en
e
r
atio
n
,
is
a
s
o
f
twar
e
p
r
o
ce
s
s
th
at
p
r
o
d
u
ce
s
o
u
tp
u
t
d
ata
in
n
atu
r
al
lan
g
u
a
g
e.
NL
U
an
d
NL
G
ca
n
be
s
ee
n
as
p
ar
ticip
an
ts
in
a
n
atu
r
al
co
m
m
u
n
icatio
n
p
r
o
c
ess
.
Ad
v
an
ce
s
in
NL
P
tech
n
o
lo
g
y
ar
e
d
u
e
to
ef
f
o
r
ts
en
ab
lin
g
m
ac
h
in
es
to
u
n
d
e
r
s
tan
d
,
in
t
er
p
r
et,
an
d
g
en
er
ate
h
u
m
an
lan
g
u
ag
e
ac
cu
r
ately
an
d
ef
f
ec
tiv
ely
.
NL
P
tech
n
o
lo
g
y
aim
s
to
lead
to
more
ef
f
ec
tiv
e
in
ter
ac
tio
n
s
b
etwe
en
h
u
m
an
s
an
d
c
o
m
p
u
te
r
s
in
v
ar
io
u
s
ap
p
l
icatio
n
s
.
T
h
is
in
clu
d
es
m
ee
tin
g
th
e
ch
allen
g
es
of
p
r
o
ce
s
s
in
g
d
if
f
er
e
n
t
lan
g
u
ag
e
s
an
d
co
m
p
lex
lan
g
u
a
g
e
s
tr
u
ctu
r
es,
as
well
as
u
n
d
er
s
tan
d
in
g
co
n
tex
t
in
b
o
t
h
wr
itten
an
d
s
p
o
k
e
n
f
o
r
m
s
.
A
d
v
an
ce
m
e
n
ts
in
NL
P
h
av
e
i
m
p
r
o
v
e
d
m
ac
h
in
e
t
r
an
s
latio
n
,
s
p
ee
ch
r
ec
o
g
n
itio
n
,
an
d
tex
t
g
en
er
atio
n
,
im
p
ac
t
in
g
lan
g
u
ag
e
b
ar
r
ier
s
an
d
en
ab
lin
g
g
lo
b
al
c
o
m
m
u
n
icatio
n
[
2
]
–
[
7
]
.
T
h
e
d
ev
elo
p
m
e
n
t
of
s
o
p
h
is
ticated
ch
atb
o
ts
an
d
v
ir
tu
al
ass
is
tan
ts
d
r
iv
en
by
NL
P
allo
ws
b
u
s
in
ess
es
to
p
r
o
v
id
e
2
4
/
7
cu
s
to
m
er
s
u
p
p
o
r
t,
en
h
a
n
cin
g
cu
s
to
m
er
s
atis
f
ac
tio
n
[
8
]
–
[
1
0
]
.
NL
P
m
o
d
els
lik
e
GPT
-
3
an
d
GPT
-
4
ca
n
g
en
er
ate
h
u
m
a
n
-
lik
e
te
x
t,
wr
ite
ar
ticles,
an
d
s
u
m
m
ar
ize
d
o
cu
m
en
ts
[
1
1
]
–
[
1
4
]
.
T
h
ese
m
o
d
els
can
p
r
o
v
id
e
tailo
r
ed
f
ee
d
b
ac
k
,
im
p
r
o
v
in
g
ed
u
ca
tio
n
al
o
u
tco
m
es
an
d
m
a
k
in
g
lear
n
in
g
more
en
g
a
g
in
g
[
1
5
]
–
[
1
8
]
.
NL
P
to
o
ls
can
an
aly
ze
v
ast
am
o
u
n
ts
of
tex
tu
al
d
ata,
s
u
ch
as
s
o
cial
m
ed
ia
p
o
s
ts
,
r
ev
iews,
an
d
s
u
r
v
ey
s
[
1
9
]
–
[
2
3
]
.
T
h
e
ad
v
an
tag
e
f
o
r
b
u
s
in
ess
es
is
th
at
th
ey
ca
n
g
ai
n
d
ee
p
in
s
ig
h
ts
in
to
c
u
s
to
m
er
o
p
in
io
n
s
an
d
m
ar
k
et
tr
e
n
d
s
,
an
d
en
ab
le
d
ata
-
d
r
iv
e
n
d
ec
is
io
n
-
m
ak
in
g
.
As
NL
P
tech
n
o
lo
g
ie
s
co
n
tin
u
e
to
ev
o
l
v
e,
th
ey
h
o
ld
th
e
p
o
ten
tial
to
f
u
r
th
er
t
r
an
s
f
o
r
m
in
ter
ac
tio
n
s
b
etwe
en
h
u
m
a
n
s
an
d
c
o
m
p
u
te
r
s
.
2.
NL
P
T
E
CH
NO
L
O
G
Y
B
AC
K
G
RO
UND
NL
P
is
o
r
g
an
ized
in
to
s
ix
m
ai
n
s
tag
es,
r
an
g
in
g
f
r
o
m
th
e
s
i
m
p
lest
to
th
e
m
o
s
t
ch
allen
g
in
g
,
as
s
h
o
wn
in
Fig
u
r
e
1.
Mo
r
p
h
o
lo
g
y
s
tu
d
i
es
th
e
s
m
allest
u
n
its
of
m
ea
n
in
g
in
a
lan
g
u
ag
e,
k
n
o
w
n
as
m
o
r
p
h
em
es,
a
n
d
th
eir
ar
r
an
g
em
e
n
t
with
in
wo
r
d
s
.
T
h
e
m
o
r
p
h
o
lo
g
ical
s
tr
u
ctu
r
e
of
th
e
w
o
r
d
m
ay
in
clu
d
e
p
r
ef
ix
es,
r
o
o
ts
,
s
u
f
f
ix
es,
cir
cu
m
f
ix
es,
an
d
in
f
i
x
es
th
at
m
o
d
if
y
its
m
ea
n
in
g
.
In
NL
P
,
s
y
s
tem
s
aim
to
u
n
d
er
s
tan
d
th
e
m
ea
n
in
g
of
wo
r
d
s
by
r
ec
o
g
n
izin
g
th
e
m
ea
n
in
g
of
th
eir
co
n
s
titu
en
t
m
o
r
p
h
em
es.
Fo
r
ex
am
p
le,
th
e
wo
r
d
"k
in
d
n
ess
"
co
n
s
is
ts
of
two
p
ar
ts
:
"k
in
d
"
an
d
"
n
ess
.
"
E
ac
h
m
o
r
p
h
em
e
h
as
its
o
wn
m
ea
n
in
g
:
"k
in
d
"
is
a
f
r
ee
m
o
r
p
h
em
e
b
ec
a
u
s
e
it
can
s
tan
d
alo
n
e
as
a
wo
r
d
,
wh
ile
"
n
ess
"
is
a
bound
m
o
r
p
h
em
e.
"
Ness
"
is
a
s
u
f
f
ix
th
at
cr
ea
tes
n
o
u
n
s
an
d
in
d
icate
s
a
s
tate,
co
n
d
itio
n
,
or
q
u
ality
[
2
4
]
th
at
ca
n
n
o
t
ex
is
t
on
its
o
wn
.
Fig
u
r
e
1.
Six
lev
els
of
NL
P
[
1
]
L
ex
ical
an
aly
s
is
is
th
e
p
r
o
ce
s
s
of
d
eter
m
in
i
n
g
t
h
e
m
ea
n
in
g
of
wo
r
d
s
an
d
p
h
r
ases
in
a
tex
t.
T
h
is
p
r
o
ce
s
s
h
elp
s
to
d
etec
t
th
e
r
elatio
n
s
h
ip
b
etwe
en
in
d
iv
id
u
al
wo
r
d
s
in
th
e
s
en
ten
ce
an
d
in
v
o
lv
es
b
r
ea
k
in
g
th
e
tex
t
d
o
wn
in
to
m
ea
n
in
g
f
u
l
el
em
en
ts
s
u
ch
as
wo
r
d
s
,
p
ar
a
g
r
ap
h
s
,
a
n
d
s
en
ten
ce
s
.
T
h
e
p
o
ten
tial
p
ar
t
of
th
e
s
p
ee
ch
of
a
w
o
r
d
is
also
ass
ig
n
ed
in
th
e
lex
ical
an
al
y
s
is
.
A
wo
r
d
ca
n
h
a
v
e
d
if
f
er
e
n
t
m
ea
n
in
g
s
an
d
be
a
n
o
u
n
or
ad
jectiv
e
,
b
u
t
it
i
s
p
ar
t
of
s
p
ee
ch
a
n
d
le
x
ical
m
ea
n
i
n
g
ca
n
o
n
ly
be
d
e
d
u
ce
d
in
co
n
te
x
t
with
o
th
er
wo
r
d
s
in
th
e
p
h
r
ase
or
s
en
ten
ce
.
T
h
is
le
v
el
m
ay
u
tili
ze
a
co
llectio
n
wi
th
a
h
ig
h
ly
s
tr
u
ctu
r
e
d
f
o
r
m
of
lan
g
u
ag
e
lex
em
es,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
Th
e
a
d
va
n
c
es in
n
a
tu
r
a
l la
n
g
u
a
g
e
p
r
o
ce
s
s
in
g
tech
n
o
lo
g
y
a
n
d
its
imp
a
ct
o
n
… (
N
a
d
ezh
d
a
B
o
r
is
o
va
)
2327
ca
lled
a
lex
ico
n
.
A
lex
em
e
is
a
m
in
im
al
u
n
it
r
ep
r
esen
ted
in
th
e
lex
ico
n
th
at
p
air
s
th
e
c
o
llectio
n
of
f
o
r
m
s
(
s
en
s
es)
th
at
a
s
in
g
le
m
o
r
p
h
e
m
e
can
tak
e.
L
ex
ical
an
aly
s
is
is
o
f
ten
in
v
o
lv
ed
in
th
e
in
itial
s
tag
e
of
m
an
y
NL
P
task
s
an
d
is
f
o
llo
wed
by
s
y
n
ta
x
an
aly
s
is
an
d
s
em
an
tic
an
aly
s
is
.
Sy
n
tactic
or
s
y
n
tax
a
n
aly
s
is
,
also
k
n
o
wn
as
"p
ar
s
in
g
,
"
is
u
s
ed
to
an
al
y
ze
th
e
wo
r
d
s
in
a
s
en
ten
ce
ac
co
r
d
in
g
to
f
o
r
m
al
g
r
am
m
a
r
r
u
les.
Gr
am
m
atica
l
r
u
les
a
r
e
ap
p
lied
to
ca
teg
o
r
ize
an
d
g
r
o
u
p
wo
r
d
s
in
to
p
h
r
ases
.
It
in
v
o
l
v
es
id
en
tify
in
g
th
e
p
ar
ts
of
s
p
ee
ch
(
POS),
d
eter
m
in
in
g
th
e
r
elatio
n
s
h
ip
s
b
etwe
en
th
em
(
s
u
c
h
as
s
u
b
ject
-
v
er
b
ag
r
ee
m
en
t)
,
an
d
co
n
s
tr
u
ctin
g
a
p
a
r
s
e
tr
ee
to
r
ep
r
esen
t
th
e
h
ier
ar
ch
ic
al
s
tr
u
ctu
r
e
of
th
e
s
en
ten
ce
.
T
h
e
d
iag
r
am
in
Fig
u
r
e
2
s
h
o
ws
a
p
ar
s
e
tr
ee
f
o
r
th
e
s
en
ten
ce
"He
n
r
y
r
ea
d
th
e
book"
u
s
in
g
a
to
p
-
d
o
wn
p
a
r
s
in
g
m
ec
h
a
n
is
m
.
Fig
u
r
e
2.
A
p
ar
s
e
tr
ee
f
o
r
th
e
s
en
ten
ce
“He
n
r
y
r
ea
d
t
h
e
b
o
o
k
”
In
to
p
-
d
o
wn
p
ar
s
in
g
,
tr
ee
c
o
n
s
tr
u
ctio
n
s
ar
e
p
ar
s
ed
f
r
o
m
th
e
to
p
to
th
e
b
o
tto
m
,
an
d
t
h
e
in
p
u
t
is
r
ea
d
f
r
o
m
lef
t
to
r
ig
h
t.
T
h
e
r
o
o
t
n
o
d
e
is
g
en
er
ated
f
ir
s
t,
f
o
llo
we
d
by
th
e
f
o
r
m
atio
n
of
th
e
leaf
n
o
d
es.
To
tag
th
e
wo
r
d
s
with
th
eir
ap
p
r
o
p
r
iate
POS
tag
s
,
th
e
f
o
llo
win
g
s
et
of
g
r
am
m
ar
r
u
les
is
u
s
ed
:
−
S
-
>
NP
VP,
wh
er
e
S
(
s
en
ten
ce
)
,
NP
(
n
o
u
n
p
h
r
ase)
,
an
d
VP
(
v
er
b
p
h
r
ase)
−
NP
-
>
PP
N,
wh
er
e
PPN
(
p
r
o
p
er
n
o
u
n
)
−
VP
-
>
V
NP,
wh
er
e
VP
(
v
er
b
p
h
r
ase)
an
d
V
(
v
e
r
b
)
−
NP
-
>
Dete
r
m
in
er
N
−
PPN
-
>
Hen
r
y
−
V
-
>
r
ea
d
−
Dete
r
m
in
er
-
>
th
e
−
N
-
>
b
o
o
k
Sy
n
tax
an
aly
s
is
is
a
cr
u
cial
s
t
ep
in
en
s
u
r
in
g
th
at
th
e
s
tr
u
ctu
r
e
of
a
s
en
ten
ce
is
s
y
n
tactica
lly
co
r
r
ec
t.
Sy
n
tax
tech
n
iq
u
es
s
u
ch
as
s
tem
m
in
g
an
d
lem
m
atiza
tio
n
a
r
e
f
r
eq
u
en
tly
u
s
ed
to
u
n
d
e
r
s
tan
d
an
d
in
ter
p
r
et
th
e
m
ea
n
in
g
of
th
e
tex
t.
B
o
th
te
ch
n
iq
u
es
r
e
d
u
ce
wo
r
d
co
m
p
l
ex
ity
to
s
im
p
le
r
f
o
r
m
s
with
less
v
ar
iatio
n
[
2
5
]
.
Stem
m
in
g
u
s
es
alg
o
r
ith
m
s
to
m
atch
p
atter
n
s
a
n
d
r
em
o
v
e
s
u
f
f
ix
es
to
r
ed
u
ce
an
in
f
lecte
d
wo
r
d
to
its
s
tem
.
L
em
m
atiza
tio
n
u
s
es
a
d
ictio
n
ar
y
to
r
e
d
u
ce
th
e
w
o
r
d
to
its
m
ea
n
in
g
f
u
l
b
ase
f
o
r
m
,
ca
lled
a
lem
m
a.
C
o
n
s
tr
u
ctin
g
th
e
d
ictio
n
a
r
y
f
o
r
th
e
lem
m
atiza
tio
n
alg
o
r
ith
m
r
eq
u
ir
es
a
d
ee
p
u
n
d
er
s
tan
d
in
g
of
lin
g
u
is
tics
,
u
n
lik
e
s
tem
m
in
g
.
Fo
r
ex
am
p
le
,
th
e
r
esu
lts
f
o
r
th
e
wo
r
d
"stu
d
ies"
ar
e
as
f
o
llo
ws:
−
Stem
m
in
g
:
s
u
f
f
ix
=
“
-
es
”,
s
tem
=
“stu
d
i”
−
L
em
m
atiza
tio
n
:
th
ir
d
p
er
s
o
n
,
s
in
g
u
lar
n
u
m
b
e
r
,
p
r
esen
t
te
n
s
e
of
th
e
v
e
r
b
s
tu
d
y
;
lem
m
a=
“stu
d
y
”.
T
h
e
f
ir
s
t
s
tep
in
s
em
an
tic
a
n
aly
s
is
,
k
n
o
wn
as
lex
ical
s
e
m
an
tics
,
in
v
o
lv
es
ex
am
in
in
g
d
ictio
n
ar
y
d
ef
in
itio
n
s
of
wo
r
d
s
.
To
u
n
d
er
s
tan
d
th
e
m
ea
n
i
n
g
of
wo
r
d
s
in
co
n
tex
t,
th
e
r
elatio
n
s
h
ip
s
b
etwe
en
in
d
iv
id
u
al
wo
r
d
s
ar
e
an
aly
ze
d
by
ex
a
m
in
in
g
t
h
e
g
r
am
m
atica
l
s
tr
u
ctu
r
e
of
a
s
p
ec
if
ic
s
en
te
n
ce
.
Fo
r
ex
am
p
le,
s
em
an
ti
c
an
aly
s
is
p
r
o
v
id
es
co
n
te
x
t
f
o
r
u
n
d
er
s
tan
d
in
g
th
e
f
o
llo
win
g
s
e
n
ten
ce
s
:
−
Hen
r
y
k
n
o
ws
a
little
B
u
lg
ar
ian
,
d
ef
in
i
n
g
B
u
lg
ar
ian
as
a
lan
g
u
ag
e.
−
Hen
r
y
k
n
o
ws
a
s
m
all
B
u
lg
ar
ian
,
d
ef
in
i
n
g
B
u
lg
ar
ian
as
a
n
ati
o
n
ality
.
Sem
an
tic
an
aly
s
is
in
v
o
lv
es
d
i
s
am
b
ig
u
atin
g
wo
r
d
s
with
m
u
l
tip
le
m
ea
n
in
g
s
.
T
h
is
p
r
o
ce
s
s
s
elec
ts
th
e
s
in
g
le
m
ea
n
in
g
of
p
o
ly
s
em
o
u
s
wo
r
d
s
in
to
a
s
em
an
tic
r
e
p
r
esen
tatio
n
of
th
e
s
en
ten
ce
.
Fo
r
ex
am
p
le,
th
e
s
en
ten
ce
"He
n
r
y
s
aw
th
e
m
an
with
th
e
telesco
p
e"
is
g
r
a
m
m
atica
lly
co
r
r
ec
t
but
am
b
i
g
u
o
u
s
b
ec
au
s
e
it
i
s
u
n
clea
r
wh
eth
er
Hen
r
y
u
s
ed
a
telesco
p
e
to
s
ee
th
e
m
an
or
wh
eth
er
th
e
m
an
,
he
s
aw
was
h
o
ld
in
g
a
telesco
p
e.
Var
io
u
s
m
eth
o
d
s
(
e.
g
.
,
c
o
n
s
id
er
in
g
th
e
f
r
e
q
u
en
c
y
of
s
en
s
e
in
a
g
iv
en
co
r
p
u
s
or
tak
in
g
in
t
o
ac
co
u
n
t
th
e
lo
ca
l
co
n
tex
t)
can
be
u
tili
ze
d
f
o
r
d
i
s
am
b
ig
u
atio
n
[
2
6
]
.
Sem
an
tic
an
aly
s
is
is
a
co
m
p
lex
p
r
o
ce
s
s
th
at
aim
s
to
im
itat
e
h
u
m
an
u
n
d
e
r
s
tan
d
in
g
of
wo
r
d
s
,
p
h
r
ases
,
an
d
s
en
ten
ce
m
ea
n
i
n
g
s
,
p
o
s
in
g
an
o
n
g
o
in
g
ch
alle
n
g
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
2
,
Ap
r
il
20
25
:
2
3
2
5
-
2
3
3
3
2328
In
th
e
p
h
ase
of
d
is
co
u
r
s
e
in
teg
r
atio
n
,
th
e
m
ea
n
in
g
of
a
s
en
ten
ce
is
d
eter
m
in
ed
by
th
e
m
ea
n
in
g
of
th
e
s
en
ten
ce
s
th
at
p
r
ec
ed
e
it.
In
t
u
r
n
,
th
e
m
ea
n
in
g
of
th
e
c
u
r
r
e
n
t
s
en
ten
ce
im
p
ac
ts
th
e
m
ea
n
i
n
g
of
t
h
e
f
o
llo
win
g
s
en
ten
ce
.
An
ex
a
m
p
le
of
p
r
o
c
ess
in
g
in
th
is
p
h
ase
is
an
ap
h
o
r
a
r
eso
lu
tio
n
.
An
ap
h
o
r
a
r
eso
l
u
tio
n
is
th
e
p
r
o
ce
s
s
of
id
e
n
tify
in
g
th
e
li
n
k
b
etwe
en
a
r
ep
ea
ted
r
ef
er
e
n
ce
(
a
n
a
n
a
p
h
o
r
,
m
o
s
t
c
o
m
m
o
n
ly
in
th
e
f
o
r
m
of
a
p
r
o
n
o
u
n
)
an
d
th
e
p
r
ev
io
u
s
m
en
tio
n
of
t
h
e
en
tity
to
wh
ich
it
r
ef
er
s
.
An
o
th
er
ex
am
p
le
of
p
r
o
ce
s
s
in
g
in
th
is
p
h
ase
is
th
e
r
ec
o
g
n
itio
n
of
d
is
co
u
r
s
e
s
tr
u
ctu
r
e.
Dis
co
u
r
s
e
s
tr
u
ctu
r
e
can
p
r
ed
ict
wh
er
e
s
p
ec
if
ic
in
f
o
r
m
atio
n
will
be
f
o
u
n
d
in
s
tr
u
ctu
r
ed
d
o
cu
m
en
ts
with
f
o
r
m
al
p
r
o
p
er
ties
[
2
7
]
s
u
c
h
as
in
tr
o
d
u
ctio
n
,
th
esis
,
m
ain
id
e
a,
co
n
cl
u
s
io
n
,
a
n
d
r
ef
er
en
ce
s
.
Pra
g
m
atic
an
aly
s
is
is
th
e
f
in
a
l
lev
el
of
NL
P,
f
o
cu
s
in
g
on
t
h
e
in
ten
tio
n
al
u
s
e
of
lan
g
u
ag
e
in
r
ea
l
-
wo
r
ld
s
itu
atio
n
s
to
e
x
tr
ac
t
ad
d
itio
n
al
m
ea
n
in
g
by
ex
a
m
in
in
g
th
e
co
n
tex
t.
It
o
p
e
r
ates
b
ased
on
a
s
et
of
r
u
les
th
at
o
u
tlin
e
co
o
p
er
ativ
e
d
ial
o
g
u
es
a
n
d
r
elies
on
an
u
n
d
er
s
tan
d
in
g
of
th
e
g
en
er
al
p
r
in
cip
les
of
h
u
m
a
n
co
m
m
u
n
icatio
n
to
d
is
ce
r
n
th
e
in
ten
d
ed
ef
f
ec
t.
Fo
r
in
s
tan
ce
,
t
h
e
p
h
r
ase
"Op
en
th
e
win
d
o
w"
is
in
ter
p
r
eted
as
a
r
eq
u
est
r
ath
er
t
h
an
an
o
r
d
e
r
w
h
en
u
s
ed
in
a
s
en
ten
ce
[
2
8
]
.
Ph
o
n
o
lo
g
y
is
a
p
r
elim
in
ar
y
s
tep
th
at
o
cc
u
r
s
b
ef
o
r
e
m
o
r
p
h
o
l
o
g
ical
an
al
y
s
is
.
It
in
v
o
lv
es
th
e
s
tu
d
y
of
s
p
ee
ch
s
o
u
n
d
s
in
h
u
m
an
lan
g
u
ag
e
an
d
how
th
ey
ar
e
tr
an
s
f
o
r
m
ed
in
to
r
ec
o
g
n
izab
le
m
o
r
p
h
em
es
an
d
wo
r
d
s
.
Fo
llo
win
g
th
is
tr
an
s
f
o
r
m
atio
n
,
th
e
s
ix
lev
els
of
p
r
o
ce
s
s
in
g
m
en
tio
n
ed
ea
r
lier
a
r
e
u
tili
ze
d
.
T
h
e
in
itial
lev
els
f
o
cu
s
on
s
m
aller
u
n
its
of
an
aly
s
is
an
d
h
av
e
b
ee
n
e
x
p
lo
r
ed
in
g
r
ea
ter
d
etail.
Mo
r
eo
v
er
,
th
ese
u
n
its
ar
e
g
o
v
er
n
ed
by
r
u
les,
wh
ich
ar
e
more
f
ea
s
ib
le
to
im
p
lem
en
t.
T
h
is
is
th
e
m
ain
r
ea
s
o
n
wh
y
N
L
P
s
y
s
tem
s
ten
d
to
m
o
s
tly
im
p
lem
en
t
m
o
d
u
les
f
o
r
p
r
o
ce
s
s
in
g
at
l
o
wer
lev
els.
Var
io
u
s
NPL
task
s
can
be
p
er
f
o
r
m
ed
in
d
if
f
e
r
en
t
s
ec
to
r
s
,
but
a
s
et
of
f
u
n
d
am
en
tal
task
s
r
ec
u
r
f
r
eq
u
en
tly
,
an
d
th
at
s
et
is
m
u
ch
more
in
-
d
ep
th
s
tu
d
ie
d
an
d
r
eliab
ly
im
p
lem
en
te
d
.
T
h
e
f
o
ll
o
win
g
s
ec
tio
n
lis
ts
s
o
m
e
of
th
e
ad
d
itio
n
al
r
esear
ch
task
s
d
o
n
e
in
NL
P.
3.
H
I
G
H
E
R
-
L
E
VE
L
T
ASK
S
IN
NATUR
AL
L
ANG
UAG
E
P
RO
CE
SS
I
NG
T
h
is
s
ec
tio
n
o
u
tlin
es
th
e
k
ey
task
s
in
NL
P,
with
t
h
eir
o
b
jectiv
es
an
d
a
p
p
licatio
n
s
t
h
at
d
r
iv
e
its
ad
v
an
ce
m
e
n
ts
.
It
also
h
i
g
h
lig
h
ts
th
e
im
p
ac
t
of
th
e
o
r
etica
l
a
d
v
an
ce
m
e
n
ts
on
p
r
ac
tical
im
p
r
o
v
em
en
ts
in
NL
P
task
s
.
NL
P
h
as
a
wid
e
r
an
g
e
of
ap
p
licatio
n
s
,
an
d
r
esear
ch
e
d
ar
ea
s
in
clu
d
e
m
ac
h
i
n
e
tr
an
s
latio
n
,
em
ail
s
p
am
d
etec
tio
n
,
in
f
o
r
m
atio
n
ex
tr
ac
ti
o
n
,
s
u
m
m
ar
izatio
n
,
an
d
q
u
esti
o
n
an
s
wer
in
g
.
L
an
g
u
ag
e
m
o
d
elin
g
s
ee
k
s
to
p
r
ed
ict
s
eq
u
en
ce
s
of
r
ec
o
g
n
ized
wo
r
d
s
an
d
p
h
o
n
em
es.
Fo
r
th
is
p
u
r
p
o
s
e,
b
ased
on
th
e
h
is
to
r
y
of
p
r
e
v
io
u
s
wo
r
d
s
,
a
p
r
o
b
ab
ili
ty
of
th
e
n
ex
t
wo
r
d
is
ass
ig
n
ed
in
th
e
s
eq
u
e
n
ce
.
Dif
f
er
en
t
ty
p
es
of
lan
g
u
ag
e
m
o
d
els,
in
clu
d
in
g
s
tatis
tical,
n
eu
r
al
n
etwo
r
k
s
,
an
d
d
ee
p
lea
r
n
in
g
,
a
r
e
u
tili
ze
d
to
r
eso
lv
e
th
e
am
b
ig
u
ity
in
lo
n
g
er
s
en
ten
ce
s
wh
en
p
r
o
ce
s
s
ed
with
r
ea
lis
tic
g
r
am
m
ar
[
2
9
]
.
T
h
e
lan
g
u
a
g
e
m
o
d
els
th
at
h
av
e
b
ee
n
d
ev
elo
p
ed
f
i
n
d
ap
p
licatio
n
s
in
m
a
n
y
N
L
P
task
s
,
s
u
ch
as
m
ac
h
in
e
tr
an
s
latio
n
,
s
p
ee
ch
r
ec
o
g
n
itio
n
,
co
n
ten
t
g
e
n
er
ati
o
n
,
a
n
d
q
u
esti
o
n
a
n
s
wer
in
g
.
L
an
g
u
a
g
e
m
o
d
elin
g
p
lay
s
a
cr
itical
r
o
le
in
im
p
r
o
v
in
g
th
e
ca
p
ab
ilit
ies
of
ch
atb
o
ts
an
d
v
ir
tu
al
ass
is
tan
ts
to
u
n
d
er
s
tan
d
an
d
g
e
n
er
at
e
tex
t
s
im
ilar
to
a
h
u
m
an
.
T
ex
t
ca
teg
o
r
izatio
n
ass
ig
n
s
p
r
ed
ef
in
e
d
ca
te
g
o
r
ies
or
i
n
d
i
ce
s
to
d
o
cu
m
e
n
ts
b
ased
on
th
eir
tex
t
co
n
ten
t.
T
h
is
p
r
o
ce
s
s
is
u
s
ed
in
ap
p
licatio
n
s
s
u
ch
as
em
ail
s
p
am
f
ilter
s
.
T
h
e
ch
allen
g
e
of
NL
P
tech
n
o
lo
g
y
in
th
ese
ap
p
licatio
n
s
is
to
ex
tr
ac
t
m
ea
n
in
g
f
r
o
m
tex
t
an
d
th
en
c
lass
if
y
a
m
ess
ag
e
as
s
p
am
.
Sev
er
al
ty
p
es
of
s
p
am
f
ilter
s
[
3
0
]
ar
e
u
s
ed
f
o
r
th
is
p
u
r
p
o
s
e,
in
clu
d
in
g
c
o
n
ten
t
f
ilt
er
s
an
d
h
ea
d
er
f
ilter
s
,
wh
ich
lo
o
k
f
o
r
co
u
n
ter
f
eit
in
f
o
r
m
atio
n
in
th
e
co
n
ten
t
or
h
ea
d
er
of
an
em
ail.
Ad
d
itio
n
ally
,
th
er
e
ar
e
g
en
er
al
b
lack
lis
t
f
ilter
s
an
d
r
u
le
-
b
ased
f
ilter
s
,
wh
ich
s
to
p
em
ails
f
r
o
m
ad
d
r
ess
es
in
clu
d
ed
in
t
h
e
"b
lack
lis
t"
or
co
r
r
esp
o
n
d
to
cr
iter
ia
d
ef
in
ed
by
th
e
u
s
er
,
s
u
ch
as
in
clu
d
in
g
a
p
ar
ticu
lar
wo
r
d
.
Per
m
is
s
io
n
f
i
lter
s
r
eq
u
ir
e
th
e
s
en
d
e
r
to
be
p
r
e
-
ap
p
r
o
v
e
d
by
th
e
r
ec
ip
ien
t,
wh
ile
c
h
allen
g
e
r
esp
o
n
s
e
f
ilter
s
r
e
q
u
ir
e
th
e
s
en
d
er
to
e
n
ter
a
co
d
e
to
g
ain
p
er
m
is
s
io
n
to
s
en
d
th
e
em
ail.
I
n
f
o
r
m
atio
n
ex
tr
ac
tio
n
r
ef
e
r
s
to
u
s
in
g
co
m
p
u
tatio
n
al
m
eth
o
d
s
to
ex
tr
ac
t
r
elev
a
n
t
p
iece
s
of
in
f
o
r
m
atio
n
f
r
o
m
u
n
s
tr
u
ctu
r
ed
or
s
em
i
-
s
tr
u
ctu
r
ed
tex
t
d
o
cu
m
e
n
ts
.
E
x
am
p
les
of
th
ese
ty
p
es
of
in
p
u
t
d
o
cu
m
e
n
ts
in
clu
d
e
em
ails
,
web
p
ag
es,
n
ews
ar
ticles,
r
esea
r
ch
p
ap
er
s
,
an
d
s
o
cial
m
ed
ia,
am
o
n
g
o
th
e
r
s
,
an
d
th
e
o
u
tp
u
t
c
o
n
s
is
ts
of
th
e
p
er
tin
en
t
in
f
o
r
m
atio
n
f
r
o
m
th
e
d
o
cu
m
e
n
ts
ac
co
r
d
in
g
to
s
p
ec
if
ic
cr
iter
ia.
I
n
f
o
r
m
atio
n
e
x
tr
ac
tio
n
tech
n
o
l
o
g
ies
ar
e
u
s
ed
to
an
aly
ze
d
o
c
u
m
en
ts
,
u
n
co
v
e
r
r
ele
v
an
t
k
n
o
wled
g
e,
an
d
p
r
esen
t
it
in
a
s
tr
u
ct
u
r
ed
f
o
r
m
at
[
3
1
]
.
A
lar
g
e
am
o
u
n
t
of
in
f
o
r
m
atio
n
av
ailab
le
n
ec
ess
itates
th
e
u
s
e
of
au
to
m
atic
a
n
d
ef
f
icien
t
to
o
ls
f
o
r
in
f
o
r
m
atio
n
ex
tr
ac
tio
n
an
d
r
etr
iev
al
to
co
m
p
en
s
ate
f
o
r
h
u
m
an
in
ef
f
ec
tiv
e
n
ess
.
T
h
e
p
r
i
m
ar
y
g
o
al
of
an
i
n
f
o
r
m
atio
n
r
etr
ie
v
al
(
I
R
)
s
y
s
tem
is
to
p
r
o
v
id
e
a
lis
t
of
r
elate
d
d
o
cu
m
en
ts
in
r
esp
o
n
s
e
to
a
u
s
er
q
u
er
y
.
T
r
a
d
itio
n
al
s
ea
r
ch
en
g
i
n
es
u
s
e
w
eb
cr
awle
r
s
to
in
d
ex
d
o
c
u
m
en
ts
b
ased
on
a
lis
t
of
k
ey
wo
r
d
s
th
e
y
co
n
tain
.
T
h
en
,
em
p
lo
y
in
g
r
an
k
in
g
alg
o
r
ith
m
s
,
th
e
in
d
ex
ed
d
o
c
u
m
en
ts
ar
e
s
o
r
ted
in
d
ec
r
ea
s
in
g
o
r
d
er
.
T
h
is
r
an
k
in
g
f
ea
tu
r
e
d
i
s
tin
g
u
is
h
es
IR
f
r
o
m
o
th
er
NL
P
ap
p
licatio
n
s
.
Ov
er
th
e
y
ea
r
s
,
v
ar
io
u
s
r
an
k
i
n
g
m
o
d
els
h
av
e
b
ee
n
d
e
v
elo
p
e
d
,
r
an
g
in
g
f
r
o
m
v
ec
to
r
s
p
ac
e
m
o
d
els
in
tr
o
d
u
ce
d
in
th
e
70
th
to
later
lear
n
in
g
-
to
-
r
an
k
m
o
d
els
an
d
n
eu
r
al
r
an
k
i
n
g
m
o
d
els
[
3
2
]
.
Desp
ite
th
e
s
u
cc
ess
f
u
l
ap
p
licatio
n
of
th
ese
m
eth
o
d
s
in
m
a
n
y
d
if
f
er
en
t
IR
ap
p
licatio
n
s
,
s
ea
r
ch
en
g
in
es
s
till
h
av
e
lim
itatio
n
s
,
as
th
ey
p
r
o
v
id
e
o
n
ly
a
lis
t
of
d
o
cu
m
e
n
ts
an
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
Th
e
a
d
va
n
c
es in
n
a
tu
r
a
l la
n
g
u
a
g
e
p
r
o
ce
s
s
in
g
tech
n
o
lo
g
y
a
n
d
its
imp
a
ct
o
n
… (
N
a
d
ezh
d
a
B
o
r
is
o
va
)
2329
no
clea
r
an
s
wer
to
u
s
er
q
u
er
i
es.
E
f
f
o
r
ts
ar
e
th
u
s
f
o
c
u
s
ed
on
in
teg
r
atin
g
in
f
o
r
m
atio
n
ex
tr
ac
tio
n
tech
n
o
lo
g
ies
with
in
f
o
r
m
atio
n
r
etr
iev
al
to
o
v
er
co
m
e
th
ese
lim
itatio
n
s
[
3
3
]
.
C
o
n
v
er
s
atio
n
al
ag
en
ts
ar
e
d
ialo
g
u
e
s
y
s
tem
s
d
esig
n
ed
to
im
itate
h
u
m
an
in
ter
ac
tio
n
s
in
n
atu
r
al
co
n
v
er
s
atio
n
.
So
m
e
c
o
n
v
e
r
s
atio
n
al
ag
e
n
ts
u
s
e
o
n
l
y
tex
t
as
i
n
p
u
t
a
n
d
o
u
tp
u
t,
lik
e
a
ch
atb
o
t.
Oth
er
s
ar
e
more
co
m
p
lex
an
d
in
v
o
lv
e
s
p
ee
ch
in
p
u
t
an
d
o
u
t
p
u
t,
s
u
ch
as
th
e
v
ir
tu
al
ass
is
tan
ts
Alex
a
a
n
d
Siri.
W
ah
d
e
an
d
Vir
g
o
lin
[
3
4
]
also
m
en
tio
n
e
d
th
e
so
-
ca
lled
em
b
o
d
ied
c
o
n
v
er
s
atio
n
al
a
g
en
ts
,
wh
ich
ty
p
ically
h
a
v
e
an
an
im
ated
v
is
u
al
r
ep
r
esen
tatio
n
lik
e
an
an
im
ated
f
ac
e
or
b
o
d
y
o
n
s
cr
ee
n
.
Ho
wev
e
r
,
th
e
m
ajo
r
task
of
all
th
ese
ty
p
es
of
c
o
n
v
e
r
s
atio
n
al
ag
en
t
s
is
how
th
ey
p
er
f
o
r
m
an
d
m
an
ag
e
th
e
d
ialo
g
u
e
ca
p
a
b
ilit
ies.
T
h
e
f
ir
s
t
p
atter
n
-
b
ased
ch
atb
o
ts
m
atch
th
e
u
s
er
'
s
in
p
u
t
to
a
s
et
of
p
atter
n
s
an
d
th
en
ap
p
ly
r
u
les
to
f
o
r
m
u
lat
e
a
r
esp
o
n
s
e.
Mo
r
e
m
o
d
er
n
p
atter
n
-
b
ased
ch
atb
o
ts
u
s
e
th
e
ar
tific
ial
in
tellig
en
ce
m
ar
k
u
p
la
n
g
u
a
g
e
,
s
p
ec
if
i
ca
lly
d
ev
elo
p
e
d
to
d
ef
in
e
tem
p
late
-
m
atch
in
g
r
u
les.
I
n
f
o
r
m
atio
n
-
r
etr
iev
al
c
h
atb
o
ts
p
r
o
d
u
ce
r
esp
o
n
s
es
by
s
elec
tin
g
an
ap
p
r
o
p
r
iate
s
en
ten
ce
f
r
o
m
a
d
ialo
g
u
e
c
o
r
p
u
s
,
wh
ile
g
e
n
er
ativ
e
ch
at
b
o
t
s
p
r
o
d
u
ce
th
eir
r
esp
o
n
s
es
u
s
in
g
s
tatis
tical
m
o
d
els.
Per
s
o
n
al
v
ir
tu
al
ass
is
tan
ts
g
en
er
ally
p
r
o
v
id
e
s
p
ee
ch
r
ec
o
g
n
it
io
n
as
an
ad
v
an
ce
d
f
ea
tu
r
e
a
n
d
co
m
b
in
e
c
h
atb
o
t
f
u
n
ctio
n
ality
with
task
-
o
r
ien
te
d
ca
p
ab
ilit
ies.
T
ex
t
s
u
m
m
ar
izatio
n
in
v
o
l
v
es
id
en
tify
in
g
th
e
k
ey
in
f
o
r
m
atio
n
in
a
d
o
cu
m
en
t
an
d
p
r
o
d
u
cin
g
a
s
u
m
m
ar
y
th
at
p
r
esen
ts
th
e
m
o
s
t
r
elev
an
t
in
f
o
r
m
atio
n
f
r
o
m
th
e
o
r
ig
in
al
c
o
n
ten
t.
T
h
er
e
a
r
e
v
a
r
io
u
s
class
if
icatio
n
s
f
o
r
au
to
m
atic
t
ex
t
s
u
m
m
ar
izatio
n
b
ased
on
d
if
f
er
en
t
f
ac
to
r
s
[
3
5
]
,
[
3
6
]
.
T
h
ese
in
clu
d
e
s
in
g
le
-
d
o
cu
m
e
n
t
an
d
m
u
lti
-
d
o
c
u
m
en
t
s
u
m
m
ar
izatio
n
b
ased
on
th
e
n
u
m
b
er
of
d
o
cu
m
en
ts
an
d
e
x
tr
ac
tiv
e,
ab
s
tr
ac
tiv
e,
an
d
h
y
b
r
id
ap
p
r
o
ac
h
es
b
ased
on
s
u
m
m
ar
izatio
n
m
eth
o
d
s
.
S
u
m
m
ar
ies
can
also
be
ca
teg
o
r
ized
as
g
en
er
ic
or
q
u
er
y
-
f
o
cu
s
ed
s
u
m
m
ar
ies,
d
e
p
en
d
in
g
on
th
e
ch
ar
ac
ter
is
tics
of
t
h
e
o
u
tp
u
t
s
u
m
m
ar
y
.
T
h
e
s
u
m
m
ar
izin
g
s
y
s
tem
can
be
ca
teg
o
r
ized
as
m
o
n
o
li
n
g
u
al,
m
u
ltil
in
g
u
al,
or
cr
o
s
s
-
lin
g
u
al
b
ased
on
th
e
in
p
u
t
an
d
o
u
tp
u
t
lan
g
u
a
g
es.
Ad
d
itio
n
ally
,
th
e
alg
o
r
ith
m
u
s
ed
f
o
r
g
en
e
r
atin
g
th
e
s
u
m
m
ar
i
es
can
be
s
u
p
er
v
is
ed
or
u
n
s
u
p
er
v
is
ed
.
T
h
er
e
ar
e
o
th
er
class
if
icatio
n
s
b
ased
on
t
h
e
len
g
th
,
ty
p
e
of
o
u
tp
u
t
co
n
te
n
t,
an
d
d
o
m
ain
of
a
u
to
m
atic
te
x
t
s
u
m
m
ar
izatio
n
.
Qu
esti
o
n
an
s
wer
in
g
is
a
cr
iti
ca
l
task
in
NL
P
th
at
en
ab
les
m
ac
h
in
es
to
u
n
d
er
s
tan
d
an
d
r
esp
o
n
d
to
q
u
esti
o
n
s
ask
ed
by
h
u
m
a
n
s
in
n
atu
r
al
lan
g
u
ag
e.
Similar
to
o
th
er
ar
ea
s
in
NL
P,
q
u
esti
o
n
-
an
s
wer
in
g
h
as
also
b
ee
n
im
p
ac
ted
by
th
e
p
r
o
g
r
e
s
s
of
d
ee
p
lear
n
in
g
m
eth
o
d
s
.
Deep
lear
n
in
g
m
eth
o
d
s
s
u
p
p
o
r
t
an
d
im
p
r
o
v
e
p
er
f
o
r
m
an
ce
f
o
r
d
if
f
e
r
en
t
task
s
in
th
e
q
u
esti
o
n
-
an
s
wer
in
g
p
r
o
ce
s
s
.
An
o
v
er
v
iew
of
q
u
esti
o
n
an
s
wer
in
g
in
d
ee
p
lear
n
in
g
c
o
n
tex
ts
[
3
7
]
ca
teg
o
r
ized
it
in
to
th
e
f
o
llo
win
g
th
r
ee
ty
p
es:
r
u
le
-
b
ased
q
u
esti
o
n
an
s
wer
in
g
,
q
u
esti
o
n
an
s
wer
in
g
in
t
h
e
er
a
of
d
ee
p
lea
r
n
in
g
,
an
d
v
is
u
al
q
u
esti
o
n
an
s
wer
in
g
.
An
e
x
am
p
l
e
of
th
e
ea
r
ly
wo
r
k
on
r
u
le
-
b
ased
q
u
esti
o
n
a
n
s
wer
in
g
is
m
en
tio
n
e
d
by
Gr
ee
n
et
al
.
[
3
8
]
,
wh
er
e
q
u
esti
o
n
s
r
elate
d
to
b
aseb
all
g
am
es
wer
e
an
s
wer
ed
u
s
in
g
a
g
am
e
d
atab
ase.
T
h
e
b
aseb
a
ll
s
y
s
tem
in
v
o
lv
es
s
y
n
tactic
an
aly
s
is
an
d
co
n
ten
t
an
aly
s
is
to
ex
tr
ac
t
th
e
wo
r
d
s
in
th
e
in
p
u
t
q
u
esti
o
n
.
Vir
tu
al
p
er
s
o
n
al
ass
i
s
tan
ts
ar
e
co
m
m
o
n
ex
am
p
les
of
q
u
esti
o
n
an
s
wer
in
g
in
th
e
e
r
a
of
d
ee
p
lear
n
in
g
.
B
lo
ck
,
a
b
ili
n
ea
r
f
u
s
io
n
m
o
d
el
b
ased
on
s
u
p
er
-
d
iag
o
n
al
ten
s
o
r
d
ec
o
m
p
o
s
itio
n
[
3
9
]
,
is
cited
as
an
ex
am
p
le
f
o
r
th
e
v
is
u
al
q
u
esti
o
n
-
an
s
wer
task
.
Hao
et
al
.
[
4
0
]
also
p
r
esen
ted
a
s
y
s
tem
atic
r
ev
iew
of
th
e
r
ec
en
t
d
ev
elo
p
m
en
t
of
d
ee
p
lear
n
in
g
m
eth
o
d
s
f
o
r
q
u
esti
o
n
a
n
s
wer
in
g
.
T
h
is
s
u
r
v
e
y
co
v
er
s
th
e
s
co
p
e,
n
etwo
r
k
s
tr
u
ctu
r
e
ch
ar
ac
ter
is
tics
an
d
t
h
eir
ef
f
ec
tiv
en
ess
,
m
eth
o
d
o
l
o
g
ical
in
n
o
v
atio
n
s
,
d
atasets
,
an
d
ap
p
licatio
n
s
.
T
h
e
f
ield
of
m
ac
h
in
e
tr
an
s
latio
n
in
v
o
lv
es
tr
an
s
latin
g
p
h
r
ase
s
f
r
o
m
one
n
atu
r
al
lan
g
u
ag
e
to
an
o
th
er
.
Go
o
g
le
T
r
an
s
late
is
one
of
th
e
m
o
s
t
p
o
p
u
lar
a
p
p
licatio
n
s
in
th
is
f
ield
.
C
h
allen
g
es
in
m
ac
h
in
e
tr
an
s
latio
n
in
clu
d
e
m
ain
tain
i
n
g
t
h
e
m
ea
n
in
g
of
s
en
ten
ce
s
tr
u
ctu
r
es
a
n
d
g
r
am
m
ar
ac
r
o
s
s
d
if
f
e
r
en
t
lan
g
u
ag
es.
Var
i
o
u
s
ap
p
r
o
ac
h
es,
each
with
its
o
wn
ad
v
a
n
tag
es
an
d
d
is
ad
v
an
tag
e
s
,
h
av
e
b
ee
n
u
s
ed
in
m
ac
h
in
e
tr
an
s
latio
n
s
y
s
tem
s
.
T
h
ese
ap
p
r
o
ac
h
es
in
clu
d
e
r
u
le
-
b
ased
,
co
r
p
u
s
-
b
ased
,
an
d
tr
an
s
f
er
-
b
ased
ap
p
r
o
ac
h
es
[
4
1
]
.
T
h
e
r
u
le
-
b
ased
ap
p
r
o
ac
h
,
also
k
n
o
w
n
as
th
e
c
lass
ical
ap
p
r
o
ac
h
to
m
ac
h
i
n
e
t
r
an
s
latio
n
,
r
elies
on
lin
g
u
is
tic
in
f
o
r
m
atio
n
ab
o
u
t
lan
g
u
ag
es
o
b
tain
ed
f
r
o
m
b
ilin
g
u
al
d
ictio
n
a
r
ies
an
d
g
r
am
m
a
r
.
E
ac
h
in
p
u
t
s
en
ten
ce
is
co
n
v
er
ted
to
an
o
u
tp
u
t
s
en
ten
ce
b
ased
on
th
e
m
o
r
p
h
o
lo
g
ical,
s
y
n
tactic,
a
n
d
s
em
an
tic
an
aly
s
is
of
b
o
th
lan
g
u
a
g
es
in
v
o
lv
ed
.
As
a
r
esu
lt,
a
r
u
le
-
b
ased
s
y
s
tem
r
eq
u
ir
es
s
y
n
tax
an
d
s
em
an
tic
an
aly
s
is
as
well
as
g
en
er
ati
o
n
.
T
r
an
s
f
er
-
b
ased
m
ac
h
in
e
tr
an
s
latio
n
ac
h
ie
v
es
h
ig
h
ac
cu
r
ac
y
an
d
p
r
o
d
u
c
es
ad
eq
u
ate
tr
an
s
latio
n
s
f
r
o
m
an
in
ter
m
ed
iat
e
r
ep
r
esen
tatio
n
th
at
m
im
ics
th
e
m
ea
n
in
g
of
th
e
o
r
ig
i
n
al
s
en
ten
ce
.
Ho
wev
er
,
a
d
r
awb
ac
k
of
th
is
ap
p
r
o
ac
h
is
th
at
r
u
les
m
u
s
t
be
ap
p
lied
at
ev
er
y
tr
an
s
latio
n
s
tep
.
C
o
r
p
u
s
-
b
ased
m
ac
h
i
n
e
tr
an
s
latio
n
is
an
alter
n
ativ
e
ap
p
r
o
ac
h
th
at
u
s
es
a
b
ilin
g
u
al
p
ar
allel
co
r
p
u
s
to
ac
q
u
ir
e
tr
a
n
s
latio
n
k
n
o
wled
g
e
.
Statis
tica
l
m
ac
h
in
e
lear
n
in
g
aim
s
to
f
in
d
co
r
r
esp
o
n
d
in
g
elem
en
ts
in
one
lan
g
u
ag
e
to
elem
en
ts
in
an
o
th
er
lan
g
u
ag
e,
an
d
th
e
m
o
s
t
ap
p
r
o
p
r
iate
tr
an
s
latio
n
is
th
e
one
to
wh
ich
th
e
s
y
s
tem
ass
ig
n
s
th
e
h
ig
h
est
p
r
o
b
ab
ilit
y
.
E
ac
h
tr
an
s
latio
n
ap
p
r
o
ac
h
h
as
its
o
wn
ad
v
an
tag
es
an
d
d
is
ad
v
an
tag
es,
with
th
e
p
r
o
s
of
one
ap
p
r
o
ac
h
o
f
te
n
co
m
p
en
s
atin
g
f
o
r
th
e
co
n
s
of
an
o
th
er
.
T
h
is
h
as
m
o
tiv
ated
th
e
d
ev
elo
p
m
en
t
of
h
y
b
r
id
m
ac
h
i
n
e
tr
an
s
latio
n
,
wh
ich
in
v
o
lv
es
u
s
in
g
m
u
ltip
le
m
ac
h
in
e
tr
a
n
s
latio
n
a
p
p
r
o
ac
h
es
with
in
a
s
in
g
le
s
y
s
tem
.
T
o
p
ic
m
o
d
elin
g
is
u
s
ed
in
v
a
r
io
u
s
d
o
m
ain
s
to
a
n
aly
ze
th
e
to
p
ical
s
tr
u
ctu
r
e
of
a
lar
g
e
co
llectio
n
of
d
o
cu
m
e
n
ts
.
It
is
one
of
th
e
m
o
s
t
p
o
p
u
lar
s
tatis
tical
t
o
o
ls
f
o
r
an
aly
zin
g
b
io
in
f
o
r
m
atic,
s
o
cial,
an
d
en
v
ir
o
n
m
en
tal
d
ata,
as
well
as
s
im
ilar
d
ata
f
r
o
m
lar
g
e
d
at
asets
.
Selectin
g
th
e
r
ig
h
t
m
et
h
o
d
s
f
o
r
e
x
tr
ac
tin
g
m
ea
n
in
g
f
u
l
d
ata
f
r
o
m
d
atasets
is
cr
u
cial
to
ev
er
y
d
ata
an
aly
s
is
p
r
o
ject.
Stan
d
ar
d
to
p
ic
m
o
d
el
ap
p
r
o
ac
h
es
ar
e
d
iv
id
ed
in
to
f
i
v
e
g
r
o
u
p
s
[
4
2
]
:
b
asic
m
o
d
els,
m
o
d
els
with
ad
v
an
ce
d
to
p
ic
r
elatio
n
s
h
ip
s
,
tim
e
-
b
ased
to
p
ic
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
2
,
Ap
r
il
20
25
:
2
3
2
5
-
2
3
3
3
2330
m
o
d
els,
s
h
o
r
t
-
tex
t
o
p
tim
ized
to
p
ic
m
o
d
els,
an
d
o
t
h
er
s
ig
n
i
f
ican
t
to
p
ic
m
o
d
el
d
esig
n
s
.
T
h
e
class
if
icatio
n
of
m
eth
o
d
s
is
f
lex
ib
le
an
d
allo
ws
f
o
r
th
e
u
p
d
atin
g
of
n
ew
m
e
th
o
d
s
.
Fu
et
al
.
[
4
3
]
also
d
is
cu
s
s
es
m
eth
o
d
s
of
to
p
ic
m
o
d
elin
g
,
alo
n
g
with
th
e
ir
f
ea
tu
r
es
an
d
lim
itatio
n
s
an
d
th
e
av
ailab
le
to
o
ls
.
Alth
o
u
g
h
th
e
task
s
m
ay
v
ar
y
,
it
is
im
p
o
r
tan
t
to
estab
lis
h
a
s
t
r
o
n
g
f
o
u
n
d
atio
n
to
d
ev
elo
p
a
v
ar
iety
of
NL
P
ap
p
licatio
n
s
,
wh
ich
co
v
er
s
lear
n
in
g
th
e
tech
n
o
l
o
g
ies
th
at
can
be
u
s
ed
.
NL
P
is
a
cr
itical
co
m
p
o
n
en
t
of
ar
tific
ial
in
tellig
en
ce
an
d
p
r
o
v
id
es
s
o
lu
tio
n
s
to
r
ea
l
-
wo
r
ld
p
r
o
b
lem
s
th
at
h
av
e
th
e
p
o
ten
tial
to
ch
an
g
e
th
e
way
co
m
p
an
ies
o
p
e
r
ate.
T
h
e
f
o
llo
win
g
s
ec
tio
n
p
r
esen
ts
th
e
u
s
e
of
NL
P
ap
p
licatio
n
s
f
o
r
v
a
r
io
u
s
s
ec
to
r
s
.
4.
DIS
CU
SS
I
O
N
C
h
atGPT
r
ep
r
esen
ts
a
s
ig
n
if
ican
t
ad
v
an
ce
m
e
n
t
in
NL
P
tech
n
o
lo
g
y
f
o
r
b
o
th
o
r
g
a
n
izatio
n
s
an
d
in
d
iv
id
u
als.
It
can
be
u
s
ed
f
o
r
cu
s
to
m
er
s
er
v
ice,
tr
a
n
s
latio
n
,
s
u
m
m
ar
izatio
n
,
an
d
c
o
n
te
n
t
p
r
o
d
u
ctio
n
.
T
h
is
tech
n
o
lo
g
y
h
as
en
o
r
m
o
u
s
p
o
t
en
tial
an
d
ch
allen
g
es
f
o
r
b
u
s
in
ess
es,
s
o
cieties,
an
d
p
eo
p
le
in
g
en
er
al.
In
th
e
ed
u
ca
tio
n
s
ec
to
r
,
h
o
wev
e
r
,
th
er
e
ar
e
well
-
k
n
o
wn
d
if
f
icu
ltie
s
with
u
s
in
g
g
en
er
ativ
e
ar
tific
ial
in
tellig
en
ce
due
to
th
e
lack
of
estab
lis
h
ed
r
u
le
s
an
d
m
o
r
al
s
tan
d
ar
d
s
[
4
4
]
–
[
4
6
]
.
Sp
e
n
n
em
a
n
n
[
4
7
]
p
r
esen
te
d
th
e
o
p
p
o
r
t
u
n
ities
an
d
p
r
o
b
lem
s
with
NL
P
m
o
d
els
lik
e
C
h
atGPT
an
d
Go
o
g
l
e
B
ar
d
an
d
how
th
e
y
co
u
ld
r
ev
o
lu
tio
n
ize
h
ig
h
e
r
ed
u
ca
tio
n
teac
h
in
g
an
d
lear
n
i
n
g
.
T
h
e
ar
ticle
d
is
cu
s
s
es
th
e
a
d
v
an
tag
es
of
u
s
in
g
NL
P
m
o
d
els
f
o
r
p
er
s
o
n
alize
d
lear
n
in
g
a
n
d
on
-
d
em
an
d
s
u
p
p
o
r
t,
s
u
c
h
as
p
r
o
v
id
in
g
f
ee
d
b
ac
k
an
d
ass
is
tan
ce
an
d
o
f
f
er
in
g
r
eso
u
r
ce
s
to
s
tu
d
en
ts
.
T
h
e
a
u
th
o
r
also
h
i
g
h
lig
h
ts
th
e
p
o
ten
tial
d
r
awb
ac
k
s
of
NL
P
m
o
d
els,
s
u
ch
as
t
h
e
p
o
te
n
tial
lo
s
s
of
h
u
m
an
in
ter
ac
tio
n
an
d
eth
ical
is
s
u
es.
To
tack
le
th
ese
is
s
u
es
,
u
n
iv
er
s
ities
s
h
o
u
ld
g
u
ar
an
te
e
th
at
NL
P
m
o
d
els
ar
e
u
s
ed
to
en
h
an
ce
,
r
ath
e
r
th
an
r
ep
lace
,
h
u
m
a
n
in
ter
ac
tio
n
.
Fo
r
in
s
tan
ce
,
in
v
o
lv
in
g
s
tu
d
en
ts
in
th
e
cr
ea
tio
n
an
d
im
p
lem
en
tatio
n
of
NL
P
m
o
d
els
to
m
ee
t
th
eir
s
p
ec
if
ic
n
ee
d
s
an
d
p
r
ef
e
r
en
ce
s
.
Mo
r
eo
v
er
,
u
n
iv
e
r
s
ity
s
taf
f
m
u
s
t
be
tr
ain
ed
to
u
s
e
an
d
ad
ap
t
n
ew
tech
n
o
lo
g
y
an
d
p
r
o
v
id
e
s
tu
d
en
ts
with
r
eso
u
r
c
es
an
d
s
u
p
p
o
r
t
to
ef
f
ec
tiv
ely
u
s
e
th
e
m
o
d
els.
I
n
s
titu
tio
n
s
s
h
o
u
ld
also
estab
lis
h
eth
ical
s
tan
d
ar
d
s
f
o
r
th
e
u
s
e
of
NL
P
m
o
d
els
to
s
af
eg
u
ar
d
s
tu
d
en
t
p
r
iv
ac
y
.
A
d
d
itio
n
ally
,
in
s
titu
tio
n
s
s
h
o
u
ld
ev
alu
ate
th
e
p
o
s
itiv
e
an
d
n
eg
ativ
e
asp
ec
ts
of
u
s
in
g
NL
P
m
o
d
els
in
h
i
g
h
er
e
d
u
ca
tio
n
wh
ile
e
n
s
u
r
in
g
r
esp
o
n
s
ib
le
u
s
e
to
en
h
a
n
ce
s
tu
d
en
t
l
ea
r
n
in
g
.
No
wad
ay
s
,
C
h
atGPT
is
wid
e
ly
u
s
ed
f
o
r
task
s
s
u
ch
as
lan
g
u
ag
e
tr
an
s
latio
n
,
q
u
esti
o
n
an
s
wer
in
g
,
s
u
m
m
ar
izatio
n
,
a
n
d
c
o
n
ten
t
c
r
ea
tio
n
.
I
ts
ab
ilit
y
to
g
e
n
er
ate
co
n
tex
tu
ally
s
u
itab
le
r
esp
o
n
s
e
s
m
ak
es
it
p
o
s
s
ib
le
to
ass
i
s
t
wi
th
ac
tiv
ities
lik
e
m
ain
tain
in
g
s
er
v
ice
q
u
ality
by
an
s
wer
in
g
co
m
m
o
n
in
q
u
ir
ies.
C
h
atGPT
can
a
ls
o
aid
in
d
ata
an
aly
s
is
an
d
r
ese
ar
ch
by
g
en
e
r
atin
g
r
ep
o
r
ts
an
d
s
u
m
m
ar
ies
f
r
o
m
h
u
g
e
d
ata
b
ases
.
B
i
s
was
[
4
8
]
ex
p
lo
r
ed
th
e
p
o
ten
tial
u
s
e
of
C
h
atGPT
f
o
r
g
lo
b
al
war
m
i
n
g
.
T
h
e
a
u
th
o
r
ac
k
n
o
wled
g
e
s
ask
in
g
C
h
atGPT
q
u
esti
o
n
s
ab
o
u
t
its
u
s
e
in
clim
ate
ch
an
g
e
r
esear
ch
.
W
h
ile
s
o
m
e
of
th
e
m
en
tio
n
ed
u
s
es
ar
e
cu
r
r
en
tly
p
o
s
s
ib
le,
o
th
er
s
h
o
ld
p
o
te
n
tial
f
o
r
th
e
f
u
tu
r
e
.
Var
io
u
s
clim
ate
s
ce
n
ar
io
s
b
ased
on
a
d
iv
er
s
e
s
e
t
of
d
ata
in
p
u
ts
ar
e
u
tili
ze
d
to
im
p
r
o
v
e
th
e
ac
c
u
r
ac
y
of
th
e
clim
ate
p
r
o
jectio
n
s
.
T
h
e
d
e
v
elo
p
m
e
n
t
of
C
h
atGPT
is
a
wate
r
s
h
ed
m
o
m
en
t
in
th
e
h
is
to
r
y
of
NL
P
an
d
ar
tific
ial
in
tellig
en
ce
.
I
ts
o
n
g
o
in
g
d
ev
el
o
p
m
en
t
is
an
ticip
ated
to
h
av
e
a
s
ig
n
if
ican
t
im
p
ac
t
on
co
n
v
er
s
atio
n
al
ar
tific
ial
in
tellig
en
ce
,
with
ef
f
ec
ts
th
at
will
u
n
d
o
u
b
t
ed
ly
ex
ten
d
ac
r
o
s
s
all
s
ec
to
r
s
of
s
o
ciety
.
An
aly
zin
g
cu
s
to
m
er
r
e
v
iews
u
s
in
g
NL
P
can
p
r
o
v
id
e
v
alu
a
b
le
in
s
ig
h
t
in
to
cu
s
to
m
er
'
s
o
p
in
io
n
s
ab
o
u
t
a
b
r
a
n
d
a
n
d
its
p
r
o
d
u
cts
or
s
e
r
v
ices.
By
em
p
lo
y
in
g
s
en
tim
e
n
t
an
aly
s
is
an
d
tex
t
class
if
icat
io
n
,
c
o
m
p
an
ies
ca
n
g
ain
an
u
n
d
e
r
s
tan
d
in
g
of
o
v
er
all
p
o
s
itiv
e
or
n
eg
ativ
e
s
en
tim
en
t
ab
o
u
t
th
e
b
r
an
d
an
d
i
d
en
ti
f
y
wh
at
cu
s
to
m
e
r
s
ap
p
r
ec
iate
or
d
is
lik
e
ab
o
u
t
th
eir
o
f
f
er
in
g
s
.
Fu
r
th
er
m
o
r
e,
m
ac
h
in
e
lear
n
in
g
a
n
d
NL
P
tech
n
iq
u
es
ar
e
u
s
ed
to
d
etec
t
an
d
r
ec
o
g
n
ize
em
o
tio
n
s
in
cu
s
to
m
er
r
ev
iews.
E
v
alu
atio
n
of
p
r
o
d
u
ct
r
ev
iews
can
be
u
tili
ze
d
to
p
r
ed
ict
p
r
icin
g
tr
e
n
d
s
,
ass
ess
ca
m
p
aig
n
s
,
an
d
s
h
ap
e
an
e
-
c
o
m
m
er
ce
m
ar
k
etin
g
s
tr
ateg
y
[
4
9
]
–
[
5
1
]
.
NL
P
is
also
u
s
ed
to
p
r
o
ce
s
s
an
d
an
aly
ze
cu
s
to
m
er
s
er
v
ice
s
u
r
v
ey
s
to
b
etter
u
n
d
e
r
s
tan
d
th
e
ch
allen
g
es
th
at
co
n
s
u
m
er
s
ar
e
ex
p
er
ien
cin
g
.
T
h
is
in
f
o
r
m
ati
o
n
is
cr
itical
f
o
r
en
h
an
cin
g
c
u
s
to
m
er
s
atis
f
ac
tio
n
,
wh
ich
will
in
cr
ea
s
e
cu
s
to
m
e
r
r
eten
tio
n
an
d
p
r
o
m
o
te
p
o
s
itiv
e
wo
r
d
-
of
-
m
o
u
th
.
A
s
y
s
tem
atic
r
ev
iew
h
as
o
u
tlin
ed
th
e
ap
p
licatio
n
of
ar
t
if
icial
in
tellig
en
ce
an
d
NL
P
in
cu
s
to
m
er
s
er
v
ice
f
r
o
m
2
0
1
5
to
2
0
2
2
[
5
2
]
.
T
h
e
r
ev
iew
u
tili
ze
d
f
iv
e
s
cien
tific
d
atab
ases
to
an
aly
ze
an
d
r
ef
i
n
e
all
r
elev
an
t
p
ap
er
s
.
Acc
o
r
d
in
g
to
th
e
au
th
o
r
s
,
ch
atb
o
ts
an
d
q
u
esti
o
n
-
an
s
wer
i
n
g
s
y
s
tem
s
ar
e
u
s
ed
in
ten
m
a
in
ar
ea
s
,
m
o
s
tly
in
g
e
n
er
al,
s
o
cial
m
ed
ia,
an
d
e
-
co
m
m
er
ce
.
Oth
e
r
m
ain
a
r
ea
s
ar
e
h
ea
lth
ca
r
e,
telec
o
m
m
u
n
icatio
n
s
,
b
o
o
k
in
g
,
th
e
co
n
s
tr
u
ctio
n
in
d
u
s
tr
y
,
b
an
k
in
g
,
en
er
g
y
u
tili
ties
,
an
d
m
ar
k
etin
g
.
In
all
of
th
ese
ar
e
as,
cu
s
to
m
er
s
ex
p
ec
t
an
im
m
e
d
iate
r
esp
o
n
s
e
to
a
q
u
esti
o
n
,
lead
i
n
g
to
th
e
p
o
p
u
l
ar
ity
of
NL
P
-
p
o
wer
e
d
c
h
atb
o
t
s
an
d
d
ig
ital
ass
is
tan
ts
s
u
ch
as
Go
o
g
le
Ass
is
tan
t,
Alex
a,
an
d
Siri.
T
h
ese
ap
p
s
en
ab
le
cu
s
to
m
er
s
to
co
n
d
u
ct
o
n
lin
e
s
ea
r
ch
es
or
m
ak
e
p
h
o
n
e
ca
lls
u
s
in
g
th
eir
v
o
ices
an
d
r
ec
eiv
e
th
e
r
elev
an
t
in
f
o
r
m
atio
n
[
5
3
]
.
Acc
o
r
d
in
g
to
m
ar
k
et
d
ata
p
r
esen
t
ed
on
th
e
Yag
u
ar
a
p
latf
o
r
m
[
5
4
]
,
o
v
er
th
e
p
ast
f
ew
y
ea
r
s
,
m
o
r
e
th
a
n
50%
of
th
e
wo
r
ld
wid
e
p
o
p
u
latio
n
u
s
e
s
th
e
v
o
ice
s
ea
r
ch
d
aily
an
d
Go
o
g
le
Ass
is
tan
t
is
th
e
m
o
s
t
p
o
p
u
lar
d
ig
ital
ass
is
tan
t,
u
s
ed
by
36%
of
u
s
er
s
.
Data
f
o
r
v
o
ice
s
ea
r
ch
es
s
h
o
w
th
at
Go
o
g
le
Ass
is
tan
t
ac
cu
r
ately
u
n
d
er
s
tan
d
s
q
u
er
ies
100%
of
th
e
tim
e
an
d
p
r
o
v
id
es
co
r
r
ec
t
an
s
wer
s
alm
o
s
t
93%
of
th
e
tim
e.
On
th
e
o
th
er
h
an
d
,
Siri
an
d
Alex
a
u
n
d
e
r
s
tan
d
th
e
q
u
e
r
y
ac
cu
r
ately
9
9
.
8
%
of
th
e
tim
e
but
s
u
cc
ess
f
u
lly
d
eliv
er
an
ac
cu
r
ate
an
s
wer
ar
o
u
n
d
80%
of
th
e
tim
e
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
Th
e
a
d
va
n
c
es in
n
a
tu
r
a
l la
n
g
u
a
g
e
p
r
o
ce
s
s
in
g
tech
n
o
lo
g
y
a
n
d
its
imp
a
ct
o
n
… (
N
a
d
ezh
d
a
B
o
r
is
o
va
)
2331
In
r
ec
r
u
itm
en
t
,
jo
b
s
ee
k
er
s
can
ap
p
ly
by
s
im
p
ly
u
p
lo
a
d
in
g
or
s
en
d
in
g
th
eir
r
esu
m
es.
Ho
wev
er
,
it
is
ch
allen
g
in
g
to
f
in
d
r
eliab
le
c
an
d
id
ates
due
to
th
e
d
i
f
f
er
en
t
s
ty
les
u
s
ed
in
th
e
r
esu
m
es.
A
p
r
o
p
o
s
ed
s
m
ar
t
r
ec
r
u
itm
en
t
s
y
s
tem
in
clu
d
es
r
esu
m
e
class
if
icatio
n
to
ad
d
r
es
s
th
is
is
s
u
e
[
5
5
]
.
T
h
is
s
y
s
tem
m
ea
s
u
r
es
ap
p
lican
ts
'
tech
n
ical
p
r
o
f
icie
n
cy
th
r
o
u
g
h
au
to
m
atic
q
u
esti
o
n
ar
ies
an
d
a
n
aly
ze
s
th
eir
r
esu
m
es
f
o
r
s
y
n
t
ac
tical
an
d
s
em
an
tic
s
im
ilar
ities
b
ased
on
th
e
co
m
p
an
y
’
s
r
eq
u
ir
em
en
ts
.
NL
P
an
d
n
am
ed
en
tity
r
ec
o
g
n
itio
n
ar
e
u
s
ed
to
ex
tr
ac
t
m
ea
n
in
g
f
u
l
d
ata
r
elate
d
to
a
c
an
d
id
ate'
s
wo
r
k
e
x
p
er
ien
ce
an
d
ed
u
ca
tio
n
.
Deep
lear
n
in
g
is
u
s
ed
f
o
r
th
e
NL
P
tech
n
iq
u
e
to
s
o
r
t
an
d
r
a
n
k
th
e
r
esu
m
es.
Ad
d
itio
n
ally
,
a
p
r
ev
io
u
s
ar
ticle
d
is
cu
s
s
ed
how
r
ec
r
u
itm
en
t
tech
n
iq
u
e
r
esear
ch
an
d
co
m
p
u
ter
s
cien
c
e
in
ter
s
ec
t,
p
r
esen
tin
g
an
a
p
p
r
o
ac
h
th
at
also
u
tili
ze
d
d
ee
p
le
ar
n
in
g
a
n
d
NL
P
f
o
r
s
o
r
tin
g
r
esu
m
es
[
5
6
]
.
T
h
is
ap
p
r
o
ac
h
d
if
f
er
s
f
r
o
m
th
e
f
o
r
m
e
r
by
in
c
o
r
p
o
r
atin
g
a
k
n
o
wled
g
e
b
ase
of
m
e
r
g
e
d
o
n
to
lo
g
y
an
d
em
p
lo
y
in
g
s
y
n
tactica
l
an
d
s
em
an
tic
s
im
ilar
ity
m
ea
s
u
r
em
en
ts
f
o
r
m
atch
i
n
g
th
e
s
u
b
m
itted
an
s
wer
s
of
th
e
a
p
p
lican
ts
r
e
g
ar
d
in
g
t
h
eir
s
k
ill
s
et.
NL
P
is
ev
o
lv
in
g
,
a
n
d
s
o
m
e
of
th
e
t
en
d
en
cies
in
NL
P
ap
p
licatio
n
s
in
clu
d
e
f
u
r
th
er
d
ev
elo
p
m
e
n
t
of
lar
g
e
lan
g
u
ag
e
m
o
d
els,
m
o
d
els
th
at
ca
n
h
an
d
le
m
u
ltip
le
lan
g
u
ag
es,
a
n
d
r
ea
l
-
tim
e
NL
P
ap
p
licatio
n
s
;
th
e
e
x
p
an
d
ed
u
s
e
of
s
en
tim
en
t
a
n
aly
s
is
an
d
o
p
in
io
n
m
in
in
g
;
en
h
an
cin
g
ac
ce
s
s
ib
ilit
y
f
o
r
in
d
iv
id
u
als
with
d
is
ab
ilit
ies,
s
u
ch
as
th
r
o
u
g
h
tex
t
-
to
-
s
p
ee
c
h
an
d
s
p
ee
ch
-
to
-
tex
t;
co
m
b
in
in
g
NL
P
with
r
o
b
o
tics
;
an
d
m
ac
h
in
e
lear
n
in
g
to
c
r
ea
te
more
co
m
p
r
e
h
en
s
iv
e
s
o
lu
tio
n
s
.
5.
CO
NCLU
SI
O
N
T
h
e
v
ast
am
o
u
n
t
of
n
at
u
r
al
la
n
g
u
ag
e
te
x
t
in
th
e
d
i
g
ital
wo
r
ld
co
n
tain
s
a
wea
lth
of
k
n
o
wle
d
g
e,
b
u
t
it
is
b
ec
o
m
in
g
in
cr
ea
s
in
g
ly
ch
al
len
g
in
g
f
o
r
a
h
u
m
an
to
id
en
ti
f
y
an
d
ex
tr
ac
t
th
is
k
n
o
wled
g
e
,
esp
ec
ially
with
in
tim
e
co
n
s
tr
ain
ts
.
T
h
r
o
u
g
h
NL
P
,
th
is
tex
tu
al
k
n
o
wled
g
e
is
tr
an
s
f
o
r
m
ed
in
to
alg
o
r
ith
m
s
to
s
o
lv
e
s
p
ec
if
ic
p
r
o
b
lem
s
.
T
h
e
id
ea
th
at
d
at
a
h
o
ld
s
v
al
u
ab
le
in
s
ig
h
ts
f
o
r
b
u
s
in
ess
es
h
as
g
ain
ed
wid
e
s
p
r
ea
d
ac
ce
p
tan
ce
,
lead
in
g
b
u
s
in
ess
es
to
in
v
est
in
v
ar
io
u
s
b
u
s
in
ess
in
tellig
en
ce
s
o
lu
tio
n
s
to
en
h
an
ce
th
eir
o
p
er
atio
n
s
.
R
esear
ch
d
em
o
n
s
tr
ates
th
at
th
e
o
b
jecti
v
e
of
au
to
m
ate
d
NL
P
is
to
h
an
d
le
th
is
task
as
ef
f
icien
tly
an
d
ac
cu
r
ately
as
a
h
u
m
an
.
As
NL
P
tech
n
o
lo
g
ies
co
n
tin
u
e
to
ev
o
lv
e,
th
e
y
h
o
ld
th
e
p
o
ten
tial
to
f
u
r
th
er
t
r
an
s
f
o
r
m
in
ter
ac
tio
n
s
b
etwe
en
h
u
m
a
n
s
an
d
c
o
m
p
u
te
r
s
.
ACK
NO
WL
E
DG
E
M
E
NT
S
T
h
is
wo
r
k
was
s
u
p
p
o
r
ted
by
P
r
o
ject
No
.
RP
-
A7
/2
2
–
“Da
ta
m
in
in
g
of
s
tu
d
e
n
ts
’
b
eh
av
io
r
in
d
is
tan
ce
e
-
lear
n
in
g
s
y
s
tem
s
:
a
ca
s
e
s
tu
d
y
LMS
B
lack
b
o
ar
d
,
”
So
u
th
-
W
est
Un
iv
er
s
ity
“N.
R
il
s
k
i,”
B
lag
o
ev
g
r
ad
,
B
u
lg
ar
ia
RE
F
E
R
E
NC
E
S
[
1
]
E.
D
.
Li
d
d
y
,
“
N
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
r
o
c
e
ssi
n
g
,
”
i
n
E
n
c
y
c
l
o
p
e
d
i
a
o
f
L
i
b
r
a
ry
a
n
d
I
n
f
o
rm
a
t
i
o
n
S
c
i
e
n
c
e
,
2
n
d
e
d
.
,
N
Y
:
M
a
r
c
e
l
D
e
c
k
e
r
,
I
n
c
.
,
2
0
0
1
.
[
2
]
M
.
K
.
N
y
e
i
n
a
n
d
K
.
M
.
S
o
e
,
“
S
o
u
r
c
e
si
d
e
p
r
e
-
o
r
d
e
r
i
n
g
u
s
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
f
o
r
E
n
g
l
i
sh
-
M
y
a
n
mar
mac
h
i
n
e
t
r
a
n
s
l
a
t
i
o
n
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
E
l
e
c
t
r
i
c
a
l
a
n
d
C
o
m
p
u
t
e
r
E
n
g
i
n
e
e
r
i
n
g
,
v
o
l
.
1
1
,
n
o
.
5
,
p
p
.
4
5
1
3
–
4
5
2
1
,
O
c
t
.
2
0
2
1
,
d
o
i
:
1
0
.
1
1
5
9
1
/
i
j
e
c
e
.
v
1
1
i
5
.
p
p
4
5
1
3
-
4
5
2
1
.
[
3
]
A
.
A
b
d
u
l
k
a
r
e
e
m,
T
.
E
.
S
o
m
e
f
u
n
,
O
.
K
.
C
h
i
n
e
d
u
m,
a
n
d
F
.
A
g
b
e
t
u
y
i
,
“
D
e
s
i
g
n
a
n
d
i
mp
l
e
m
e
n
t
a
t
i
o
n
o
f
s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
s
y
s
t
e
m
i
n
t
e
g
r
a
t
e
d
w
i
t
h
i
n
t
e
r
n
e
t
o
f
t
h
i
n
g
s,
”
I
n
t
e
rn
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
El
e
c
t
ri
c
a
l
a
n
d
C
o
m
p
u
t
e
r
E
n
g
i
n
e
e
ri
n
g
,
v
o
l
.
1
1
,
n
o
.
2
,
p
p
.
1
7
9
6
–
1
8
0
3
,
A
p
r
.
2
0
2
1
,
d
o
i
:
1
0
.
1
1
5
9
1
/
i
j
e
c
e
.
v
1
1
i
2
.
p
p
1
7
9
6
-
1
8
0
3
.
[
4
]
H
.
C
h
o
i
e
t
a
l
.
,
“
S
p
o
k
e
n
‐
t
o
‐
w
r
i
t
t
e
n
t
e
x
t
c
o
n
v
e
r
s
i
o
n
f
o
r
e
n
h
a
n
c
e
me
n
t
o
f
K
o
r
e
a
n
–
E
n
g
l
i
s
h
r
e
a
d
a
b
i
l
i
t
y
a
n
d
mac
h
i
n
e
t
r
a
n
s
l
a
t
i
o
n
,
”
ETRI
J
o
u
rn
a
l
,
v
o
l
.
4
6
,
n
o
.
1
,
p
p
.
1
2
7
–
1
3
6
,
F
e
b
.
2
0
2
4
,
d
o
i
:
1
0
.
4
2
1
8
/
e
t
r
i
j
.
2
0
2
3
-
0
3
5
4
.
[
5
]
Z.
C
h
e
n
e
t
a
l
.
,
“
S
A
L
M
:
s
p
e
e
c
h
-
a
u
g
men
t
e
d
l
a
n
g
u
a
g
e
m
o
d
e
l
w
i
t
h
i
n
-
c
o
n
t
e
x
t
l
e
a
r
n
i
n
g
f
o
r
s
p
e
e
c
h
r
e
c
o
g
n
i
t
i
o
n
a
n
d
t
r
a
n
sl
a
t
i
o
n
,
”
i
n
I
C
AS
S
P
2
0
2
4
-
2
0
2
4
I
E
EE
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
A
c
o
u
s
t
i
c
s
,
S
p
e
e
c
h
a
n
d
S
i
g
n
a
l
Pr
o
c
e
ssi
n
g
(
I
C
A
S
S
P)
,
I
EEE,
A
p
r
.
2
0
2
4
,
p
p
.
1
3
5
2
1
–
1
3
5
2
5
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
A
S
S
P
4
8
4
8
5
.
2
0
2
4
.
1
0
4
4
7
5
5
3
.
[
6
]
J.
Li
,
T.
Ta
n
g
,
W
.
X
.
Zh
a
o
,
J.
-
Y
.
N
i
e
,
a
n
d
J.
-
R
.
W
e
n
,
“
P
r
e
-
t
r
a
i
n
e
d
l
a
n
g
u
a
g
e
mo
d
e
l
s
f
o
r
t
e
x
t
g
e
n
e
r
a
t
i
o
n
:
a
su
r
v
e
y
,
”
AC
M
C
o
m
p
u
t
i
n
g
S
u
r
v
e
y
s
,
v
o
l
.
5
6
,
n
o
.
9
,
p
p
.
1
–
3
9
,
O
c
t
.
2
0
2
4
,
d
o
i
:
1
0
.
1
1
4
5
/
3
6
4
9
4
4
9
.
[
7
]
P
.
H
u
d
e
l
so
n
a
n
d
F
.
C
h
a
p
p
u
i
s,
“
U
si
n
g
v
o
i
c
e
-
to
-
v
o
i
c
e
m
a
c
h
i
n
e
t
r
a
n
sl
a
t
i
o
n
t
o
o
v
e
r
c
o
me
l
a
n
g
u
a
g
e
b
a
r
r
i
e
r
s
i
n
c
l
i
n
i
c
a
l
c
o
mm
u
n
i
c
a
t
i
o
n
:
a
n
e
x
p
l
o
r
a
t
o
r
y
s
t
u
d
y
,
”
J
o
u
r
n
a
l
o
f
G
e
n
e
r
a
l
I
n
t
e
rn
a
l
M
e
d
i
c
i
n
e
,
v
o
l
.
3
9
,
n
o
.
7
,
p
p
.
1
0
9
5
–
1
1
0
2
,
M
a
y
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
1
6
0
6
-
0
2
4
-
0
8
6
4
1
-
w.
[
8
]
M
.
V
i
j
a
y
k
u
m
a
r
,
V
.
R
.
S
r
i
,
a
n
d
M
.
S
.
M
a
p
p
i
l
l
a
i
,
“
A
s
t
u
d
y
o
n
c
h
a
t
b
o
t
s
a
n
d
v
i
r
t
u
a
l
a
s
s
i
s
t
a
n
t
s
i
n
c
u
s
t
o
m
e
r
e
n
g
a
g
e
m
e
n
t
:
a
r
e
v
i
e
w
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
E
n
g
i
n
e
e
r
i
n
g
a
n
d
M
a
n
a
g
e
m
e
n
t
R
e
s
e
a
r
c
h
,
v
o
l
.
1
4
,
n
o
.
1
,
p
p
.
2
0
4
–
2
0
8
,
2
0
2
4
,
d
o
i
:
1
0
.
5
2
8
1
/
z
e
n
o
d
o
.
1
0
7
9
1
6
9
7
.
[
9
]
K
.
R
o
h
i
t
,
A
.
S
h
a
n
k
a
r
,
G
.
K
a
t
i
y
a
r
,
A
.
M
e
h
r
o
t
r
a
,
a
n
d
E.
A
.
A
l
z
e
i
b
y
,
“
C
o
n
su
mer
e
n
g
a
g
e
me
n
t
i
n
c
h
a
t
b
o
t
s a
n
d
v
o
i
c
e
b
o
t
s
.
A
m
u
l
t
i
p
l
e
-
e
x
p
e
r
i
me
n
t
a
p
p
r
o
a
c
h
i
n
o
n
l
i
n
e
r
e
t
a
i
l
i
n
g
c
o
n
t
e
x
t
,
”
J
o
u
r
n
a
l
o
f
Re
t
a
i
l
i
n
g
a
n
d
C
o
n
s
u
m
e
r S
e
rv
i
c
e
s
,
v
o
l
.
7
8
,
p
.
1
0
3
7
2
8
,
M
a
y
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
r
e
t
c
o
n
ser.
2
0
2
4
.
1
0
3
7
2
8
.
[
1
0
]
G
.
B
.
S
h
a
n
k
a
r
,
N
.
S
u
r
e
s
h
,
a
n
d
T.
V
.
L
a
k
s
h
m
i
,
“
A
p
p
l
i
c
a
t
i
o
n
o
f
c
h
a
t
b
o
t
s
a
n
d
v
i
r
t
u
a
l
a
ssi
s
t
a
n
t
s
i
n
t
i
c
k
e
t
b
o
o
k
i
n
g
s
y
st
e
m,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
R
e
se
a
rc
h
J
o
u
rn
a
l
o
n
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
ri
n
g
a
n
d
M
a
n
a
g
e
m
e
n
t
(
I
RJ
AEM
)
,
v
o
l
.
2
,
n
o
.
0
5
,
p
p
.
1
6
0
5
–
1
6
0
8
,
M
a
y
2
0
2
4
,
d
o
i
:
1
0
.
4
7
3
9
2
/
I
R
JA
EM
.
2
0
2
4
.
0
2
2
1
.
[
1
1
]
Q
.
K
h
r
a
i
s
h
a
,
S
.
P
u
t
,
J.
K
a
p
p
e
n
b
e
r
g
,
A
.
W
a
r
r
a
i
t
c
h
,
a
n
d
K
.
H
a
d
f
i
e
l
d
,
“
C
a
n
l
a
r
g
e
l
a
n
g
u
a
g
e
mo
d
e
l
s
r
e
p
l
a
c
e
h
u
m
a
n
s
i
n
s
y
s
t
e
m
a
t
i
c
r
e
v
i
e
w
s?
E
v
a
l
u
a
t
i
n
g
G
P
T‐
4
’
s
e
f
f
i
c
a
c
y
i
n
s
c
r
e
e
n
i
n
g
a
n
d
e
x
t
r
a
c
t
i
n
g
d
a
t
a
f
r
o
m
p
e
e
r
‐
r
e
v
i
e
w
e
d
a
n
d
g
r
e
y
l
i
t
e
r
a
t
u
r
e
i
n
m
u
l
t
i
p
l
e
l
a
n
g
u
a
g
e
s
,
”
Re
s
e
a
r
c
h
S
y
n
t
h
e
s
i
s M
e
t
h
o
d
s
,
v
o
l
.
1
5
,
n
o
.
4
,
p
p
.
6
1
6
–
6
2
6
,
J
u
l
.
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
2
/
j
r
sm.
1
7
1
5
.
[
1
2
]
H
.
S
h
a
k
i
l
,
Z
.
O
r
t
i
z
,
a
n
d
G
.
C
.
F
o
r
b
e
s,
“
U
t
i
l
i
z
i
n
g
G
P
T
t
o
e
n
h
a
n
c
e
t
e
x
t
su
mm
a
r
i
z
a
t
i
o
n
:
a
st
r
a
t
e
g
y
t
o
m
i
n
i
mi
z
e
h
a
l
l
u
c
i
n
a
t
i
o
n
s
,
”
Arx
i
v
.
o
rg
/
a
b
s
/
2
4
0
5
.
0
4
0
3
9
,
M
a
y
2
0
2
4
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
2
,
Ap
r
il
20
25
:
2
3
2
5
-
2
3
3
3
2332
[
1
3
]
Z.
P
.
W
a
n
g
,
P
.
B
h
a
n
d
a
r
y
,
Y
.
W
a
n
g
,
a
n
d
J.
H
.
M
o
o
r
e
,
“
U
s
i
n
g
G
P
T
-
4
t
o
w
r
i
t
e
a
s
c
i
e
n
t
i
f
i
c
r
e
v
i
e
w
a
r
t
i
c
l
e
:
a
p
i
l
o
t
e
v
a
l
u
a
t
i
o
n
s
t
u
d
y
,
”
Bi
o
D
a
t
a
Mi
n
i
n
g
,
v
o
l
.
1
7
,
n
o
.
1
,
p
.
1
6
,
Ju
n
.
2
0
2
4
,
d
o
i
:
1
0
.
1
1
8
6
/
s1
3
0
4
0
-
0
2
4
-
0
0
3
7
1
-
3.
[
1
4
]
M
.
A
.
K
a
c
e
n
a
,
L.
I
.
P
l
o
t
k
i
n
,
a
n
d
J.
C
.
F
e
h
r
e
n
b
a
c
h
e
r
,
“
T
h
e
u
se
o
f
a
r
t
i
f
i
c
i
a
l
i
n
t
e
l
l
i
g
e
n
c
e
i
n
w
r
i
t
i
n
g
sc
i
e
n
t
i
f
i
c
r
e
v
i
e
w
a
r
t
i
c
l
e
s,
”
C
u
rr
e
n
t
O
s
t
e
o
p
o
r
o
si
s
Re
p
o
rt
s
,
v
o
l
.
2
2
,
n
o
.
1
,
p
p
.
1
1
5
–
1
2
1
,
F
e
b
.
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
1
9
1
4
-
0
2
3
-
0
0
8
5
2
-
0.
[
1
5
]
R
.
C
a
n
d
r
a
sar
i
,
J.
M
a
k
u
l
u
a
,
Y
.
N
o
v
i
a
s
my
,
K
.
M
a
k
u
l
u
a
,
a
n
d
S
.
S
i
mi
n
t
o
,
“
G
P
T
C
h
a
t
:
u
sefu
l
o
r
n
o
t
i
n
su
p
p
o
r
t
i
n
g
l
e
a
r
n
i
n
g
i
n
h
i
g
h
e
r
e
d
u
c
a
t
i
o
n
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
L
a
n
g
u
a
g
e
a
n
d
U
b
i
q
u
i
t
o
u
s Le
a
r
n
i
n
g
,
v
o
l
.
2
,
n
o
.
2
,
J
u
n
.
2
0
2
4
,
d
o
i
:
1
0
.
7
0
1
7
7
/
i
j
l
u
l
.
v
2
i
2
.
9
6
3
.
[
1
6
]
J.
D
o
u
g
h
t
y
e
t
a
l
.
,
“
A
c
o
mp
a
r
a
t
i
v
e
s
t
u
d
y
o
f
A
I
-
g
e
n
e
r
a
t
e
d
(
G
P
T
-
4
)
a
n
d
h
u
m
a
n
-
c
r
a
f
t
e
d
M
C
Q
s
i
n
p
r
o
g
r
a
mm
i
n
g
e
d
u
c
a
t
i
o
n
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
2
6
t
h
Au
s
t
ra
l
a
si
a
n
C
o
m
p
u
t
i
n
g
Ed
u
c
a
t
i
o
n
C
o
n
f
e
r
e
n
c
e
,
N
e
w
Y
o
r
k
,
N
Y
,
U
S
A
:
A
C
M
,
Ja
n
.
2
0
2
4
,
p
p
.
1
1
4
–
123
,
d
o
i
:
1
0
.
1
1
4
5
/
3
6
3
6
2
4
3
.
3
6
3
6
2
5
6
.
[
1
7
]
J.
M
.
P
r
i
e
t
o
A
n
d
r
e
u
a
n
d
A
.
L
.
P
a
l
mei
r
a
,
“
Q
u
i
c
k
r
e
v
i
e
w
o
f
p
e
d
a
g
o
g
i
c
a
l
e
x
p
e
r
i
e
n
c
e
s
u
s
i
n
g
G
P
T
-
3
i
n
e
d
u
c
a
t
i
o
n
,
”
J
o
u
rn
a
l
o
f
T
e
c
h
n
o
l
o
g
y
a
n
d
S
c
i
e
n
c
e
Ed
u
c
a
t
i
o
n
,
v
o
l
.
1
4
,
n
o
.
2
,
p
.
6
3
3
,
M
a
r
.
2
0
2
4
,
d
o
i
:
1
0
.
3
9
2
6
/
j
o
t
se
.
2
1
1
1
.
[
1
8
]
S.
-
V
.
F
u
l
g
e
n
c
i
o
,
“
D
e
v
e
l
o
p
i
n
g
e
f
f
e
c
t
i
v
e
e
d
u
c
a
t
i
o
n
a
l
c
h
a
t
b
o
t
s
w
i
t
h
G
P
T:
i
n
s
i
g
h
t
s
f
r
o
m
a
p
i
l
o
t
st
u
d
y
i
n
a
u
n
i
v
e
r
si
t
y
su
b
j
e
c
t
,
”
T
re
n
d
s
i
n
H
i
g
h
e
r E
d
u
c
a
t
i
o
n
,
v
o
l
.
3
,
n
o
.
1
,
p
p
.
1
5
5
–
1
6
8
,
M
a
r
.
2
0
2
4
,
d
o
i
:
1
0
.
3
3
9
0
/
h
i
g
h
e
r
e
d
u
3
0
1
0
0
0
9
.
[
1
9
]
M
.
E.
A
l
z
a
h
r
a
n
i
,
T
.
H
.
H
.
A
l
d
h
y
a
n
i
,
S
.
N
.
A
l
s
u
b
a
r
i
,
M
.
M
.
A
l
t
h
o
b
a
i
t
i
,
a
n
d
A
.
F
a
h
a
d
,
“
D
e
v
e
l
o
p
i
n
g
a
n
i
n
t
e
l
l
i
g
e
n
t
sy
s
t
e
m
w
i
t
h
d
e
e
p
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
ms
f
o
r
s
e
n
t
i
me
n
t
a
n
a
l
y
s
i
s
o
f
e
-
c
o
mm
e
r
c
e
p
r
o
d
u
c
t
r
e
v
i
e
w
s,”
C
o
m
p
u
t
a
t
i
o
n
a
l
I
n
t
e
l
l
i
g
e
n
c
e
a
n
d
N
e
u
r
o
sc
i
e
n
c
e
,
v
o
l
.
2
0
2
2
,
p
p
.
1
–
1
0
,
M
a
y
2
0
2
2
,
d
o
i
:
1
0
.
1
1
5
5
/
2
0
2
2
/
3
8
4
0
0
7
1
.
[
2
0
]
N
.
M
a
l
i
k
a
n
d
M
.
B
i
l
a
l
,
“
N
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
r
o
c
e
ssi
n
g
f
o
r
a
n
a
l
y
z
i
n
g
o
n
l
i
n
e
c
u
st
o
m
e
r
r
e
v
i
e
w
s:
a
s
u
r
v
e
y
,
t
a
x
o
n
o
m
y
,
a
n
d
o
p
e
n
r
e
sea
r
c
h
c
h
a
l
l
e
n
g
e
s,
”
P
e
e
r
J
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
1
0
,
p
.
e
2
2
0
3
,
J
u
l
.
2
0
2
4
,
d
o
i
:
1
0
.
7
7
1
7
/
p
e
e
r
j
-
c
s
.
2
2
0
3
.
[
2
1
]
A
.
S
a
r
r
a
f
,
“
U
t
i
l
i
z
i
n
g
N
LP
s
e
n
t
i
me
n
t
a
n
a
l
y
si
s
a
p
p
r
o
a
c
h
t
o
c
a
t
e
g
o
r
i
z
e
a
maz
o
n
r
e
v
i
e
w
s
a
g
a
i
n
s
t
a
n
e
x
t
e
n
d
e
d
t
e
st
i
n
g
s
e
t
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
C
o
m
p
u
t
e
r
(
I
J
C
)
,
v
o
l
.
5
0
,
n
o
.
1
,
p
p
.
1
0
7
–
1
1
6
,
2
0
2
4
.
[
2
2
]
X
.
M
a
,
Y
.
Li
,
a
n
d
M
.
A
si
f
,
“
E
-
c
o
m
merce
r
e
v
i
e
w
se
n
t
i
me
n
t
a
n
a
l
y
si
s
a
n
d
p
u
r
c
h
a
s
e
i
n
t
e
n
t
i
o
n
p
r
e
d
i
c
t
i
o
n
b
a
se
d
o
n
d
e
e
p
l
e
a
r
n
i
n
g
t
e
c
h
n
o
l
o
g
y
,
”
J
o
u
r
n
a
l
o
f
O
r
g
a
n
i
za
t
i
o
n
a
l
a
n
d
En
d
U
ser
C
o
m
p
u
t
i
n
g
,
v
o
l
.
3
6
,
n
o
.
1
,
p
p
.
1
–
2
9
,
D
e
c
.
2
0
2
3
,
d
o
i
:
1
0
.
4
0
1
8
/
JO
EU
C
.
3
3
5
1
2
2
.
[
2
3
]
M
.
R
a
n
j
a
n
,
S
.
Ti
w
a
r
i
,
A
.
M
.
S
a
t
t
a
r
,
a
n
d
N
.
S
.
Ta
t
k
a
r
,
“
A
n
e
w
a
p
p
r
o
a
c
h
f
o
r
c
a
r
r
y
i
n
g
o
u
t
se
n
t
i
m
e
n
t
a
n
a
l
y
s
i
s
o
f
s
o
c
i
a
l
me
d
i
a
c
o
mm
e
n
t
s
u
s
i
n
g
n
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
r
o
c
e
ssi
n
g
,
”
i
n
R
Ai
S
E
-
2
0
2
3
,
B
a
sel
S
w
i
t
z
e
r
l
a
n
d
:
M
D
P
I
,
Ja
n
.
2
0
2
4
,
p
.
1
8
1
,
doi
:
1
0
.
3
3
9
0
/
e
n
g
p
r
o
c
2
0
2
3
0
5
9
1
8
1
.
[
2
4
]
“
C
o
l
l
i
n
s
E
n
g
l
i
sh
d
i
c
t
i
o
n
a
r
y
-
c
o
m
p
l
e
t
e
& u
n
a
b
r
i
d
g
e
d
2
0
1
2
d
i
g
i
t
a
l
e
d
i
t
i
o
n
,
”
W
i
l
l
i
a
m
C
o
l
l
i
n
s
S
o
n
s
&
Co
.
LTD
.
A
c
c
e
sse
d
:
N
o
v
.
0
2
,
2
0
2
3
.
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s
:
/
/
w
w
w
.
d
i
c
t
i
o
n
a
r
y
.
c
o
m
[
2
5
]
D
.
K
h
y
a
n
i
,
B
.
S
.
S
i
d
d
h
a
r
t
h
a
,
N
.
M
.
N
i
v
e
d
i
t
h
a
,
B
.
M
.
D
i
v
y
a
,
a
n
d
Y
.
M
.
M
a
n
u
,
“
A
n
i
n
t
e
r
p
r
e
t
a
t
i
o
n
o
f
l
e
mm
a
t
i
z
a
t
i
o
n
a
n
d
st
e
mm
i
n
g
i
n
n
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
r
o
c
e
ssi
n
g
,
”
J
o
u
rn
a
l
o
f
U
n
i
v
e
rsi
t
y
o
f
S
h
a
n
g
h
a
i
f
o
r
S
c
i
e
n
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
2
2
,
n
o
.
1
0
,
p
p
.
3
5
0
–
3
5
7
,
2
0
2
1
.
[
2
6
]
L.
G
u
o
h
u
a
n
,
Z
.
H
a
o
,
a
n
d
W
.
H
o
n
g
h
u
i
,
“
R
e
se
a
r
c
h
o
n
m
e
t
h
o
d
s
o
f
s
e
ma
n
t
i
c
d
i
s
a
mb
i
g
u
a
t
i
o
n
a
b
o
u
t
n
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
r
o
c
e
ss
i
n
g
,
”
i
n
2
0
0
9
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
Wi
r
e
l
e
ss
N
e
t
w
o
rks
a
n
d
I
n
f
o
rm
a
t
i
o
n
S
y
st
e
m
s
,
I
EEE,
D
e
c
.
2
0
0
9
,
p
p
.
3
4
7
–
3
4
9
,
d
o
i
:
1
0
.
1
1
0
9
/
W
N
I
S
.
2
0
0
9
.
2
1
.
[
2
7
]
B.
W
e
b
b
e
r
,
M
.
E
g
g
,
a
n
d
V
.
K
o
r
d
o
n
i
,
“
D
i
s
c
o
u
r
se
s
t
r
u
c
t
u
r
e
a
n
d
l
a
n
g
u
a
g
e
t
e
c
h
n
o
l
o
g
y
,
”
N
a
t
u
ra
l
L
a
n
g
u
a
g
e
En
g
i
n
e
e
ri
n
g
,
v
o
l
.
1
8
,
n
o
.
4
,
p
p
.
4
3
7
–
4
9
0
,
O
c
t
.
2
0
1
2
,
d
o
i
:
1
0
.
1
0
1
7
/
S
1
3
5
1
3
2
4
9
1
1
0
0
0
3
3
7
.
[
2
8
]
S
.
J.
a
n
d
S
.
S
w
a
my
,
“
A
p
r
i
o
r
c
a
s
e
s
t
u
d
y
o
f
n
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
r
o
c
e
ssi
n
g
o
n
d
i
f
f
e
r
e
n
t
d
o
m
a
i
n
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
El
e
c
t
ri
c
a
l
a
n
d
C
o
m
p
u
t
e
r E
n
g
i
n
e
e
ri
n
g
,
v
o
l
.
1
0
,
n
o
.
5
,
p
.
4
9
2
8
,
O
c
t
.
2
0
2
0
,
d
o
i
:
1
0
.
1
1
5
9
1
/
i
j
e
c
e
.
v
1
0
i
5
.
p
p
4
9
2
8
-
4
9
3
6
.
[
2
9
]
P
.
K
l
o
so
w
s
k
i
,
“
D
e
e
p
l
e
a
r
n
i
n
g
f
o
r
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
n
d
l
a
n
g
u
a
g
e
mo
d
e
l
l
i
n
g
,
”
i
n
2
0
1
8
S
i
g
n
a
l
Pro
c
e
ss
i
n
g
:
A
l
g
o
ri
t
h
m
s,
Arc
h
i
t
e
c
t
u
r
e
s,
Arr
a
n
g
e
m
e
n
t
s,
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s (
S
PA)
,
I
EEE,
S
e
p
.
2
0
1
8
,
p
p
.
2
2
3
–
2
2
8
,
d
o
i
:
1
0
.
2
3
9
1
9
/
S
P
A
.
2
0
1
8
.
8
5
6
3
3
8
9
.
[
3
0
]
D
.
K
h
u
r
a
n
a
,
A
.
K
o
l
i
,
K
.
K
h
a
t
t
e
r
,
a
n
d
S
.
S
i
n
g
h
,
“
N
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
r
o
c
e
ssi
n
g
:
s
t
a
t
e
o
f
t
h
e
a
r
t
,
c
u
r
r
e
n
t
t
r
e
n
d
s
a
n
d
c
h
a
l
l
e
n
g
e
s
,
”
Mu
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
.
8
2
,
n
o
.
3
,
p
p
.
3
7
1
3
–
3
7
4
4
,
Ja
n
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
1
0
4
2
-
0
2
2
-
1
3
4
2
8
-
4.
[
3
1
]
M
.
Y
.
L
a
n
d
o
l
s
i
,
L.
H
l
a
o
u
a
,
a
n
d
L
.
B
e
n
R
o
m
d
h
a
n
e
,
“
I
n
f
o
r
ma
t
i
o
n
e
x
t
r
a
c
t
i
o
n
f
r
o
m
e
l
e
c
t
r
o
n
i
c
me
d
i
c
a
l
d
o
c
u
me
n
t
s
:
st
a
t
e
o
f
t
h
e
a
r
t
a
n
d
f
u
t
u
r
e
r
e
sea
r
c
h
d
i
r
e
c
t
i
o
n
s,
”
K
n
o
w
l
e
d
g
e
a
n
d
I
n
f
o
rm
a
t
i
o
n
S
y
s
t
e
m
s
,
v
o
l
.
6
5
,
n
o
.
2
,
p
p
.
4
6
3
–
5
1
6
,
F
e
b
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
0
1
1
5
-
0
2
2
-
0
1
7
7
9
-
1.
[
3
2
]
Y
.
F
a
n
e
t
a
l
.
,
“
P
r
e
-
t
r
a
i
n
i
n
g
m
e
t
h
o
d
s
i
n
i
n
f
o
r
mat
i
o
n
r
e
t
r
i
e
v
a
l
,
”
Fo
u
n
d
a
t
i
o
n
s
a
n
d
T
re
n
d
s
®
i
n
I
n
f
o
rm
a
t
i
o
n
R
e
t
r
i
e
v
a
l
,
v
o
l
.
1
6
,
n
o
.
3
,
p
p
.
1
7
8
–
3
1
7
,
2
0
2
2
,
d
o
i
:
1
0
.
1
5
6
1
/
1
5
0
0
0
0
0
1
0
0
.
[
3
3
]
J.
L
i
n
,
“
A
p
r
o
p
o
s
e
d
c
o
n
c
e
p
t
u
a
l
f
r
a
m
e
w
o
r
k
f
o
r
a
r
e
p
r
e
s
e
n
t
a
t
i
o
n
a
l
a
p
p
r
o
a
c
h
t
o
i
n
f
o
r
mat
i
o
n
r
e
t
r
i
e
v
a
l
,
”
AC
M
S
I
G
I
R
F
o
ru
m
,
v
o
l
.
5
5
,
n
o
.
2
,
p
p
.
1
–
2
9
,
D
e
c
.
2
0
2
1
,
d
o
i
:
1
0
.
1
1
4
5
/
3
5
2
7
5
4
6
.
3
5
2
7
5
5
2
.
[
3
4
]
M
.
W
a
h
d
e
a
n
d
M
.
V
i
r
g
o
l
i
n
,
“
C
o
n
v
e
r
sa
t
i
o
n
a
l
a
g
e
n
t
s
:
t
h
e
o
r
y
a
n
d
a
p
p
l
i
c
a
t
i
o
n
s,”
i
n
H
a
n
d
b
o
o
k
o
n
C
o
m
p
u
t
e
r
L
e
a
r
n
i
n
g
a
n
d
I
n
t
e
l
l
i
g
e
n
c
e
,
p
p
.
4
9
7
–
544
,
d
o
i
:
1
0
.
1
1
4
2
/
9
7
8
9
8
1
1
2
4
7
3
2
3
_
0
0
1
2
.
[
3
5
]
D
.
Y
a
d
a
v
,
J.
D
e
sa
i
,
a
n
d
A
.
K
.
Y
a
d
a
v
,
“
A
u
t
o
ma
t
i
c
t
e
x
t
su
mm
a
r
i
z
a
t
i
o
n
me
t
h
o
d
s:
a
c
o
mp
r
e
h
e
n
si
v
e
r
e
v
i
e
w
,
”
Arx
i
v
.
O
r
g
/
A
b
s/
2
2
0
4
.
0
1
8
4
9
,
M
a
r
.
2
0
2
2
.
[
3
6
]
A
.
P
.
W
i
d
y
a
ss
a
r
i
e
t
a
l
.
,
“
R
e
v
i
e
w
o
f
a
u
t
o
ma
t
i
c
t
e
x
t
s
u
mm
a
r
i
z
a
t
i
o
n
t
e
c
h
n
i
q
u
e
s
&am
p
;
me
t
h
o
d
s,”
J
o
u
r
n
a
l
o
f
K
i
n
g
S
a
u
d
U
n
i
v
e
rsi
t
y
-
C
o
m
p
u
t
e
r
a
n
d
I
n
f
o
rm
a
t
i
o
n
S
c
i
e
n
c
e
s
,
v
o
l
.
3
4
,
n
o
.
4
,
p
p
.
1
0
2
9
–
1
0
4
6
,
A
p
r
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
k
s
u
c
i
.
2
0
2
0
.
0
5
.
0
0
6
.
[
3
7
]
A
.
T
o
r
f
i
,
R
.
A
.
S
h
i
r
v
a
n
i
,
Y
.
K
e
n
e
sh
l
o
o
,
N
.
T
a
v
a
f
,
a
n
d
E
.
A
.
F
o
x
,
“
N
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
r
o
c
e
ssi
n
g
a
d
v
a
n
c
e
m
e
n
t
s
b
y
d
e
e
p
l
e
a
r
n
i
n
g
:
a
su
r
v
e
y
,
”
Arx
i
v
.
O
r
g
/
Ab
s
/
2
0
0
3
.
0
1
2
0
0
,
M
a
r
.
2
0
2
0
.
[
3
8
]
B
.
F
.
G
r
e
e
n
,
A
.
K
.
W
o
l
f
,
C
.
C
h
o
ms
k
y
,
a
n
d
K
.
La
u
g
h
e
r
y
,
“
B
a
se
b
a
l
l
:
a
n
a
u
t
o
ma
t
i
c
q
u
e
st
i
o
n
-
a
n
sw
e
r
e
r
,
”
i
n
Pa
p
e
rs
p
r
e
se
n
t
e
d
a
t
t
h
e
Ma
y
9
-
1
1
,
1
9
6
1
,
w
e
s
t
e
r
n
j
o
i
n
t
I
RE
-
A
I
EE
-
AC
M
c
o
m
p
u
t
e
r
c
o
n
f
e
r
e
n
c
e
o
n
-
I
RE
-
AI
EE
-
A
C
M
’
6
1
(
W
e
st
e
rn
)
,
N
e
w
Y
o
r
k
,
N
e
w
Y
o
r
k
,
U
S
A
:
A
C
M
P
r
e
ss,
1
9
6
1
,
p
.
2
1
9
,
d
o
i
:
1
0
.
1
1
4
5
/
1
4
6
0
6
9
0
.
1
4
6
0
7
1
4
.
[
3
9
]
H
.
B
e
n
-
y
o
u
n
e
s
,
R
.
C
a
d
e
n
e
,
N
.
T
h
o
m
e
,
a
n
d
M
.
C
o
r
d
,
“
B
LO
C
K
:
b
i
l
i
n
e
a
r
s
u
p
e
r
d
i
a
g
o
n
a
l
f
u
s
i
o
n
f
o
r
v
i
su
a
l
q
u
e
st
i
o
n
a
n
sw
e
r
i
n
g
a
n
d
v
i
s
u
a
l
r
e
l
a
t
i
o
n
sh
i
p
d
e
t
e
c
t
i
o
n
,
”
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
AAAI
C
o
n
f
e
re
n
c
e
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
.
3
3
,
n
o
.
0
1
,
p
p
.
8
1
0
2
–
8
1
0
9
,
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
8
1
0
2
.
[
4
0
]
T.
H
a
o
,
X
.
Li
,
Y
.
H
e
,
F
.
L.
W
a
n
g
,
a
n
d
Y
.
Q
u
,
“
R
e
c
e
n
t
p
r
o
g
r
e
ss
i
n
l
e
v
e
r
a
g
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
m
e
t
h
o
d
s
f
o
r
q
u
e
st
i
o
n
a
n
sw
e
r
i
n
g
,
”
N
e
u
ra
l
C
o
m
p
u
t
i
n
g
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
3
4
,
n
o
.
4
,
p
p
.
2
7
6
5
–
2
7
8
3
,
F
e
b
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
7
/
s
0
0
5
2
1
-
021
-
0
6
7
4
8
-
3.
[
4
1
]
M
.
D
.
O
k
p
o
r
,
“
M
a
c
h
i
n
e
t
r
a
n
s
l
a
t
i
o
n
a
p
p
r
o
a
c
h
e
s:
i
ss
u
e
s
a
n
d
c
h
a
l
l
e
n
g
e
s,”
I
J
C
S
I
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
I
ss
u
e
s
,
v
o
l
.
1
1
,
n
o
.
5
,
p
p
.
1
5
9
–
1
6
5
,
2
0
1
4
.
[
4
2
]
I
.
V
a
y
a
n
s
k
y
a
n
d
S
.
A
.
P
.
K
u
m
a
r
,
“
A
r
e
v
i
e
w
o
f
t
o
p
i
c
m
o
d
e
l
i
n
g
me
t
h
o
d
s,
”
I
n
f
o
rm
a
t
i
o
n
S
y
s
t
e
m
s
,
v
o
l
.
9
4
,
p
.
1
0
1
5
8
2
,
D
e
c
.
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
s.
2
0
2
0
.
1
0
1
5
8
2
.
[
4
3
]
Q
.
F
u
,
Y
.
Z
h
u
a
n
g
,
J.
G
u
,
Y
.
Zh
u
,
a
n
d
X
.
G
u
o
,
“
A
g
r
e
e
i
n
g
t
o
d
i
sa
g
r
e
e
:
c
h
o
o
si
n
g
a
mo
n
g
e
i
g
h
t
t
o
p
i
c
-
mo
d
e
l
i
n
g
me
t
h
o
d
s,
”
B
i
g
D
a
t
a
Re
se
a
rc
h
,
v
o
l
.
2
3
,
p
.
1
0
0
1
7
3
,
F
e
b
.
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
b
d
r
.
2
0
2
0
.
1
0
0
1
7
3
.
[
4
4
]
Y
.
K
.
D
w
i
v
e
d
i
e
t
a
l
.
,
“
O
p
i
n
i
o
n
p
a
p
e
r
:
‘
so
w
h
a
t
i
f
C
h
a
t
G
P
T
w
r
o
t
e
i
t
?
’
mu
l
t
i
d
i
sc
i
p
l
i
n
a
r
y
p
e
r
s
p
e
c
t
i
v
e
s
o
n
o
p
p
o
r
t
u
n
i
t
i
e
s,
c
h
a
l
l
e
n
g
e
s
a
n
d
i
mp
l
i
c
a
t
i
o
n
s
o
f
g
e
n
e
r
a
t
i
v
e
c
o
n
v
e
r
sa
t
i
o
n
a
l
A
I
f
o
r
r
e
sea
r
c
h
,
p
r
a
c
t
i
c
e
a
n
d
p
o
l
i
c
y
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
I
n
f
o
rm
a
t
i
o
n
Ma
n
a
g
e
m
e
n
t
,
v
o
l
.
7
1
,
p
.
1
0
2
6
4
2
,
A
u
g
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
j
i
n
f
o
m
g
t
.
2
0
2
3
.
1
0
2
6
4
2
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
Th
e
a
d
va
n
c
es in
n
a
tu
r
a
l la
n
g
u
a
g
e
p
r
o
ce
s
s
in
g
tech
n
o
lo
g
y
a
n
d
its
imp
a
ct
o
n
… (
N
a
d
ezh
d
a
B
o
r
is
o
va
)
2333
[
4
5
]
G
.
K
e
n
d
a
l
l
a
n
d
J.
A
.
Te
i
x
e
i
r
a
d
a
S
i
l
v
a
,
“
R
i
s
k
s
o
f
a
b
u
se
o
f
l
a
r
g
e
l
a
n
g
u
a
g
e
m
o
d
e
l
s
,
l
i
k
e
C
h
a
t
G
P
T,
i
n
sc
i
e
n
t
i
f
i
c
p
u
b
l
i
s
h
i
n
g
:
A
u
t
h
o
r
s
h
i
p
,
p
r
e
d
a
t
o
r
y
p
u
b
l
i
s
h
i
n
g
,
a
n
d
p
a
p
e
r
m
i
l
l
s,”
L
e
a
r
n
e
d
Pu
b
l
i
s
h
i
n
g
,
v
o
l
.
3
7
,
n
o
.
1
,
p
p
.
5
5
–
6
2
,
Jan
.
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
2
/
l
e
a
p
.
1
5
7
8
.
[
4
6
]
K
.
F
u
c
h
s,
“
E
x
p
l
o
r
i
n
g
t
h
e
o
p
p
o
r
t
u
n
i
t
i
e
s
a
n
d
c
h
a
l
l
e
n
g
e
s
o
f
N
LP
m
o
d
e
l
s
i
n
h
i
g
h
e
r
e
d
u
c
a
t
i
o
n
:
i
s
C
h
a
t
G
P
T
a
b
l
e
s
si
n
g
o
r
a
c
u
r
s
e
?
,
”
Fro
n
t
i
e
rs
i
n
E
d
u
c
a
t
i
o
n
,
v
o
l
.
8
,
M
a
y
2
0
2
3
,
d
o
i
:
1
0
.
3
3
8
9
/
f
e
d
u
c
.
2
0
2
3
.
1
1
6
6
6
8
2
.
[
4
7
]
D
.
H
.
R
.
S
p
e
n
n
e
ma
n
n
,
“
Ex
p
l
o
r
i
n
g
e
t
h
i
c
a
l
b
o
u
n
d
a
r
i
e
s
:
c
a
n
C
h
a
t
G
P
T
b
e
p
r
o
mp
t
e
d
t
o
g
i
v
e
a
d
v
i
c
e
o
n
h
o
w
t
o
c
h
e
a
t
i
n
u
n
i
v
e
r
si
t
y
a
ssi
g
n
m
e
n
t
s?
”
Pr
e
p
r
i
n
t
s
,
A
u
g
.
1
7
,
2
0
2
3
,
d
o
i
:
1
0
.
2
0
9
4
4
/
p
r
e
p
r
i
n
t
s
2
0
2
3
0
8
.
1
2
7
1
.
v
1
.
[
4
8
]
S
.
S
.
B
i
sw
a
s,
“
P
o
t
e
n
t
i
a
l
u
se
o
f
c
h
a
t
G
P
T
i
n
g
l
o
b
a
l
w
a
r
mi
n
g
,
”
A
n
n
a
l
s
o
f
Bi
o
m
e
d
i
c
a
l
E
n
g
i
n
e
e
r
i
n
g
,
v
o
l
.
5
1
,
n
o
.
6
,
p
p
.
1
1
2
6
–
1
1
2
7
,
Ju
n
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
0
4
3
9
-
0
2
3
-
0
3
1
7
1
-
8.
[
4
9
]
A
.
A
l
i
b
a
si
c
a
n
d
T.
P
o
p
o
v
i
c
,
“
A
p
p
l
y
i
n
g
n
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
r
o
c
e
ssi
n
g
t
o
a
n
a
l
y
z
e
c
u
s
t
o
m
e
r
s
a
t
i
sf
a
c
t
i
o
n
,
”
i
n
2
0
2
1
2
5
t
h
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
I
n
f
o
rm
a
t
i
o
n
T
e
c
h
n
o
l
o
g
y
(
I
T
)
,
I
EEE,
F
e
b
.
2
0
2
1
,
p
p
.
1
–
4
,
d
o
i
:
1
0
.
1
1
0
9
/
I
T5
1
5
2
8
.
2
0
2
1
.
9
3
9
0
1
1
1
.
[
5
0
]
X
.
Li
n
,
“
S
e
n
t
i
m
e
n
t
a
n
a
l
y
s
i
s
o
f
e
-
c
o
m
merce
c
u
st
o
mer
r
e
v
i
e
w
s
b
a
se
d
o
n
n
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
r
o
c
e
ssi
n
g
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
2
0
2
0
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
Bi
g
D
a
t
a
a
n
d
Ar
t
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
,
N
e
w
Y
o
r
k
,
N
Y
,
U
S
A
:
A
C
M
,
A
p
r
.
2
0
2
0
,
p
p
.
3
2
–
36
,
d
o
i
:
1
0
.
1
1
4
5
/
3
4
3
6
2
8
6
.
3
4
3
6
2
9
3
.
[
5
1
]
U
.
S
i
n
g
h
,
A
.
S
a
r
a
sw
a
t
,
H
.
K
.
A
z
a
d
,
K
.
A
b
h
i
s
h
e
k
,
a
n
d
S
.
S
h
i
t
h
a
r
t
h
,
“
To
w
a
r
d
s i
m
p
r
o
v
i
n
g
e
-
c
o
mm
e
r
c
e
c
u
st
o
mer r
e
v
i
e
w
a
n
a
l
y
s
i
s f
o
r
sen
t
i
m
e
n
t
d
e
t
e
c
t
i
o
n
,
”
S
c
i
e
n
t
i
f
i
c
R
e
p
o
r
t
s
,
v
o
l
.
1
2
,
n
o
.
1
,
p
.
2
1
9
8
3
,
D
e
c
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
3
8
/
s4
1
5
9
8
-
0
2
2
-
2
6
4
3
2
-
3.
[
5
2
]
M
.
M
a
s
h
a
a
b
i
,
A
.
A
l
o
t
a
i
b
i
,
H
.
Q
u
d
a
i
h
,
R
.
A
l
n
a
sh
w
a
n
,
a
n
d
H
.
A
l
-
K
h
a
l
i
f
a
,
“
N
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
r
o
c
e
ss
i
n
g
i
n
c
u
s
t
o
m
e
r
ser
v
i
c
e
:
a
sy
st
e
ma
t
i
c
r
e
v
i
e
w
,
”
Arx
i
v
.
o
r
g
/
a
b
s/
2
2
1
2
.
0
9
5
2
3
,
D
e
c
.
2
0
2
2
.
[
5
3
]
P
.
A
.
O
l
u
j
i
m
i
a
n
d
A
.
A
d
e
-
I
b
i
j
o
l
a
,
“
N
LP
t
e
c
h
n
i
q
u
e
s
f
o
r
a
u
t
o
ma
t
i
n
g
r
e
sp
o
n
s
e
s
t
o
c
u
s
t
o
mer
q
u
e
r
i
e
s
:
a
sy
s
t
e
m
a
t
i
c
r
e
v
i
e
w
,
”
D
i
sc
o
v
e
r
Art
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
,
v
o
l
.
3
,
n
o
.
1
,
p
.
2
0
,
M
a
y
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
7
/
s
4
4
1
6
3
-
0
2
3
-
0
0
0
6
5
-
5.
[
5
4
]
A
.
M
o
s
b
y
,
“
6
9
v
o
i
c
e
s
e
a
r
c
h
s
t
a
t
i
s
t
i
c
s
2
0
2
4
(
u
s
a
g
e
&
d
e
m
o
g
r
a
p
h
i
c
s)
,
”
Y
a
g
u
a
r
a
.
c
o
,
h
t
t
p
s:
/
/
w
w
w
.
y
a
g
u
a
r
a
.
c
o
/
v
o
i
c
e
-
se
a
r
c
h
-
st
a
t
i
st
i
c
s
/
#
:
~
:
t
e
x
t
=
S
i
r
i
S
t
a
t
i
st
i
c
s,
d
r
i
v
i
n
g
o
r
i
n
t
h
e
i
r
c
a
r
s
(a
c
c
e
sse
d
:
A
u
g
.
1
5
,
2
0
2
4
)
.
[
5
5
]
R
.
P
o
n
n
a
b
o
y
i
n
a
,
R
.
M
a
k
a
l
a
,
a
n
d
E
.
V
.
R
e
d
d
y
,
“
S
mart
r
e
c
r
u
i
t
m
e
n
t
sy
s
t
e
m
u
si
n
g
d
e
e
p
l
e
a
r
n
i
n
g
w
i
t
h
n
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
r
o
c
e
ss
i
n
g
,
”
i
n
I
n
t
e
l
l
i
g
e
n
t
S
y
s
t
e
m
s
a
n
d
S
u
st
a
i
n
a
b
l
e
C
o
m
p
u
t
i
n
g
.
S
m
a
rt
I
n
n
o
v
a
t
i
o
n
,
S
y
s
t
e
m
s
a
n
d
T
e
c
h
n
o
l
o
g
i
e
s
,
2
0
2
2
,
p
p
.
6
4
7
–
655
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
981
-
19
-
0
0
1
1
-
2
_
5
8
.
[
5
6
]
C
.
M
a
d
d
u
m
a
g
e
,
D
.
S
e
n
e
v
i
r
a
t
h
n
e
,
I
.
G
a
y
a
s
h
a
n
,
T
.
S
h
e
h
a
n
,
a
n
d
S
.
S
u
m
a
t
h
i
p
a
l
a
,
“
I
n
t
e
l
l
i
g
e
n
t
r
e
c
r
u
i
t
m
e
n
t
s
y
s
t
e
m
,
”
i
n
2
0
1
9
I
E
E
E
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
f
o
r
C
o
n
v
e
r
g
e
n
c
e
i
n
T
e
c
h
n
o
l
o
g
y
(
I
2
C
T
)
,
I
E
E
E
,
M
a
r
.
2
0
1
9
,
p
p
.
1
–
6
,
d
o
i
:
1
0
.
1
1
0
9
/
I
2
C
T
4
5
6
1
1
.
2
0
1
9
.
9
0
3
3
8
3
6
.
B
I
O
G
RAP
H
I
E
S
OF
AUTH
O
RS
Na
d
e
z
h
d
a
Bo
r
is
o
v
a
re
c
e
iv
e
d
th
e
M
.
S
.
a
n
d
P
h
.
D.
d
e
g
re
e
s
in
c
o
m
p
u
ter
sc
ien
c
e
fro
m
S
o
u
t
h
-
Wes
t
Un
i
v
e
rsity
“
Ne
o
fit
Ril
sk
i
,
”
Blag
o
e
v
g
ra
d
,
Bu
lg
a
ria,
in
2
0
0
0
a
n
d
2
0
1
5
,
re
sp
e
c
ti
v
e
ly
.
C
u
rre
n
t
ly
,
sh
e
is
a
n
a
ss
istan
t
p
r
o
fe
ss
o
r
a
t
t
h
e
De
p
a
rt
m
e
n
t
o
f
In
f
o
rm
a
ti
c
s,
S
o
u
t
h
-
Wes
t
Un
iv
e
rsity
.
He
r
a
re
a
o
f
re
se
a
rc
h
is
NLP
a
n
d
d
a
ta
m
in
in
g
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
n
b
o
riso
v
a
@s
wu
.
b
g
.
Elena
K
a
r
a
shtra
n
o
v
a
is
a
n
a
ss
o
c
iate
p
ro
fe
ss
o
r
a
t
th
e
De
p
a
rtme
n
t
o
f
I
n
fo
rm
a
ti
c
s,
S
o
u
t
h
-
Wes
t
Un
i
v
e
rsity
.
He
r
a
re
a
o
f
re
se
a
rc
h
is
sta
ti
stic
,
NLP
a
n
d
d
a
ta
m
in
i
n
g
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
h
e
len
@s
wu
.
b
g
.
Ir
e
n
a
Ata
n
a
so
v
a
is
a
n
a
ss
o
c
iate
p
r
o
fe
ss
o
r
a
n
d
h
e
a
d
o
f
t
h
e
De
p
a
rtme
n
t
o
f
In
fo
rm
a
ti
c
s,
S
o
u
t
h
-
Wes
t
Un
i
v
e
r
sity
.
He
r
a
re
a
o
f
re
se
a
rc
h
is
a
rti
ficia
l
in
telli
g
e
n
t
a
n
d
d
a
ta
m
in
in
g
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
iren
a
tm@
sw
u
.
b
g
.
Evaluation Warning : The document was created with Spire.PDF for Python.