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.
14
,
No
.
6
,
Dec
em
b
er
20
24
,
p
p
.
6
7
1
1
~
6
7
2
2
I
SS
N:
2088
-
8
7
0
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijece.
v
14
i
6
.
pp
6
7
1
1
-
6
7
2
2
6711
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m
Aspect
-
ba
sed sen
t
iment ana
ly
sis
:
n
a
tural la
ng
ua
g
e
understa
nding
for implici
t
revi
ew
Su
ha
riy
a
nto
1,
2
,
Riy
a
na
rt
o
Sa
rno
1
,
Cha
s
t
ine F
a
t
icha
h
1
,
Ra
chm
a
d Abdu
lla
h
1,
3
1
D
e
p
a
r
t
me
n
t
o
f
I
n
f
o
r
mat
i
c
s
,
F
a
c
u
l
t
y
o
f
I
n
t
e
l
l
i
g
e
n
t
El
e
c
t
r
i
c
a
l
a
n
d
I
n
f
o
r
m
a
t
i
c
s
Te
c
h
n
o
l
o
g
y
,
I
n
st
i
t
u
t
T
e
k
n
o
l
o
g
i
S
e
p
u
l
u
h
N
o
p
e
m
b
e
r
,
S
u
r
a
b
a
y
a
,
I
n
d
o
n
e
si
a
2
F
a
c
u
l
t
y
o
f
C
o
mp
u
t
e
r
S
c
i
e
n
c
e
,
D
i
a
n
N
u
sw
a
n
t
o
r
o
U
n
i
v
e
r
s
i
t
y
,
S
e
m
a
r
a
n
g
,
I
n
d
o
n
e
s
i
a
3
D
e
p
a
r
m
e
n
t
o
f
M
a
n
a
g
e
me
n
t
,
S
e
k
o
l
a
h
Ti
n
g
g
i
I
l
mu
E
k
o
n
o
m
i
I
n
d
o
n
e
si
a
,
S
u
r
a
b
a
y
a
,
I
n
d
o
n
e
si
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Ma
y
5
,
2
0
2
4
R
ev
is
ed
Au
g
2
,
2
0
2
4
Acc
ep
ted
Au
g
6
,
2
0
2
4
Th
e
d
iffere
n
t
ty
p
e
s
o
f
imp
li
c
it
re
v
iew
s
sh
o
u
ld
b
e
we
ll
u
n
d
e
rsto
o
d
so
th
a
t
t
h
e
d
e
v
e
lo
p
e
d
e
x
trac
ti
o
n
tec
h
n
iq
u
e
c
a
n
so
lv
e
a
ll
p
ro
b
lem
s
in
imp
li
c
it
re
v
iew
s
a
n
d
p
r
o
d
u
c
e
p
re
c
ise
term
s
o
f
a
s
p
e
c
ts
a
n
d
o
p
in
i
o
n
s.
We
p
ro
p
o
se
a
n
a
sp
e
c
t
-
b
a
se
d
se
n
ti
m
e
n
t
a
n
a
ly
sis
(A
BS
A)
m
e
th
o
d
wit
h
n
a
t
u
ra
l
lan
g
u
a
g
e
u
n
d
e
rsta
n
d
i
n
g
fo
r
im
p
li
c
it
re
v
ie
ws
b
a
se
d
o
n
se
n
ten
c
e
a
n
d
wo
rd
stru
c
tu
re
.
We
b
u
il
t
a
te
x
t
e
x
trac
ti
o
n
m
e
th
o
d
u
sin
g
a
m
a
c
h
in
e
lea
rn
in
g
a
lg
o
rit
h
m
r
u
le
with
a
d
e
e
p
u
n
d
e
rsta
n
d
in
g
o
f
d
iffere
n
t
ty
p
e
s
o
f
se
n
ten
c
e
s
a
n
d
wo
r
d
s.
F
u
rth
e
rm
o
re
,
th
e
a
sp
e
c
t
c
a
teg
o
ry
o
f
e
a
c
h
re
v
iew
is
d
e
term
in
e
d
b
y
m
e
a
su
rin
g
th
e
wo
rd
sim
il
a
rit
y
b
e
twe
e
n
th
e
a
sp
e
c
t
term
s
c
o
n
tain
e
d
i
n
e
a
c
h
re
v
iew
a
n
d
a
sp
e
c
t
k
e
y
wo
rd
s
e
x
trac
ted
f
ro
m
Wi
k
i
p
e
d
ia.
Bi
d
irec
ti
o
n
a
l
e
n
c
o
d
e
r
re
p
re
se
n
tatio
n
s
fr
o
m
tran
sf
o
rm
e
rs
(
BERT
)
e
m
b
e
d
d
in
g
a
n
d
se
m
a
n
ti
c
sim
il
a
rit
y
a
re
u
se
d
to
m
e
a
su
re
th
e
wo
rd
sim
il
a
rit
y
v
a
lu
e
.
M
o
r
e
o
v
e
r,
th
e
p
ro
p
o
se
d
ABSA
m
e
th
o
d
u
se
s
BERT
,
a
h
y
b
ri
d
lex
ico
n
,
a
n
d
m
a
n
u
a
l
we
ig
h
ti
n
g
o
f
o
p
i
n
i
o
n
term
s.
Th
e
p
u
r
p
o
se
o
f
th
e
h
y
b
ri
d
lex
ico
n
a
n
d
t
h
e
m
a
n
u
a
l
we
ig
h
t
in
g
o
f
o
p
in
io
n
te
rm
s
is
to
u
p
d
a
te
th
e
e
x
isti
n
g
le
x
ico
n
a
n
d
so
lv
e
th
e
p
r
o
b
lem
o
f
we
ig
h
ti
n
g
wo
rd
s
a
n
d
p
h
ra
se
s
o
f
o
p
i
n
io
n
te
rm
s.
Th
e
e
v
a
lu
a
ti
o
n
re
su
lt
s
we
re
v
e
ry
g
o
o
d
,
wit
h
a
v
e
ra
g
e
F1
-
sc
o
re
s
o
f
9
3
.
8
4
%
f
o
r
a
sp
e
c
t
c
a
teg
o
riza
ti
o
n
a
n
d
9
2
.
4
2
%
fo
r
ABSA.
K
ey
w
o
r
d
s
:
Asp
ec
t
-
b
ased
s
en
tim
en
t a
n
aly
s
is
B
id
ir
ec
tio
n
al
en
co
d
er
r
ep
r
esen
tatio
n
s
f
r
o
m
tr
an
s
f
o
r
m
er
s
Ma
ch
in
e
lear
n
in
g
Natu
r
al
lan
g
u
ag
e
u
n
d
er
s
tan
d
in
g
Sem
an
tic
s
im
ilar
ity
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
iy
an
ar
to
Sar
n
o
Dep
ar
tm
en
t o
f
I
n
f
o
r
m
atics,
Facu
lty
o
f
I
n
tellig
en
t E
lectr
ical
a
n
d
I
n
f
o
r
m
atics T
ec
h
n
o
lo
g
y
,
I
n
s
titu
t T
ek
n
o
lo
g
i
Sep
u
lu
h
No
p
e
m
b
er
Su
r
ab
ay
a,
I
n
d
o
n
esia
E
m
ail:
r
iy
an
ar
to
@
its
.
ac
.
id
1.
I
NT
RO
D
UCT
I
O
N
Asp
ec
t
-
b
ased
s
en
tim
en
t
an
al
y
s
is
(
AB
S
A)
r
esear
ch
o
n
e
x
p
li
cit
s
en
ten
ce
ca
s
es
[
1
]
h
as
le
d
to
m
a
n
y
d
ev
elo
p
m
e
n
ts
in
ap
p
ly
in
g
d
if
f
er
en
t
m
eth
o
d
s
f
o
r
AB
SA,
wh
ich
wo
r
k
o
n
ex
p
licit
s
en
ten
ce
ca
s
es
an
d
im
p
licit
s
en
ten
ce
ca
s
es.
T
h
e
d
ev
elo
p
m
en
t
o
f
AB
SA
r
esear
ch
r
elate
d
to
im
p
licit
ca
s
es
h
as
r
ea
ch
ed
ex
tr
ac
tio
n
w
o
r
k
wh
er
e
asp
ec
ts
an
d
o
p
i
n
io
n
s
c
an
n
o
t
b
e
id
e
n
tifie
d
in
a
s
en
te
n
ce
.
Sev
er
al
r
esear
ch
er
s
h
av
e
ca
r
r
ied
o
u
t
r
esear
c
h
r
elate
d
to
im
p
licit
o
p
in
io
n
e
x
tr
ac
tio
n
.
Ho
wev
er
,
litt
le
wo
r
k
h
as
b
ee
n
d
o
n
e
r
elate
d
t
o
ex
tr
ac
tin
g
im
p
licit
o
p
in
io
n
s
.
T
h
e
im
p
licit
o
p
in
io
n
ex
tr
ac
ti
o
n
[
2
]
ca
n
id
e
n
tify
w
o
r
d
s
th
at
co
n
tain
e
x
p
r
ess
io
n
s
o
f
o
p
in
io
n
th
at
ca
n
n
o
t
b
e
r
ec
o
g
n
ized
as
o
p
i
n
io
n
ter
m
s
f
o
r
ex
p
licit
asp
ec
t
f
e
atu
r
es.
A
s
tu
d
y
u
s
in
g
a
b
ag
o
f
wo
r
d
s
[
3
]
wo
r
k
e
d
o
n
ca
s
es
o
f
im
p
licit
s
en
ten
ce
s
th
at
ca
n
n
o
t
b
e
id
en
tifie
d
as
asp
ec
t
f
ea
tu
r
es
o
r
r
ec
o
g
n
ized
b
y
o
p
in
i
o
n
f
ea
tu
r
es.
R
esear
ch
o
n
im
p
licit
o
p
in
io
n
ex
tr
ac
tio
n
b
y
id
e
n
tify
in
g
ex
p
li
cit
asp
ec
ts
[
4
]
h
as
also
b
ee
n
co
n
d
u
cte
d
u
s
in
g
th
e
co
-
o
cc
u
r
r
en
ce
ter
m
a
p
p
r
o
ac
h
to
id
en
tify
im
p
licit
asp
ec
ts
th
r
o
u
g
h
im
p
licit
o
p
in
io
n
s
.
An
o
th
er
im
p
licit
o
p
in
io
n
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.
14
,
No
.
6
,
Dec
em
b
e
r
20
24
:
6
7
1
1
-
6
7
2
2
6712
ex
tr
ac
tio
n
[
5
]
ca
n
ca
teg
o
r
ize
v
er
b
s
an
d
ad
v
e
r
b
s
o
f
n
o
u
n
s
as
im
p
licit
o
p
in
io
n
s
.
T
h
e
im
p
licit o
p
in
io
n
e
x
tr
ac
tio
n
b
ased
o
n
c
o
-
o
cc
u
r
r
en
ce
asp
ec
t
ter
m
s
in
a
co
r
p
u
s
[
6
]
h
as
b
ee
n
ca
r
r
ied
o
u
t
f
o
r
th
e
ca
s
e
o
f
co
m
p
o
u
n
d
s
en
ten
ce
s
.
T
h
en
,
im
p
licit
o
p
i
n
io
n
ex
t
r
ac
tio
n
in
co
m
p
o
u
n
d
an
d
co
m
p
l
ex
s
en
ten
ce
s
[
7
]
was
also
ca
r
r
ied
o
u
t.
Ho
wev
e
r
,
n
o
n
e
o
f
t
h
e
ab
o
v
e
-
m
en
tio
n
ed
s
tu
d
ies
lo
o
k
ed
at
th
e
ex
tr
ac
tio
n
o
f
d
if
f
e
r
en
t
ty
p
es
o
f
im
p
licit
o
p
in
io
n
s
en
ten
ce
s
in
d
etail.
I
m
p
licit
s
en
ten
ce
s
,
b
o
th
asp
ec
t
an
d
o
p
in
io
n
ca
s
es,
ca
n
also
o
cc
u
r
in
s
ev
er
al
d
if
f
er
e
n
t
ty
p
es
o
f
s
en
ten
ce
s
.
Sen
ten
ce
s
ca
n
b
e
d
iv
id
ed
in
to
f
o
u
r
ty
p
es b
ased
o
n
th
eir
s
tr
u
ctu
r
e:
s
im
p
le,
co
m
p
o
u
n
d
,
co
m
p
lex
,
a
n
d
co
m
p
o
u
n
d
-
c
o
m
p
lex
s
en
ten
ce
s
.
T
h
e
ty
p
e
o
f
s
en
ten
ce
will
ce
r
tain
ly
in
f
l
u
en
ce
t
h
e
ex
t
r
ac
tio
n
o
f
asp
ec
t
a
n
d
o
p
in
io
n
wo
r
d
s
.
T
h
is
is
b
ec
a
u
s
e
o
f
th
e
is
s
u
es
we
ex
p
lain
e
d
at
th
e
b
eg
in
n
in
g
o
f
th
is
s
ec
tio
n
.
T
h
er
ef
o
r
e,
it
is
n
ec
ess
ar
y
to
b
u
ild
a
m
eth
o
d
to
ex
tr
ac
t
asp
ec
t
an
d
o
p
i
n
io
n
wo
r
d
s
th
at
ar
e
ex
p
lici
t
an
d
im
p
licit
in
d
is
am
b
ig
u
o
u
s
an
d
am
b
ig
u
o
u
s
s
en
ten
ce
s
b
ased
o
n
th
e
t
y
p
e
o
f
s
en
ten
ce
s
tr
u
ctu
r
e.
Am
b
ig
u
o
u
s
s
en
ten
ce
f
ac
to
r
s
h
av
e
m
o
tiv
ated
s
tu
d
ies
r
elate
d
to
s
en
tim
en
t
an
al
y
s
is
in
th
e
ca
s
e
o
f
im
p
licit
o
p
in
io
n
s
en
ten
ce
s
.
A
m
b
ig
u
o
u
s
s
en
ten
ce
s
h
av
e
m
a
n
y
s
u
b
-
f
ac
t
o
r
s
th
at
af
f
ec
t
th
e
ac
cu
r
ac
y
v
alu
e
i
n
ex
tr
ac
tin
g
asp
ec
t
an
d
o
p
in
io
n
ter
m
s
f
r
o
m
a
s
en
ten
c
e
[
8
]
.
T
h
ese
s
u
b
-
f
ac
t
o
r
s
in
clu
d
e
n
e
g
atio
n
(
co
n
ju
n
ctio
n
an
aly
s
is
an
d
p
u
n
ctu
atio
n
m
a
r
k
s
)
,
co
n
tr
ast
tr
an
s
itio
n
,
an
d
in
ten
s
if
ier
s
.
Ho
wev
er
,
th
e
ex
p
licit
an
d
im
p
licit
ex
tr
ac
tio
n
o
f
asp
ec
ts
an
d
o
p
i
n
io
n
s
in
d
is
am
b
ig
u
o
u
s
an
d
a
m
b
ig
u
o
u
s
s
en
ten
ce
s
b
ased
o
n
t
h
e
ty
p
e
o
f
s
en
ten
ce
s
tr
u
ctu
r
e
h
as n
o
t
b
ee
n
s
tu
d
ied
in
d
ep
th
.
B
ased
o
n
th
ese
is
s
u
es,
th
is
p
ap
er
p
r
o
p
o
s
es
a
m
et
h
o
d
b
ase
d
o
n
a
co
m
p
r
eh
e
n
s
iv
e
u
n
d
er
s
tan
d
in
g
o
f
p
r
ev
io
u
s
s
tu
d
ies
o
n
im
p
licit
s
en
ten
ce
ca
s
es,
esp
ec
ially
im
p
l
icit
o
p
in
io
n
s
.
Fo
r
ex
am
p
le,
i
n
a
r
estau
r
an
t
r
ev
iew
ca
s
e,
“
Th
is
p
la
ce
is
th
e
mo
s
t
Ja
p
a
n
ese
it
ca
n
ev
er
g
et
.
”
T
h
e
wo
r
d
Ja
p
a
n
ese
ca
n
b
e
eith
e
r
an
ad
jectiv
e
o
r
a
n
o
u
n
.
Me
a
n
wh
ile,
s
u
p
p
o
s
e
th
e
s
en
ten
ce
is
ex
tr
ac
ted
u
s
in
g
p
ar
t
o
f
s
p
ee
ch
(
POS)
tag
g
in
g
[
9
]
;
th
e
r
esu
lt
is
th
at
Ja
p
a
n
ese
is
lab
eled
J
J
,
r
ep
r
ese
n
tin
g
an
a
d
jectiv
e.
T
h
e
tag
la
b
els
ar
e:
T
h
is
(
DT
)
p
lace
(
NN)
is
(
VB
Z
)
th
e
(
DT
)
m
o
s
t
(
R
B
S)
J
ap
an
ese
(
J
J
)
it
(
P
R
P)
ca
n
(
MD
)
ev
er
(
R
B
)
g
et
(
VB
)
.
(
.
)
.
I
n
p
r
e
v
io
u
s
s
tu
d
ies,
th
e
r
esu
lts
o
b
tain
ed
f
r
o
m
th
e
ex
tr
ac
tio
n
o
f
asp
ec
t
an
d
o
p
in
io
n
ter
m
s
wer
e
p
la
ce
an
d
Ja
p
a
n
ese
,
r
esp
ec
tiv
ely
,
with
“
R
E
STAU
R
ANT
”
as
th
e
asp
ec
t
ca
teg
o
r
y
o
b
tain
e
d
,
b
ec
au
s
e
th
ey
wer
e
n
o
t
a
b
le
to
e
x
t
r
ac
t
Ja
p
a
n
ese
with
“
FOOD
”
as
th
e
asp
ec
t
ca
teg
o
r
y
.
T
h
is
r
esear
ch
aim
s
to
ac
cu
r
ately
ex
tr
ac
t
p
air
s
o
f
asp
ec
ts
an
d
o
p
in
i
o
n
s
,
wh
ich
ar
e
ex
p
licit
an
d
im
p
licit,
b
ased
o
n
th
e
f
o
u
r
s
en
ten
ce
s
tr
u
ctu
r
es
s
im
p
le,
co
m
p
o
u
n
d
,
co
m
p
l
ex
,
an
d
co
m
p
o
u
n
d
-
co
m
p
lex
.
Mo
r
eo
v
e
r
,
th
e
r
es
u
lts
o
f
th
ese
p
air
s
will
b
e
p
r
o
ce
s
s
ed
to
d
eter
m
in
e
th
e
asp
ec
t
ca
teg
o
r
y
an
d
s
en
tim
en
t p
o
lar
ity
o
f
a
r
e
v
iew.
W
e
u
s
e
a
r
e
s
tau
r
an
t
r
ev
iew
d
ataset
th
at
i
s
well
-
an
n
o
tated
b
ec
au
s
e
it
co
n
tain
s
r
ev
iews
w
i
th
v
ar
io
u
s
ty
p
es
o
f
s
en
ten
ce
s
.
T
h
e
d
ataset
was
p
r
o
ce
s
s
ed
in
th
e
p
r
e
-
p
r
o
ce
s
s
in
g
s
tag
e.
As
in
p
u
t,
we
also
u
s
ed
asp
ec
t
ca
teg
o
r
y
k
ey
wo
r
d
s
tak
en
f
r
o
m
ar
ticles
o
n
W
ik
ip
ed
ia.
Fro
m
t
h
e
p
r
e
-
p
r
o
ce
s
s
in
g
r
esu
lts
,
we
ca
r
r
ied
o
u
t
t
h
e
tex
t
ex
tr
ac
tio
n
p
r
o
ce
s
s
.
W
e
b
u
ilt
a
s
et
o
f
r
u
les
b
ased
o
n
s
en
te
n
ce
-
wo
r
d
r
elatio
n
s
as
a
p
r
o
p
o
s
ed
tex
t
ex
tr
ac
tio
n
m
eth
o
d
to
g
et
asp
ec
ts
an
d
o
p
in
io
n
ter
m
s
th
at
ar
e
e
x
p
licit
an
d
im
p
licit.
I
n
th
e
tex
t
ex
t
r
ac
tio
n
,
o
u
r
m
eth
o
d
ex
tr
ac
ts
s
en
ten
ce
s
b
ased
o
n
th
eir
s
tr
u
ctu
r
e:
s
im
p
le,
co
m
p
o
u
n
d
,
c
o
m
p
lex
,
an
d
co
m
p
o
u
n
d
-
co
m
p
lex
s
en
te
n
ce
s
.
T
h
ese
s
en
ten
ce
s
ar
e
also
cla
s
s
if
ied
in
to
ex
p
licit
asp
ec
t
a
n
d
ex
p
licit
o
p
in
i
o
n
(
E
AE
O)
;
im
p
licit
asp
ec
t
an
d
ex
p
licit
o
p
in
io
n
(
I
AE
O)
;
ex
p
licit
asp
ec
t
an
d
im
p
licit
o
p
in
io
n
(
E
AI
O)
;
an
d
im
p
licit
asp
ec
t
an
d
im
p
licit
o
p
in
io
n
(
I
AI
O)
.
T
h
e
n
,
in
th
e
asp
ec
t
ca
teg
o
r
izatio
n
s
tag
e
,
th
e
asp
ec
t
ter
m
s
s
u
cc
ess
f
u
lly
o
b
tain
ed
u
s
in
g
b
id
ir
ec
tio
n
al
en
co
d
er
r
ep
r
ese
n
tatio
n
s
f
r
o
m
tr
an
s
f
o
r
m
er
s
(
B
E
R
T
)
em
b
ed
d
in
g
an
d
s
em
an
tic
s
im
ilar
ity
ar
e
p
r
o
ce
s
s
ed
to
o
b
tain
asp
ec
t c
at
eg
o
r
ies b
ased
o
n
f
o
u
r
class
es:
am
b
ian
ce
,
f
o
o
d
,
s
er
v
ice,
an
d
p
r
ice
.
Fu
r
th
er
m
o
r
e
,
in
th
e
AB
SA
s
tag
e,
th
e
v
alu
e
o
f
s
en
tim
en
t
p
o
lar
ity
is
m
ea
s
u
r
e
d
u
s
in
g
th
e
B
E
R
T
s
en
tim
en
t
an
aly
s
is
m
eth
o
d
b
y
im
p
r
o
v
i
n
g
th
e
le
x
ico
n
b
ased
o
n
th
e
o
p
in
io
n
lex
ico
n
an
d
t
h
e
im
p
licit
asp
ec
t
lex
ico
n
.
W
e
m
an
u
ally
ad
d
ed
a
d
d
itio
n
al
o
p
in
i
o
n
ter
m
weig
h
tin
g
to
ca
teg
o
r
ize
s
en
tim
en
t p
o
l
ar
ity
b
ased
o
n
f
o
u
r
class
es
:
v
er
y
p
o
s
itiv
e,
p
o
s
itiv
e,
n
eg
ativ
e,
an
d
v
er
y
n
e
g
ativ
e.
Fin
ally
,
we
co
n
d
u
cte
d
an
ev
alu
atio
n
s
tag
e
t
o
m
ea
s
u
r
e
th
e
s
u
cc
ess
o
f
th
e
p
r
o
p
o
s
ed
m
eth
o
d
f
o
r
asp
ec
t
ca
teg
o
r
izatio
n
an
d
AB
SA
b
y
m
ea
s
u
r
in
g
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all,
an
d
F1
-
s
co
r
e
.
2.
RE
L
AT
E
D
WO
RK
2
.
1
.
P
re
-
pro
ce
s
s
ing
T
h
e
p
r
e
-
p
r
o
ce
s
s
in
g
s
tag
e
aim
s
to
p
r
o
ce
s
s
r
aw
d
ata
in
to
d
a
ta
th
at
is
r
ea
d
y
t
o
b
e
p
r
o
ce
s
s
ed
.
I
n
th
is
s
tag
e,
to
k
en
izatio
n
is
ca
r
r
ied
o
u
t,
b
r
ea
k
i
n
g
d
o
wn
th
e
to
k
e
n
s
in
a
d
o
cu
m
en
t
to
b
e
tak
e
n
as
ter
m
s
.
T
h
en
,
th
e
s
to
p
wo
r
d
r
e
m
o
v
al
p
r
o
ce
s
s
is
ca
r
r
ied
o
u
t
b
y
r
em
o
v
in
g
u
n
n
e
ce
s
s
ar
y
ter
m
s
co
n
tain
ed
in
th
e
to
k
en
ized
r
esu
lts
.
Fin
ally
,
th
e
d
ata
p
r
o
ce
s
s
ed
b
y
s
to
p
wo
r
d
r
em
o
v
al
is
f
u
r
th
er
p
r
o
ce
s
s
ed
b
y
s
tem
m
in
g
,
wh
er
e
ea
ch
wo
r
d
is
r
ed
u
ce
d
t
o
its
r
o
o
t.
2
.
2
.
K
ey
wo
rds
f
o
r
a
s
pect
ca
t
e
g
o
r
ies
T
h
e
r
estau
r
an
t
asp
ec
t
ca
teg
o
r
i
es
wer
e
tak
en
f
r
o
m
th
e
d
ef
in
i
tio
n
s
an
d
v
ar
iab
les
o
f
asp
ec
t
ca
teg
o
r
ies
d
escr
ib
ed
in
p
r
ev
io
u
s
s
tu
d
ies
[
1
0
]
,
[
1
1
]
.
T
h
is
r
esear
ch
u
s
ed
f
o
u
r
r
estau
r
an
t
asp
ec
t
ca
teg
o
r
ies:
atm
o
s
p
h
er
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
A
s
p
ec
t
-
b
a
s
ed
s
en
timen
t a
n
a
l
ysis
:
n
a
tu
r
a
l la
n
g
u
a
g
e
u
n
d
ers
ta
n
d
in
g
fo
r
imp
licit
r
ev
iew
(
S
u
h
a
r
iya
n
to
)
6713
f
o
o
d
,
p
r
ice,
an
d
s
er
v
ice.
Me
a
n
wh
ile,
th
e
asp
ec
t
k
ey
wo
r
d
s
wer
e
tak
en
f
r
o
m
W
ik
ip
ed
ia
p
ag
es
r
elate
d
to
th
e
d
ef
in
itio
n
an
d
v
a
r
iab
les o
f
asp
ec
t c
ateg
o
r
ies.
2
.
3
.
T
ex
t
ex
t
ra
ct
i
o
n
I
n
s
en
tim
en
t
an
aly
s
is
,
tex
t
ex
t
r
ac
tio
n
aim
s
to
g
et
asp
ec
t
an
d
o
p
in
io
n
wo
r
d
s
.
T
ex
t
ex
tr
ac
tio
n
r
esear
ch
h
as
p
r
o
d
u
ce
d
m
a
n
y
wo
r
k
s
[
1
2
]
.
T
h
e
p
r
esen
t
r
esear
c
h
d
i
s
cu
s
s
es
p
r
o
b
lem
s
th
at
p
r
ev
io
u
s
s
tu
d
ies
h
av
e
n
o
t
wo
r
k
ed
o
u
t
i
n
d
e
p
th
.
W
e
d
is
cu
s
s
tex
t
ex
tr
ac
tio
n
b
ased
o
n
two
m
ain
task
s
:
s
en
ten
ce
ex
tr
ac
tio
n
an
d
wo
r
d
ex
tr
ac
tio
n
.
2
.
3
.
1
.
Sente
nce
ex
t
ra
ct
io
n
E
n
g
lis
h
s
en
ten
ce
ex
tr
ac
tio
n
is
v
er
y
clo
s
ely
r
elate
d
to
th
e
s
en
ten
ce
ty
p
e
in
a
r
ev
iew.
B
ased
o
n
th
eir
s
tr
u
ctu
r
e,
E
n
g
lis
h
s
en
ten
ce
s
co
n
s
is
t
o
f
f
o
u
r
class
es:
s
im
p
le,
co
m
p
o
u
n
d
,
co
m
p
lex
,
an
d
co
m
p
o
u
n
d
-
c
o
m
p
lex
s
en
ten
ce
s
[
1
3
]
.
An
in
-
d
ep
th
ap
p
r
o
ac
h
is
n
ee
d
ed
to
ex
tr
ac
t
th
e
tex
t
in
r
ev
iew
ca
s
es
alo
n
g
with
p
o
s
s
ib
le
s
en
ten
ce
s
tr
u
ctu
r
es.
An
alg
o
r
ith
m
m
ay
wo
r
k
well
o
n
s
im
p
le
s
en
ten
ce
ca
s
es
b
ec
au
s
e
th
ey
o
n
ly
co
n
tain
a
n
asp
ec
t
an
d
an
o
p
in
io
n
wo
r
d
.
I
t
is
d
if
f
er
en
t
f
o
r
ca
s
es
with
o
t
h
er
s
en
ten
ce
s
tr
u
ctu
r
es
th
at
m
ay
h
av
e
m
o
r
e
th
a
n
o
n
e
asp
ec
t
an
d
o
p
i
n
io
n
wo
r
d
.
Mo
r
eo
v
e
r
,
if
th
e
asp
ec
t
an
d
o
p
in
io
n
wo
r
d
s
ar
e
n
o
t
lo
ca
te
d
clo
s
e
to
g
eth
er
i
n
a
s
en
ten
ce
,
th
en
er
r
o
r
s
in
ch
o
o
s
in
g
th
e
asp
ec
t
an
d
o
p
in
io
n
w
o
r
d
s
will
m
o
s
t
lik
ely
o
cc
u
r
w
h
en
u
s
in
g
a
s
im
p
le
alg
o
r
ith
m
.
T
h
e
f
o
llo
win
g
ar
e
s
o
m
e
illu
s
tr
atio
n
s
o
f
s
en
ten
ce
e
x
tr
ac
tio
n
f
r
o
m
r
e
v
iew
ca
s
es.
−
A
s
im
p
le
s
en
ten
ce
is
a
s
en
te
n
ce
th
at
o
n
ly
c
o
n
s
is
ts
o
f
o
n
e
s
u
b
ject
an
d
o
n
e
p
r
e
d
icate
.
Fo
r
ex
am
p
le,
Th
e
p
erfect
s
p
o
t.
−
A
co
m
p
o
u
n
d
s
en
ten
ce
c
o
n
s
is
ts
o
f
two
o
r
m
o
r
e
s
u
b
jects
an
d
p
r
ed
icate
s
co
n
n
ec
ted
b
y
a
c
o
n
ju
n
ctio
n
.
T
h
is
co
n
ju
n
ctio
n
c
o
n
s
is
ts
o
f
a
n
d
,
o
r
,
b
u
t,
eith
er,
an
d
n
eith
er
.
Fo
r
ex
am
p
le,
Th
e
fo
o
d
w
a
s
w
ell
p
r
ep
a
r
ed
a
n
d
t
h
e
s
ervice
w
a
s
imp
ec
ca
b
le.
−
A
co
m
p
lex
s
en
ten
ce
co
n
s
is
ts
o
f
two
o
r
m
o
r
e
s
u
b
jects
an
d
p
r
ed
icate
s
with
a
m
ain
an
d
a
s
u
b
o
r
d
in
ate
clau
s
e
co
n
n
ec
ted
b
y
a
s
u
b
o
r
d
in
atio
n
.
T
h
is
s
u
b
o
r
d
i
n
atio
n
ex
te
n
d
s
th
e
r
ea
ch
o
f
th
e
m
ai
n
clau
s
e
b
y
ad
d
in
g
a
s
tatem
en
t
o
f
tim
e,
p
lace
,
p
er
s
o
n
,
th
in
g
,
o
r
r
ea
s
o
n
.
Fo
r
ex
am
p
le,
My
fr
ien
d
g
o
t
th
e
mu
s
h
r
o
o
m
p
iz
z
a
,
w
h
ich
ta
s
ted
b
etter.
−
A
co
m
p
o
u
n
d
-
co
m
p
lex
s
en
ten
ce
co
n
s
is
ts
o
f
th
r
ee
o
r
m
o
r
e
s
u
b
jects
an
d
p
r
ed
icate
s
.
T
h
is
ty
p
e
o
f
s
en
ten
ce
co
n
s
is
ts
o
f
a
co
n
ju
n
ctio
n
an
d
a
s
u
b
o
r
d
in
atio
n
in
th
e
s
en
ten
ce
.
Fo
r
ex
am
p
le,
Th
e
a
mb
ien
ce
is
p
r
etty
a
n
d
n
ice
fo
r
co
n
ve
r
s
a
tio
n
,
s
o
a
ca
s
u
a
l lu
n
c
h
h
ere
w
o
u
ld
p
r
o
b
a
b
l
y
b
e
b
est.
2
.
3
.
2
.
Wo
rd
ex
t
ra
c
t
io
n
I
n
E
n
g
lis
h
,
wo
r
d
ex
tr
ac
tio
n
i
s
v
er
y
clo
s
ely
r
elate
d
to
th
e
task
o
f
r
etr
iev
in
g
t
h
e
tar
g
et
wo
r
d
s
f
o
r
f
u
r
th
er
p
r
o
ce
s
s
in
g
.
Ou
r
m
et
h
o
d
e
x
tr
ac
ts
wo
r
d
s
as
asp
ec
t
a
n
d
o
p
in
io
n
ter
m
s
b
ased
o
n
POS
tag
g
in
g
with
3
6
lab
els.
T
h
en
,
it
d
iv
id
es
th
e
r
e
v
iews
in
to
f
o
u
r
class
es,
n
am
e
ly
:
E
AE
O,
I
AE
O,
E
AI
O,
an
d
I
AI
O.
E
x
tr
ac
tin
g
ex
p
licit
asp
ec
t
an
d
o
p
in
io
n
ter
m
s
ca
n
b
e
ea
s
ily
d
o
n
e
w
h
en
t
h
er
e
is
NN
as
th
e
asp
ec
t
ter
m
lab
el
an
d
J
J
as
th
e
o
p
in
io
n
ter
m
lab
el
in
t
h
e
s
en
te
n
ce
.
Ho
wev
e
r
,
if
a
s
en
ten
ce
is
im
p
licit
with
o
u
t
NN
o
r
J
J
,
it
will
p
o
s
e
a
s
p
ec
i
f
ic
p
r
o
b
lem
in
th
e
te
x
t
ex
tr
ac
tio
n
.
T
h
is
d
i
v
is
io
n
o
f
wo
r
d
class
es
aim
s
to
id
en
tify
an
d
co
r
r
ec
tly
ex
tr
ac
t
t
h
e
asp
ec
t
an
d
o
p
in
io
n
ter
m
s
t
h
at
ar
e
e
x
p
licit
an
d
im
p
licit.
T
a
b
le
1
illu
s
tr
ates
tex
t
ex
tr
ac
tio
n
f
o
r
s
en
ten
ce
an
d
wo
r
d
ex
tr
ac
tio
n
ca
s
es.
T
h
e
wo
r
d
ex
t
r
ac
tio
n
co
n
ce
p
t im
p
lem
e
n
ted
i
s
b
ased
o
n
th
e
f
o
llo
win
g
f
o
u
r
ca
s
es:
−
I
n
th
e
E
AE
O
ca
s
e,
asp
ec
t
a
n
d
o
p
in
io
n
wo
r
d
s
ar
e
ex
tr
ac
ted
b
y
id
en
tif
y
in
g
s
en
te
n
ce
s
co
n
ta
in
in
g
NN
as
th
e
asp
ec
t te
r
m
an
d
J
J
as th
e
o
p
in
i
o
n
ter
m
.
−
I
n
th
e
I
AE
O
ca
s
e,
asp
ec
t
an
d
o
p
in
io
n
wo
r
d
s
ar
e
e
x
tr
ac
ted
b
y
id
en
tif
y
in
g
s
en
ten
ce
s
th
at
d
o
n
o
t
c
o
n
tain
NN
as th
e
asp
ec
t te
r
m
b
u
t c
o
n
tain
J
J
as th
e
o
p
in
io
n
ter
m
.
−
I
n
th
e
E
AI
O
ca
s
e,
asp
ec
t
an
d
o
p
in
io
n
wo
r
d
s
ar
e
ex
tr
ac
ted
b
y
id
en
tify
in
g
s
en
ten
ce
s
co
n
tai
n
in
g
NN
as
th
e
asp
ec
t te
r
m
b
u
t n
o
t JJ
as th
e
o
p
in
io
n
ter
m
.
−
I
n
th
e
I
AI
O
ca
s
e,
asp
ec
t
an
d
o
p
in
io
n
wo
r
d
s
ar
e
ex
tr
ac
ted
b
y
id
en
tify
in
g
s
en
ten
ce
s
th
at
d
o
n
o
t
co
n
tain
NN
as th
e
asp
ec
t te
r
m
an
d
d
o
n
o
t
co
n
tain
J
J
as th
e
o
p
in
io
n
ter
m
.
T
ab
le
1
.
T
h
e
p
r
o
p
o
s
ed
d
ataset
r
ep
r
esen
tatio
n
I
D
R
e
v
i
e
w
s
I
D
S
e
n
t
e
n
c
e
s
R
e
v
i
e
w
s
1
0
0
1
0
th
e
wa
it
ti
m
e
h
e
re
fo
r
th
e
f
o
o
d
e
v
e
n
m
o
re
m
a
k
e
s m
y
sto
m
a
c
h
v
e
r
y
ru
m
b
li
n
g
1
0
0
2
0
th
e
p
o
rti
o
n
h
e
re
fo
r
t
h
e
fo
o
d
stil
l
m
a
k
e
s m
y
sto
m
a
c
h
v
e
ry
r
u
m
b
li
n
g
2
.
4
.
Sy
no
ny
m
-
ba
s
ed
t
er
m
ex
pa
ns
io
n
Sy
n
o
n
y
m
-
b
ased
ter
m
ex
p
a
n
s
io
n
ex
p
an
d
s
n
o
u
n
s
an
d
ad
jec
tiv
es
d
en
o
tin
g
th
e
asp
ec
t
an
d
o
p
in
io
n
wo
r
d
s
.
T
h
is
ex
p
an
s
io
n
p
r
o
ce
s
s
i
s
ca
r
r
ied
o
u
t
u
s
in
g
W
o
r
d
Net
[
1
4
]
.
T
h
e
n
,
th
e
ex
p
a
n
s
io
n
r
esu
lt
will
b
e
ex
tr
ac
ted
f
r
o
m
wo
r
d
s
y
n
o
n
y
m
s
with
th
e
h
ig
h
est s
im
ilar
ity
v
alu
e
th
r
o
u
g
h
th
e
asp
ec
t c
ateg
o
r
y
v
ar
ia
b
les.
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.
14
,
No
.
6
,
Dec
em
b
e
r
20
24
:
6
7
1
1
-
6
7
2
2
6714
2
.
5
.
Def
ini
t
io
n
-
ba
s
ed
t
er
m
ex
pa
ns
io
n
Def
in
itio
n
-
b
ased
ter
m
ex
p
a
n
s
io
n
is
u
s
ed
to
o
b
tain
h
ig
h
er
ac
cu
r
ac
y
o
f
asp
ec
t a
n
d
o
p
in
i
o
n
wo
r
d
p
air
s
.
I
f
s
y
n
o
n
y
m
-
b
ased
ex
p
a
n
s
io
n
is
u
s
ed
to
ex
p
an
d
ex
p
licit
n
o
u
n
s
an
d
a
d
jectiv
es,
th
en
we
a
s
s
u
m
e
th
at
ex
p
licit
n
o
u
n
s
a
n
d
ad
jectiv
es
ca
n
also
h
av
e
im
p
licit
m
ea
n
i
n
g
s
.
Fo
r
e
x
am
p
le,
th
e
w
a
itin
g
time
h
ere
fo
r
th
e
fo
o
d
ma
ke
s
my
s
to
ma
ch
feel
h
u
n
g
r
y
ev
en
mo
r
e
,
wh
er
e
th
e
r
esu
lts
o
f
th
e
n
o
u
n
a
n
d
ad
jectiv
e
p
air
s
s
to
ma
ch
an
d
h
u
n
g
r
y
h
av
e
an
im
p
licit
m
ea
n
in
g
r
elat
ed
to
t
h
e
asp
ec
t
ca
te
g
o
r
y
v
a
r
iab
le.
T
h
e
r
ef
o
r
e
,
f
ir
s
t,
f
r
o
m
th
e
r
esu
lts
o
f
p
r
ev
io
u
s
s
tu
d
ies
[
1
5
]
,
we
c
o
m
p
iled
a
lis
t
o
f
ex
p
licit
o
p
in
i
o
n
te
r
m
s
f
r
o
m
e
x
p
licit
asp
ec
ts
an
d
ea
ch
ca
te
g
o
r
y
.
Fu
r
th
er
m
o
r
e
,
s
u
p
p
o
s
e
th
e
o
p
i
n
io
n
ter
m
s
in
th
e
r
e
v
iew
ar
e
n
o
t
p
r
esen
t
in
th
e
lis
t
o
f
ev
er
y
asp
ec
t
ca
teg
o
r
y
th
at
h
as
b
ee
n
c
o
m
p
iled
;
in
th
at
ca
s
e,
we
will
m
ar
k
th
e
o
p
i
n
io
n
te
r
m
as
im
p
licit.
Fo
r
e
x
am
p
le,
t
h
e
wo
r
d
h
u
n
g
r
y
is
n
o
t
lis
ted
in
th
e
p
air
o
f
e
x
p
li
cit
asp
ec
t
ter
m
s
w
a
itin
g
time
an
d
fo
o
d
,
s
o
th
e
wo
r
d
h
u
n
g
r
y
is
lab
eled
as
an
im
p
licit o
p
in
io
n
te
r
m
.
Def
in
itio
n
-
b
ased
ex
p
an
s
io
n
i
s
b
ased
o
n
n
o
u
n
s
an
d
v
er
b
s
with
a
n
o
u
n
-
s
u
b
ject
r
elatio
n
s
h
ip
with
o
p
in
io
n
ter
m
s
lab
eled
as
im
p
licit.
T
h
en
,
s
im
p
le
E
n
g
lis
h
W
ik
tio
n
ar
y
ex
p
an
d
s
n
o
u
n
s
a
n
d
v
er
b
s
th
at
d
e
n
o
te
ter
m
s
o
f
im
p
licit
asp
ec
t
an
d
i
m
p
licit
o
p
in
io
n
.
M
o
r
eo
v
e
r
,
B
E
R
T
em
b
ed
d
in
g
is
u
s
ed
to
g
e
t
th
e
ter
m
s
o
f
ea
c
h
s
en
ten
ce
in
th
e
r
elate
d
W
ik
tio
n
ar
y
lin
k
.
Fin
ally
,
th
e
ex
p
a
n
d
i
n
g
ter
m
s
with
th
e
h
ig
h
est
s
im
ilar
ity
v
alu
e
to
th
e
asp
ec
t te
r
m
s
in
th
e
r
ev
iew
ar
e
tak
en
as h
id
d
e
n
ter
m
s
.
2
.
6
.
Na
t
ura
l la
ng
ua
g
e
un
dersta
nd
ing
Mo
s
t
n
atu
r
al
lan
g
u
ag
e
u
n
d
e
r
s
tan
d
in
g
(
NL
U)
u
s
e
n
e
u
r
al
n
etwo
r
k
m
o
d
els
th
at
s
tar
t
b
y
b
r
ea
k
i
n
g
d
o
wn
a
s
eq
u
en
ce
o
f
s
tr
in
g
s
in
to
in
d
iv
id
u
al
wo
r
d
s
,
p
h
r
ases
,
o
r
en
t
ir
e
s
en
ten
ce
s
,
a
p
r
o
ce
s
s
k
n
o
w
n
as
to
k
en
izatio
n
.
T
h
e
to
k
en
izer
b
u
ild
s
v
o
ca
b
u
lar
y
an
d
co
n
v
er
ts
wo
r
d
s
eq
u
en
ce
s
in
to
in
teg
er
s
eq
u
en
ce
s
.
E
ac
h
in
teg
er
is
m
ap
p
ed
to
a
v
alu
e
in
t
h
e
d
icti
o
n
ar
y
th
at
e
n
co
d
es
th
e
e
n
tire
co
r
p
u
s
,
with
th
e
k
ey
s
in
th
e
d
ictio
n
ar
y
b
ein
g
th
e
v
o
ca
b
u
lar
y
te
r
m
its
elf
.
Similar
ly
,
th
e
B
E
R
T
m
o
d
el
p
r
o
ce
s
s
es
in
p
u
t
to
k
e
n
s
in
th
r
ee
s
tep
s
,
w
h
er
e
ea
c
h
to
k
en
is
th
e
s
u
m
o
f
t
h
r
ee
em
b
e
d
d
in
g
s
: T
o
k
en
em
b
ed
d
i
n
g
,
Seg
m
e
n
t e
m
b
ed
d
in
g
,
an
d
Po
s
itio
n
em
b
e
d
d
in
g
.
2
.
7
.
T
er
m
f
re
qu
ency
-
inv
er
s
e
co
rpus
f
re
qu
e
ncy
T
er
m
f
r
eq
u
en
cy
-
in
v
er
s
e
co
r
p
u
s
f
r
eq
u
en
c
y
(
T
F
-
I
C
F)
[
1
6
]
weig
h
ts
ter
m
s
b
ased
o
n
in
f
o
r
m
atio
n
f
r
o
m
d
o
cu
m
e
n
t
d
in
clu
s
ter
c
.
First,
th
e
f
r
eq
u
en
cy
o
f
ter
m
t
i
n
cl
u
s
ter
c
is
d
e
n
o
ted
b
y
.
T
h
e
n
,
th
e
n
u
m
b
er
o
f
clu
s
ter
s
th
at
ex
is
t
c
an
d
th
e
n
u
m
b
er
o
f
clu
s
ter
s
co
n
tain
in
g
th
e
ter
m
t
,
,
ar
e
u
s
ed
to
ca
lcu
late
th
e
in
v
er
s
e
clu
s
ter
f
r
eq
u
en
cy
t
d
en
o
ted
b
y
u
s
in
g
th
e
eq
u
atio
n
.
Fin
ally
,
th
e
T
F
-
I
C
F
weig
h
tin
g
is
ca
lcu
lated
u
s
in
g
(
1
)
.
−
=
(
1
)
2
.
8
.
Sem
a
ntic
s
im
ila
rit
y
Sem
an
tic
s
im
ilar
ity
[
1
7
]
u
s
es
co
s
in
e
s
im
ilar
ity
[
1
8
]
to
m
ea
s
u
r
e
th
e
s
im
ilar
ity
b
etwe
en
a
s
p
ec
t
ter
m
an
d
asp
ec
t
v
ar
iab
le
.
T
h
e
co
s
in
e
s
im
ilar
ity
eq
u
atio
n
is
s
h
o
wn
in
(
2
)
.
is
a
v
ec
to
r
m
em
b
e
r
f
r
o
m
an
d
is
a
v
ec
to
r
m
em
b
e
r
f
r
o
m
.
(
,
)
=
∑
=
1
√
∑
(
)
2
=
1
√
∑
(
)
2
=
1
(
2
)
2
.
9
.
B
E
R
T
B
E
R
T
[
1
9
]
is
u
s
ed
to
d
er
iv
e
wo
r
d
co
n
tex
t
b
ased
o
n
th
e
s
e
m
an
tic
co
n
ce
p
t
o
f
a
s
en
ten
ce
.
W
e
d
iv
id
e
th
e
co
o
r
d
in
ates
o
f
th
e
C
ar
tesi
an
d
iag
r
am
in
to
f
o
u
r
g
r
o
u
p
s
:
v
er
y
p
o
s
itiv
e,
p
o
s
itiv
e,
n
eg
ativ
e,
a
n
d
v
er
y
n
eg
ativ
e
.
T
h
e
x
-
ax
is
in
d
icate
s
th
e
s
tr
en
g
th
o
f
th
e
s
en
tim
en
t
v
alu
e,
wh
eth
er
it
is
s
tr
o
n
g
o
r
n
o
r
m
al,
a
n
d
t
h
e
y
-
ax
is
in
d
icate
s
th
e
p
o
la
r
ity
v
a
lu
e
o
f
th
e
s
en
tim
e
n
t,
wh
eth
er
i
t is p
o
s
itiv
e
o
r
n
eg
ativ
e.
2
.
1
0
.
H
y
brid
l
ex
ico
n
A
h
y
b
r
id
lex
ic
o
n
is
th
e
co
m
b
i
n
atio
n
o
f
a
n
o
p
in
io
n
lex
ico
n
[
1
1
]
a
n
d
an
im
p
licit
o
p
in
i
o
n
le
x
ico
n
[
2
0
]
.
First,
th
e
o
p
in
i
o
n
le
x
ico
n
is
u
s
ed
to
g
et
th
e
o
p
in
io
n
ter
m
s
c
o
n
tain
ed
in
th
e
r
ev
iew.
T
h
e
n
,
if
n
o
o
p
i
n
io
n
ter
m
s
ar
e
o
b
tain
e
d
,
ca
n
d
id
ate
o
p
in
i
o
n
ter
m
s
ar
e
d
etec
ted
u
s
in
g
th
e
im
p
licit
o
p
in
io
n
lex
ico
n
.
T
h
e
ex
p
an
s
io
n
o
f
ca
n
d
id
ate
o
p
in
io
n
ter
m
s
is
b
ased
o
n
t
h
e
ad
v
e
r
b
d
e
g
r
ee
r
elati
o
n
s
h
ip
.
2
.
1
1
.
O
pin
io
n
t
er
m
deg
re
e
weig
hting
Op
in
io
n
te
r
m
d
eg
r
ee
weig
h
ti
n
g
aim
s
t
o
u
p
d
ate
th
e
ex
is
tin
g
s
en
tim
en
t
p
o
lar
ity
an
d
s
en
tim
en
t
s
tr
en
g
th
.
Pre
v
io
u
s
r
esear
ch
[
2
1
]
o
b
tain
ed
s
en
tim
en
t
p
o
lar
i
ty
an
d
s
tr
en
g
th
f
r
o
m
o
p
in
io
n
ter
m
s
.
Ho
wev
er
,
th
e
o
p
in
io
n
ter
m
s
ca
n
n
o
t b
e
p
air
ed
with
ex
is
tin
g
asp
ec
t te
r
m
s
.
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
A
s
p
ec
t
-
b
a
s
ed
s
en
timen
t a
n
a
l
ysis
:
n
a
tu
r
a
l la
n
g
u
a
g
e
u
n
d
ers
ta
n
d
in
g
fo
r
imp
licit
r
ev
iew
(
S
u
h
a
r
iya
n
to
)
6715
2.
M
E
T
H
O
D
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
d
iag
r
a
m
is
s
h
o
wn
in
Fig
u
r
e
1
.
Fig
u
r
e
1
s
h
o
ws
th
at
th
is
r
esear
ch
h
as
s
ev
er
al
s
tag
es,
in
clu
d
in
g
p
r
e
-
p
r
o
ce
s
s
in
g
,
POS
tag
g
in
g
,
en
h
a
n
ce
d
++
d
ep
en
d
en
cy
p
ar
s
in
g
,
tex
t
ex
tr
ac
tio
n
,
asp
ec
t
ca
teg
o
r
izatio
n
,
an
d
asp
ec
t
-
b
a
s
ed
s
en
tim
en
t
an
aly
s
is
.
T
h
is
r
esear
ch
u
s
es
asp
ec
t
ca
teg
o
r
y
k
e
y
wo
r
d
s
a
n
d
a
h
y
b
r
id
lex
ico
n
as th
e
co
r
p
u
s
to
ex
tr
ac
t a
s
p
ec
t a
n
d
o
p
i
n
io
n
te
r
m
s
.
Fig
u
r
e
1
.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
d
iag
r
am
3
.
1
.
Da
t
a
s
et
T
h
e
d
ataset
r
ep
r
esen
tatio
n
u
s
ed
in
th
is
s
tu
d
y
was
tak
en
f
r
o
m
th
e
Po
n
tik
i
d
ataset
[
2
2
]
with
1
,
6
5
4
tr
ain
in
g
d
ata
an
d
8
4
5
test
in
g
d
ata.
W
e
al
s
o
m
an
u
ally
ad
d
ed
s
o
m
e
im
p
licit
s
en
ten
ce
s
f
o
r
th
e
d
ataset
to
im
p
r
o
v
e
th
e
ac
cu
r
ac
y
o
f
th
e
m
ea
s
u
r
em
en
t.
T
h
e
im
p
licit sen
ten
ce
d
ataset
co
n
s
is
t
s
o
f
8
0
0
tr
ain
in
g
d
ata
an
d
2
0
0
test
in
g
d
ata.
T
h
e
to
tal
p
r
o
p
o
s
ed
d
ataset
is
3
,
4
9
9
s
en
ten
ce
s
.
T
h
e
p
r
o
p
o
s
ed
d
ataset
r
ep
r
ese
n
tatio
n
is
s
h
o
wn
in
T
ab
le
1
.
3
.
2
.
P
re
-
pro
ce
s
s
ing
W
e
b
u
ilt
th
e
p
r
o
p
o
s
ed
p
r
e
-
p
r
o
ce
s
s
in
g
m
eth
o
d
co
n
s
is
ts
co
n
v
er
t
in
to
lo
wer
ca
s
e,
s
to
p
wo
r
d
r
em
o
v
al,
an
d
p
u
n
ct
u
atio
n
r
em
o
v
al.
So
m
e
s
y
m
b
o
ls
in
th
e
p
u
n
ctu
atio
n
r
em
o
v
al
p
r
o
ce
s
s
ar
e
n
o
t
r
em
o
v
ed
b
ec
a
u
s
e
th
ey
ar
e
p
ar
t
o
f
wr
itin
g
in
s
en
te
n
ce
ty
p
es
b
ased
o
n
th
eir
s
tr
u
ctu
r
e.
T
h
ese
s
y
m
b
o
ls
co
n
s
i
s
t
o
f
co
m
m
as
(
,
)
,
s
em
ico
lo
n
s
(
;)
,
co
lo
n
s
(
:)
,
s
in
g
le
q
u
o
tes
(
‘
)
,
d
o
u
b
le
q
u
o
tes
(
”)
,
e
x
clam
atio
n
m
ar
k
(
!
)
,
q
u
e
s
tio
n
m
ar
k
(
?)
,
a
n
d
co
n
n
ec
t (
-
)
.
T
a
b
le
2
s
h
o
ws th
e
illu
s
tr
atio
n
o
f
p
r
e
-
p
r
o
ce
s
s
in
g
r
esu
lts
.
T
ab
le
2
.
Pre
-
p
r
o
ce
s
s
in
g
r
esu
lt
s
I
D
R
e
v
i
e
w
s
I
D
S
e
n
t
e
n
c
e
s
S
e
n
t
e
n
c
e
R
e
s
u
l
t
s
1
0
0
1
0
Th
e
w
a
i
t
t
i
m
e
h
e
r
e
f
o
r
t
h
e
f
o
o
d
e
v
e
n
mo
r
e
mak
e
s m
y
s
t
o
m
a
c
h
v
e
r
y
r
u
mb
l
i
n
g
.
t
h
e
w
a
i
t
t
i
me
h
e
r
e
f
o
r
t
h
e
f
o
o
d
e
v
e
n
mo
r
e
m
a
k
e
s m
y
st
o
ma
c
h
v
e
r
y
r
u
m
b
l
i
n
g
1
0
0
2
0
Th
e
p
o
r
t
i
o
n
h
e
r
e
f
o
r
t
h
e
f
o
o
d
st
i
l
l
ma
k
e
s
my
st
o
m
a
c
h
v
e
r
y
r
u
mb
l
i
n
g
.
t
h
e
p
o
r
t
i
o
n
h
e
r
e
f
o
r
t
h
e
f
o
o
d
st
i
l
l
m
a
k
e
s
my
s
t
o
m
a
c
h
v
e
r
y
r
u
mb
l
i
n
g
3
.
3
.
K
ey
wo
rd
ex
t
ra
c
t
io
n
T
h
is
r
esear
ch
u
s
es
s
o
u
r
ce
s
f
r
o
m
W
ik
ip
ed
ia
[
1
3
]
f
o
r
asp
ec
t
ca
teg
o
r
y
k
e
y
wo
r
d
ex
tr
ac
tio
n
.
T
h
e
k
ey
wo
r
d
s
ar
e
u
s
ed
as
th
e
in
p
u
t
v
ar
iab
le
f
o
r
f
o
u
r
asp
ec
t
ca
t
eg
o
r
ies:
am
b
ian
ce
,
f
o
o
d
,
s
er
v
ice,
an
d
p
r
ice
.
T
h
e
o
b
tain
ed
k
ey
wo
r
d
v
ar
iab
les
ar
e
ex
tr
ac
ted
b
ased
o
n
th
e
n
u
m
b
er
o
f
ter
m
clu
s
ter
s
g
e
n
er
ated
f
r
o
m
W
ik
ip
ed
ia
f
o
r
th
e
f
o
u
r
asp
ec
t c
ateg
o
r
ies.
3
.
4
.
P
O
S t
a
g
g
ing
a
nd
enha
nced+
+
depen
dency
T
h
is
s
tag
e
p
r
o
p
o
s
es
p
r
e
p
ar
in
g
th
e
d
ata
f
o
r
ea
ch
wo
r
d
ta
g
an
d
ea
c
h
wo
r
d
r
elatio
n
as
in
p
u
t
to
d
eter
m
in
e
th
e
ca
n
d
id
ate
ter
m
s
.
T
h
e
POS
tag
g
in
g
is
ca
r
r
ie
d
o
u
t
u
s
in
g
th
e
t
o
o
l
f
r
o
m
Stan
f
o
r
d
C
o
r
e
NL
P
[
9
]
t
o
ex
tr
ac
t
th
e
wo
r
d
ty
p
e
o
f
ea
ch
ter
m
.
T
h
e
n
,
th
e
en
h
a
n
ce
d
++
d
ep
en
d
en
c
y
p
a
r
s
e
[
9
]
is
ca
r
r
ied
o
u
t
to
ex
tr
ac
t
ter
m
r
elatio
n
s
in
th
e
r
ev
iew.
T
a
b
le
3
s
h
o
ws
th
e
r
esu
lts
o
f
POS
tag
g
in
g
a
n
d
en
h
an
ce
d
++
d
ep
e
n
d
en
cy
f
o
r
R
ev
iew
[
1
0
0
1
]
.
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.
14
,
No
.
6
,
Dec
em
b
e
r
20
24
:
6
7
1
1
-
6
7
2
2
6716
T
ab
le
3
.
R
esu
lts
o
f
POS tag
g
in
g
an
d
en
h
a
n
ce
d
++
d
ep
en
d
en
cy
I
D
R
e
v
i
e
w
R
e
v
i
e
w
s
P
O
S
t
a
g
g
i
n
g
En
h
a
n
c
e
d
+
+
d
e
p
e
n
d
e
n
c
y
1
0
0
1
Th
e
w
a
i
t
t
i
m
e
h
e
r
e
f
o
r
t
h
e
f
o
o
d
e
v
e
n
mo
r
e
m
a
k
e
s m
y
st
o
m
a
c
h
v
e
r
y
r
u
mb
l
i
n
g
t
h
e
<
D
T
>
w
a
i
t
<
N
N
>
t
i
m
e
<
N
N
>
h
e
r
e
<
RB>
f
o
r <I
N
>
t
h
e
<
D
T
>
f
o
o
d
<
N
N
>
e
v
e
n
<
RB>
m
o
r
e
<
RB
R>
m
a
k
e
s
<
VBZ>
m
y
<
P
RP
$
>
st
o
m
a
c
h
<
N
N
>
v
e
r
y
<
R
B> r
u
m
b
l
i
n
g
<
VBG
>
d
e
t
(
t
i
m
e
,
t
h
e
)
,
c
o
m
p
o
u
n
d
(
t
i
m
e
,
w
a
i
t
)
,
a
d
v
m
o
d
(
t
i
m
e
,
h
e
r
e
)
,
n
m
o
d
:
f
o
r(
t
i
m
e
,
f
o
o
d
)
,
n
s
u
b
j
(
m
a
k
e
s,
t
i
m
e
)
,
c
a
s
e
(
f
o
o
d
,
f
o
r)
,
d
e
t
(
f
o
o
d
,
t
h
e
)
,
a
d
v
m
o
d
(
m
o
re
,
e
v
e
n
)
,
d
e
p
(
m
a
k
e
s,
ru
m
b
l
i
n
g
)
,
n
m
o
d
:
p
a
ss(
st
o
m
a
c
h
,
m
y
)
,
n
su
b
j
(
ru
m
b
l
i
n
g
,
st
o
m
a
c
h
)
,
a
d
v
m
o
d
(
ru
m
b
l
i
n
g
,
v
e
r
y
)
3
.
5
.
T
ex
t
e
x
t
ra
ct
i
o
n
T
h
e
p
r
o
p
o
s
ed
tex
t
ex
tr
ac
tio
n
i
s
co
m
p
iled
b
ased
o
n
f
o
u
r
d
ef
i
n
itio
n
s
:
E
AE
O,
I
AE
O,
E
AI
O,
an
d
I
AI
O.
T
er
m
s
lab
eled
as
NN
th
at
ex
i
s
t
in
th
e
asp
ec
t
k
ey
wo
r
d
v
ar
ia
b
le
ar
e
id
en
tifie
d
as
ex
p
licit
a
s
p
ec
t
ter
m
s
.
T
h
en
,
ter
m
s
lab
eled
as
J
J
th
at
ex
is
t
i
n
th
e
o
p
i
n
io
n
lex
ico
n
ar
e
id
en
t
if
ied
as
ex
p
licit
o
p
i
n
io
n
ter
m
s
.
M
o
r
e
o
v
e
r
,
w
e
a
l
s
o
a
d
d
e
d
a
p
h
r
a
s
e
e
x
t
r
a
c
ti
o
n
.
3
.
5
.
1
.
E
AE
O
ex
t
ra
ct
io
n
First,
th
e
ca
n
d
id
ate
r
esu
lts
o
f
asp
ec
t
an
d
o
p
in
io
n
ter
m
s
ar
e
p
r
o
ce
s
s
ed
to
o
b
tain
t
h
e
wo
r
d
s
th
at
in
d
icate
o
p
in
i
o
n
ter
m
s
in
th
e
o
p
in
io
n
ter
m
e
x
tr
ac
tio
n
.
T
h
e
asp
ec
t
ter
m
ex
tr
ac
tio
n
aim
s
to
g
et
th
e
wo
r
d
s
th
at
in
d
icate
asp
ec
t
ter
m
s
.
T
h
e
asp
ec
t
ter
m
ex
tr
ac
tio
n
also
p
r
o
d
u
ce
s
an
en
tity
c
o
r
p
u
s
th
at
will
b
e
u
s
ed
as
i
n
p
u
t
i
n
th
e
ex
tr
ac
tio
n
p
r
o
ce
s
s
f
o
r
ca
s
e
s
o
f
ex
p
licit
asp
ec
t
an
d
im
p
lic
it
o
p
in
io
n
.
Op
en
I
E
f
r
o
m
Stan
f
o
r
d
NL
P
ar
e
u
s
ed
to
id
en
tify
th
e
en
titi
es
in
ea
c
h
r
ev
iew,
an
d
th
ey
ar
e
s
to
r
ed
in
th
e
en
tity
co
r
p
u
s
.
L
astl
y
,
asp
ec
t
an
d
o
p
in
io
n
ca
n
d
id
ate
wo
r
d
s
ar
e
ex
tr
ac
ted
.
T
h
e
E
A
E
O
e
x
t
r
a
ct
i
o
n
c
o
n
s
i
s
ts
o
f
s
e
v
e
r
a
l
c
o
n
d
i
ti
o
n
s
,
i
n
c
l
u
d
i
n
g
:
a.
T
h
e
r
e
a
r
e
a
s
i
n
g
le
N
N
a
n
d
a
s
in
g
l
e
J
J
.
b.
T
h
e
r
e
a
r
e
a
s
i
n
g
le
N
N
a
n
d
m
u
lt
i
p
l
e
J
J
s
.
c.
T
h
e
r
e
a
r
e
m
u
l
t
i
p
l
e
N
Ns
a
n
d
a
s
i
n
g
l
e
J
J
.
d.
T
h
e
r
e
a
r
e
m
u
l
t
i
p
l
e
N
Ns
a
n
d
J
J
s
.
3
.
5
.
2
.
I
AE
O
ex
t
ra
ct
io
n
First,
th
e
ca
n
d
id
ate
r
esu
lts
o
f
asp
ec
t
an
d
o
p
in
io
n
ter
m
s
ar
e
p
r
o
ce
s
s
ed
to
o
b
tain
wo
r
d
s
th
a
t
in
d
icate
o
p
in
io
n
ter
m
s
in
th
e
o
p
in
i
o
n
t
er
m
ex
tr
ac
tio
n
.
I
n
th
e
asp
ec
t
t
er
m
ex
tr
ac
tio
n
,
we
u
s
e
th
e
im
p
licit
asp
ec
t
co
r
p
u
s
[
6
]
as
i
n
p
u
t
t
o
id
e
n
tify
w
o
r
d
s
th
at
ar
e
im
p
licit
asp
ec
t
ter
m
s
.
I
f
th
e
im
p
licit
asp
ec
t
ter
m
c
an
n
o
t
b
e
id
e
n
tifie
d
,
we
s
ea
r
ch
u
s
in
g
an
ex
p
licit
co
-
o
cc
u
r
r
en
ce
co
n
tex
t
o
p
in
io
n
ter
m
with
th
e
f
o
u
r
asp
ec
t
ca
teg
o
r
ies.
T
h
en
,
f
r
o
m
th
e
r
esu
lts
o
f
th
is
p
r
o
ce
s
s
,
an
ex
p
a
n
s
io
n
p
r
o
ce
s
s
is
ca
r
r
ied
o
u
t
b
ased
o
n
wo
r
d
s
y
n
o
n
y
m
s
to
g
et
th
e
h
ig
h
est
wo
r
d
s
im
ilar
ity
v
alu
e
f
r
o
m
th
e
co
-
o
cc
u
r
r
e
n
ce
co
n
tex
t
ter
m
r
esu
lts
f
o
r
th
e
f
o
u
r
asp
ec
t
ca
teg
o
r
ies.
T
h
e
h
ig
h
est
wo
r
d
s
im
ilar
ity
v
alu
e
is
th
en
ex
tr
ac
ted
as
th
e
im
p
licit
asp
ec
t
ter
m
.
Fu
r
th
er
m
o
r
e,
th
e
r
esu
lt
in
g
im
p
licit
asp
ec
t
ter
m
is
s
to
r
ed
in
th
e
im
p
licit
a
s
p
ec
t
co
r
p
u
s
.
T
h
is
im
p
licit
asp
ec
t
co
r
p
u
s
is
u
p
d
ated
au
t
o
m
a
tically
wh
en
a
n
ew
im
p
licit
asp
ec
t
ter
m
is
o
b
tai
n
ed
.
L
astl
y
,
asp
ec
t
a
n
d
o
p
in
io
n
ca
n
d
id
ate
w
o
r
d
s
ar
e
ex
t
r
ac
ted
.
T
h
e
I
A
E
O
e
x
t
r
a
c
t
i
o
n
c
o
n
s
i
s
t
s
o
f
t
h
e
f
o
ll
o
w
i
n
g
c
o
n
d
i
t
i
o
n
s
:
a.
T
h
e
r
e
i
s
n
o
N
N
,
b
u
t
t
h
e
r
e
is
o
n
e
s
i
n
g
l
e
J
J
.
b.
T
h
e
r
e
i
s
n
o
N
N
,
b
u
t
t
h
e
r
e
a
r
e
m
u
l
t
i
p
l
e
J
J
s
.
c.
N
N
o
r
J
J
a
r
e
m
a
r
k
e
d
a
s
i
m
p
li
c
i
t
as
p
e
c
t
t
e
r
m
s
b
a
s
e
d
o
n
t
h
e
i
m
p
l
i
c
it
as
p
e
c
t
c
o
r
p
u
s
l
is
t
.
Fo
r
e
x
a
m
p
l
e
,
t
h
e
r
e
v
i
e
w
t
h
e
f
o
o
d
i
s
e
x
p
e
n
s
i
v
e
,
w
h
e
r
e
e
x
p
e
n
s
i
v
e
i
s
a
n
o
p
i
n
i
o
n
t
e
r
m
m
a
r
k
e
d
a
s
a
n
i
m
p
l
i
c
i
t
asp
e
c
t
o
f
t
h
e
t
e
r
m
p
r
i
c
es
.
T
h
u
s
,
t
h
e
r
e
v
i
ew
c
o
n
ta
i
n
s
a
n
e
x
p
li
c
it
as
p
e
c
t
o
f
t
h
e
t
e
r
m
f
o
o
d
a
n
d
a
n
i
m
p
li
c
i
t
a
s
p
e
c
t
o
f
t
h
e
t
e
r
m
p
r
i
c
es
.
W
e
a
d
d
e
d
a
n
e
w
a
s
p
e
ct
a
n
d
o
p
i
n
i
o
n
t
e
r
m
t
a
r
g
e
t
b
a
s
ed
o
n
t
h
e
r
e
v
i
e
w
I
D
n
u
m
b
e
r
i
n
t
h
i
s
c
as
e
.
T
h
i
s
c
a
s
e
c
o
n
s
i
s
t
s
o
f
t
h
e
f
o
ll
o
w
i
n
g
co
n
d
i
t
i
o
n
s
:
−
T
h
e
r
e
a
r
e
a
s
i
n
g
le
N
N
a
n
d
a
s
in
g
l
e
J
J
.
−
T
h
e
r
e
a
r
e
a
s
i
n
g
le
N
N
a
n
d
m
u
lt
i
p
l
e
J
J
s
.
−
T
h
e
r
e
a
r
e
m
u
l
t
i
p
l
e
N
Ns
a
n
d
a
s
i
n
g
l
e
J
J
.
−
T
h
e
r
e
a
r
e
m
u
l
t
i
p
l
e
N
Ns
a
n
d
J
J
s
.
3
.
5
.
3
.
E
AIO
ex
t
ra
ct
io
n
First,
th
e
ca
n
d
id
ate
r
esu
lts
o
f
asp
ec
t
an
d
o
p
in
io
n
ter
m
s
ar
e
p
r
o
ce
s
s
ed
to
o
b
tain
in
te
n
t
wo
r
d
s
,
wh
ich
in
d
icate
im
p
licit
o
p
in
io
n
ter
m
s
in
th
e
in
ten
t
ex
tr
ac
tio
n
.
T
h
e
n
,
th
e
en
tity
co
r
p
u
s
we
co
llec
ted
is
u
s
ed
to
g
et
wo
r
d
s
d
en
o
tin
g
asp
ec
t
ter
m
s
i
n
th
e
e
n
tity
ex
tr
ac
tio
n
.
T
h
e
tw
o
ex
tr
ac
tio
n
r
esu
lts
ar
e
p
air
ed
b
ased
o
n
th
e
wo
r
d
r
elatio
n
th
at
co
n
n
ec
ts
th
e
in
ten
t
wo
r
d
s
.
W
e
u
s
e
th
e
r
esu
lts
o
f
th
ese
in
ten
t
-
en
tity
wo
r
d
p
a
ir
s
as
in
p
u
t
to
g
et
wo
r
d
d
ef
in
itio
n
s
f
r
o
m
W
ik
ip
ed
ia.
T
h
e
en
tity
-
in
ten
t
d
ef
in
it
io
n
f
r
o
m
th
e
lin
k
e
d
W
ik
ip
ed
ia
p
ag
e
is
u
s
ed
to
o
b
tain
t
h
e
o
p
in
io
n
ter
m
b
ase
d
o
n
th
e
ex
p
licit
o
p
in
io
n
ter
m
lis
t
th
at
r
esu
lted
f
r
o
m
th
e
E
AE
O
ex
t
r
ac
tio
n
.
Su
p
p
o
s
e
we
ca
n
n
o
t p
r
o
d
u
ce
a
n
o
p
in
io
n
ter
m
co
r
r
esp
o
n
d
in
g
to
an
ex
p
licit o
p
in
io
n
lis
t.
I
n
th
at
ca
s
e,
we
ex
p
an
d
b
ased
o
n
th
e
d
ef
in
itio
n
o
f
all
o
p
in
io
n
ter
m
s
id
en
tifie
d
f
r
o
m
th
e
tar
g
et
-
lin
k
ed
W
ik
ip
ed
ia
p
a
g
e.
Mo
r
eo
v
er
,
a
co
-
o
cc
u
r
r
e
n
ce
co
n
te
x
t
ter
m
is
p
e
r
f
o
r
m
e
d
o
n
th
e
e
n
tity
-
in
ten
t
p
air
s
to
g
et
th
e
wo
r
d
s
im
ilar
ity
with
th
e
h
i
g
h
est
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
A
s
p
ec
t
-
b
a
s
ed
s
en
timen
t a
n
a
l
ysis
:
n
a
tu
r
a
l la
n
g
u
a
g
e
u
n
d
ers
ta
n
d
in
g
fo
r
imp
licit
r
ev
iew
(
S
u
h
a
r
iya
n
to
)
6717
v
alu
e.
T
h
e
ter
m
with
th
e
h
ig
h
est
v
alu
e
is
m
ar
k
ed
as
th
e
im
p
licit
o
p
in
io
n
ter
m
an
d
s
av
ed
in
th
e
co
r
p
u
s
.
T
h
is
im
p
licit
o
p
in
io
n
co
r
p
u
s
is
u
p
d
ated
au
to
m
atica
lly
wh
en
n
e
w
im
p
licit
o
p
in
io
n
ter
m
s
ar
e
o
b
tain
ed
.
L
astl
y
,
th
e
ca
n
d
id
ates
o
f
asp
ec
t
an
d
o
p
in
io
n
ter
m
s
ar
e
ex
tr
ac
ted
.
T
h
e
E
A
I
O
e
x
t
r
a
c
ti
o
n
c
o
n
s
is
ts
o
f
t
h
e
f
o
ll
o
w
i
n
g
c
o
n
d
i
t
i
o
n
s
:
a.
T
h
e
r
e
i
s
a
s
i
n
g
l
e
N
N
a
n
d
n
o
J
J
.
b.
T
h
e
r
e
a
r
e
m
u
l
t
i
p
l
e
N
Ns
a
n
d
n
o
J
J
.
c.
T
h
e
r
e
a
r
e
a
n
N
N
a
n
d
a
J
J
,
b
u
t
th
e
p
a
i
r
i
n
g
t
e
r
m
o
f
N
N
a
n
d
J
J
ca
n
n
o
t
b
e
e
x
t
r
a
c
t
e
d
u
s
i
n
g
t
h
e
r
e
l
a
t
i
o
n
r
u
l
e
s
NN
a
n
d
J
J
i
n
t
h
e
E
A
E
O
e
x
t
r
a
c
ti
o
n
p
r
o
c
e
s
s
.
T
h
is
c
as
e
c
o
n
s
i
s
t
s
o
f
th
e
f
o
l
l
o
w
i
n
g
c
o
n
d
i
t
i
o
n
s
:
−
T
h
e
r
e
a
r
e
a
s
i
n
g
le
N
N
a
n
d
a
s
in
g
l
e
J
J
.
−
T
h
e
r
e
a
r
e
a
s
i
n
g
le
N
N
a
n
d
m
u
lt
i
p
l
e
J
J
s
.
−
T
h
e
r
e
a
r
e
m
u
l
t
i
p
l
e
N
Ns
a
n
d
a
s
i
n
g
l
e
J
J
.
−
T
h
e
r
e
a
r
e
m
u
l
t
i
p
l
e
N
Ns
a
n
d
J
J
s
.
W
e
m
ar
k
ce
r
tain
VE
R
B
s
as
i
m
p
licit o
p
in
io
n
ter
m
s
b
ased
o
n
th
e
o
b
j r
elatio
n
f
o
r
r
ev
iew
c
ases
w
ith
an
NN
an
d
n
o
J
J
.
T
h
en
,
t
h
e
VE
R
B
is
m
ar
k
ed
as
th
e
in
ten
t.
T
o
g
et
th
e
o
p
in
io
n
ter
m
VE
R
B
r
elatio
n
s
h
ip
with
th
e
tar
g
et
asp
ec
t
ter
m
NN,
an
ass
o
ciatio
n
r
u
le
was
b
u
ilt
b
y
tak
i
n
g
th
e
en
titi
es
in
th
e
s
en
ten
ce
b
ased
o
n
th
e
en
tit
y
co
r
p
u
s
r
esu
ltin
g
f
r
o
m
t
h
e
E
A
E
O
ex
tr
ac
tio
n
.
T
h
en
,
th
e
en
tity
-
in
ten
t
is
ex
p
an
d
e
d
b
ased
o
n
th
e
d
ef
in
itio
n
o
f
th
e
wo
r
d
tak
en
f
r
o
m
W
ik
ip
ed
ia
to
g
et
th
e
ac
tu
al
wo
r
d
m
ea
n
i
n
g
i
n
th
e
s
en
ten
ce
.
Fo
r
in
s
tan
ce
,
th
e
r
ev
iew
[
1
0
0
1
]
o
f
th
e
w
a
it
time
h
ere
fo
r
th
e
fo
o
d
ev
en
mo
r
e
ma
ke
s
my
s
to
ma
ch
ve
r
y
r
u
mb
lin
g
h
as:
th
e
asp
ec
t
ter
m
s
o
f
w
a
it
time
,
fo
o
d
,
an
d
s
to
ma
ch
;
an
d
th
er
e
is
n
o
ex
tr
ac
te
d
o
p
in
io
n
ter
m
.
T
h
e
ter
m
s
o
f
w
a
it
time
an
d
s
to
m
a
ch
ar
e
ex
tr
ac
ted
as
th
e
en
titi
es.
W
e
p
air
th
e
en
titi
es
wi
th
th
eir
in
ten
t
f
o
r
d
ef
in
itio
n
-
b
ased
ex
p
an
s
io
n
t
o
ex
tr
ac
t
th
e
tar
g
et
o
p
in
i
o
n
t
er
m
b
ased
o
n
th
e
e
x
p
licit
o
p
in
io
n
c
o
r
p
u
s
.
T
h
e
ex
tr
ac
ted
p
ai
r
in
g
ter
m
s
o
f
:
ar
e
w
a
it
time:
ma
ke
s
an
d
s
to
ma
ch
:
r
u
mb
le
.
T
h
e
p
ai
r
in
g
ter
m
s
ar
e
u
s
ed
as
th
e
in
p
u
t
f
o
r
d
ef
in
it
io
n
-
b
ased
ex
p
a
n
s
io
n
r
esu
lts
.
B
ased
o
n
th
e
d
e
f
in
itio
n
ex
p
an
s
io
n
r
esu
lt,
w
a
it
time:
ma
ke
s
d
o
es
n
o
t
h
a
v
e
a
d
ef
in
itio
n
ca
n
b
e
id
e
n
tifie
d
,
an
d
s
to
ma
ch
:
r
u
mb
le
h
as
th
r
ee
d
ef
in
itio
n
s
.
W
e
ex
tr
ac
t
th
e
ex
p
licit
o
p
i
n
io
n
ter
m
lab
eled
as
J
J
in
th
e
ex
tr
ac
ted
d
ef
in
itio
n
a
n
d
th
e
r
esu
lt
is
h
u
n
g
r
y
.
Fin
ally
,
we
ex
tr
ac
t
th
e
s
y
n
o
n
y
m
s
o
f
h
u
n
g
r
y
b
ased
o
n
th
e
d
ef
in
itio
n
as
th
e
r
esu
lt.
T
h
e
r
esu
lts
ar
e:
feelin
g
,
w
a
n
tin
g
ea
t
;
n
o
t
ea
ten
,
few
h
o
u
r
s
,
an
d
litt
le
ea
t
.
3
.
5
.
4
.
I
AIO
ex
t
ra
ct
io
n
F
i
r
s
t
,
t
h
e
c
a
n
d
i
d
at
e
r
e
s
u
l
t
s
o
f
a
s
p
e
c
t
a
n
d
o
p
i
n
i
o
n
t
e
r
m
s
a
r
e
p
r
o
c
e
s
s
e
d
t
o
o
b
t
a
i
n
w
o
r
d
s
t
h
a
t
a
r
e
l
a
b
e
l
e
d
a
s
i
m
p
li
c
i
t
o
p
i
n
i
o
n
t
e
r
m
s
.
T
h
i
s
s
t
e
p
u
s
es
i
n
p
u
t
f
r
o
m
t
h
e
i
m
p
l
i
c
i
t
o
p
i
n
i
o
n
c
o
r
p
u
s
g
e
n
e
r
a
t
e
d
f
r
o
m
t
h
e
E
A
I
O
e
x
t
r
a
c
ti
o
n
.
M
e
a
n
w
h
i
l
e
,
i
n
a
s
p
e
c
t
t
e
r
m
e
x
t
r
a
ct
i
o
n
,
w
e
u
s
e
t
h
e
i
m
p
l
i
ci
t
a
s
p
e
ct
c
o
r
p
u
s
t
h
a
t
w
a
s
u
p
d
a
t
e
d
i
n
I
A
E
O
e
x
t
r
a
ct
i
o
n
t
o
g
e
t
w
o
r
d
s
d
e
n
o
t
i
n
g
i
m
p
l
i
c
it
a
s
p
ec
t
t
e
r
m
s
.
L
a
s
t
l
y
,
w
e
e
x
t
r
a
ct
c
a
n
d
i
d
a
te
te
r
m
s
o
f
a
s
p
e
ct
a
n
d
o
p
i
n
i
o
n
.
T
h
e
I
A
I
O
e
x
t
r
a
c
t
i
o
n
c
o
n
s
is
ts
o
f
t
h
e
c
o
n
d
it
i
o
n
s
w
it
h
n
o
N
N
a
n
d
J
J
as
e
x
p
l
i
ci
t
te
r
m
s
o
f
as
p
e
c
t
a
n
d
o
p
i
n
i
o
n
.
3
.
6
.
Asp
ec
t
ca
t
eg
o
riza
t
io
n
T
h
e
p
r
o
p
o
s
ed
asp
ec
t
ca
t
e
g
o
r
i
z
ati
o
n
(
AC
)
ai
m
s
t
o
m
ea
s
u
r
e
a
n
d
d
ete
r
m
in
e
t
h
e
asp
ec
t
ca
t
e
g
o
r
ies
o
f
t
h
e
ter
m
s
g
e
n
e
r
a
te
d
f
r
o
m
th
e
te
x
t
e
x
t
r
a
cti
o
n
s
t
ag
e.
I
n
t
h
e
AC
s
tag
e,
th
e
N
L
U
f
r
am
ew
o
r
k
r
et
r
iev
es
t
h
e
ex
tr
ac
t
e
d
tar
g
et
asp
ec
t
t
er
m
s
in
d
i
ca
te
d
b
y
n
o
u
n
s
a
n
d
n
o
u
n
p
h
r
ases
.
E
x
p
li
cit
a
n
d
i
m
p
li
cit
r
ev
iews
c
an
h
av
e
o
n
e
o
r
m
o
r
e
cr
i
te
r
ia
b
ase
d
o
n
t
h
e
ta
r
g
et
as
p
ec
t
t
er
m
s
m
e
n
ti
o
n
e
d
i
n
e
ac
h
r
ev
i
ew
.
T
h
e
ta
r
g
et
asp
ec
t
te
r
m
s
a
r
e
d
ete
cte
d
b
as
e
d
o
n
m
e
n
ti
o
n
i
n
g
t
h
e
e
x
p
lic
it
as
p
e
ct
t
e
r
m
b
u
t
n
o
t
t
h
e
ac
tu
al
ta
r
g
et
o
f
t
h
e
i
m
p
li
cit
o
p
i
n
i
o
n
t
e
r
m
.
I
n
a
d
d
it
io
n
,
t
h
e
ex
p
a
n
s
i
o
n
o
f
as
p
ec
t
t
e
r
m
s
b
as
ed
o
n
wo
r
d
s
y
n
o
n
y
m
s
a
n
d
w
o
r
d
d
e
f
i
n
it
io
n
s
is
als
o
ca
r
r
ie
d
o
u
t
t
o
o
b
t
ai
n
m
o
r
e
ac
c
u
r
at
e
as
p
ec
t
ca
te
g
o
r
ies
.
W
e
u
s
e
B
E
R
T
e
m
b
e
d
d
in
g
[
1
9
]
a
n
d
Se
m
a
n
ti
c
s
i
m
il
ar
it
y
[
1
8
]
t
o
m
e
asu
r
e
t
h
e
s
im
il
ar
it
y
b
e
twe
e
n
t
h
e
asp
ec
t
t
er
m
s
an
d
t
h
e
f
o
u
r
asp
ec
t
c
at
eg
o
r
ies:
a
m
b
ia
n
ce
,
f
o
o
d
,
p
r
ic
e,
a
n
d
s
e
r
v
i
ce
.
Fo
r
illu
s
tr
atio
n
,
T
ab
le
4
s
h
o
ws
th
e
r
esu
lts
o
f
s
em
an
tic
s
im
ilar
ity
ca
lcu
latio
n
s
in
th
e
asp
ec
t
ca
teg
o
r
izatio
n
s
tag
e
o
f
I
D
R
e
v
iew
[
1
0
0
1
]
.
T
h
e
r
esu
lts
o
f
th
e
th
r
ee
ter
m
s
with
th
e
h
ig
h
est
s
co
r
es
o
b
tain
ed
in
th
e
tex
t
ex
tr
ac
tio
n
s
tag
e
r
e
p
la
ce
ter
m
s
wer
e
m
ar
k
ed
as
im
p
l
icit
in
th
e
ex
is
tin
g
r
ev
iew.
T
h
en
,
th
e
ter
m
s
u
p
d
ate
m
ea
s
u
r
ed
th
e
s
im
ilar
ity
v
alu
e
o
f
th
e
k
ey
wo
r
d
v
ar
iab
les f
r
o
m
ea
ch
asp
ec
t c
ateg
o
r
y
.
Af
ter
u
p
d
atin
g
th
e
im
p
licit
ter
m
s
,
th
e
h
ig
h
est
av
er
ag
e
to
tal
s
im
ilar
ity
v
alu
e
was
0
.
8
3
1
1
f
o
r
Ser
v
ice
asp
ec
t
ca
teg
o
r
y
.
So
,
t
h
e
asp
ec
t
ca
teg
o
r
izatio
n
r
esu
lts
f
o
r
I
D
r
e
v
iews [
1
0
0
1
]
is
“
SERVICE
”
.
T
ab
le
4
.
Sem
an
tic
s
im
ilar
ity
c
alcu
latio
n
r
esu
lts
Te
r
m
l
i
st
A
sp
e
c
t
1
A
sp
e
c
t
2
A
sp
e
c
t
3
A
sp
e
c
t
4
w
a
i
t
t
i
m
e
,
f
o
o
d
,
m
a
k
e
s,
f
e
e
l
i
n
g
,
w
a
n
t
i
n
g
e
a
t
0
.
7
9
8
4
0
.
8
2
0
3
0
.
8
3
7
8
0
.
7
0
6
5
w
a
i
t
t
i
m
e
,
f
o
o
d
,
m
a
k
e
s,
n
o
t
e
a
t
e
n
,
f
e
w
h
o
u
rs
0
.
7
8
7
6
0
.
8
2
0
7
0
.
8
3
9
9
0
.
7
2
0
9
w
a
i
t
t
i
m
e
,
f
o
o
d
,
m
a
k
e
s,
l
i
t
t
l
e
e
a
t
0
.
7
8
1
4
0
.
7
9
8
3
0
.
8
1
5
7
0
.
7
0
3
7
To
t
a
l
A
v
g
.
0
.
7
8
9
1
0
.
8
1
3
1
0
.
8
3
1
1
0
.
7
1
0
4
N
B
:
A
s
p
e
c
t
1
i
s
a
mb
i
a
n
c
e
;
A
sp
e
c
t
2
i
s fo
o
d
;
A
s
p
e
c
t
3
i
s se
r
v
i
c
e
,
a
n
d
A
s
p
e
c
t
4
i
s
p
r
i
c
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.
14
,
No
.
6
,
Dec
em
b
e
r
20
24
:
6
7
1
1
-
6
7
2
2
6718
3
.
7
.
AB
SA
T
h
e
p
r
o
p
o
s
ed
r
esear
ch
s
y
s
te
m
is
s
h
o
wn
in
Fig
u
r
e
1
.
B
ased
o
n
Fig
u
r
e
1
,
th
e
p
r
o
p
o
s
ed
AB
SA
s
tag
e
u
s
es
th
e
in
p
u
t
asp
ec
t
ca
teg
o
r
y
f
r
o
m
th
e
r
esu
lts
o
f
th
e
asp
ec
t
ca
teg
o
r
izatio
n
s
tag
e.
T
h
e
n
,
th
e
r
esu
lts
o
f
th
e
o
p
in
io
n
ter
m
s
f
r
o
m
t
h
e
tex
t
e
x
tr
ac
tio
n
s
tag
e
a
r
e
u
s
ed
as
in
p
u
t
f
o
r
d
eter
m
i
n
in
g
th
e
s
en
tim
e
n
t
p
o
lar
ity
.
W
e
u
s
e
B
E
R
T
s
en
tim
en
t a
n
aly
s
is
,
a
h
y
b
r
id
lex
ic
o
n
,
a
n
d
o
p
in
io
n
d
eg
r
ee
weig
h
tin
g
f
o
r
AB
SA m
eth
o
d
s
.
3
.
7
.
1
.
B
E
RT
s
ent
im
ent
a
na
ly
s
is
B
E
R
T
s
e
n
ti
m
e
n
t
a
n
a
l
y
s
is
[
2
3
]
i
s
u
s
e
d
t
o
e
x
t
r
a
c
t
o
p
i
n
i
o
n
w
o
r
d
s
i
n
t
o
f
o
u
r
c
l
a
s
s
e
s
:
p
o
s
it
i
v
e
,
v
e
r
y
p
o
s
i
t
i
v
e
,
n
e
g
a
ti
v
e
,
a
n
d
v
e
r
y
n
e
g
a
t
i
v
e
.
H
o
w
e
v
e
r
,
t
h
e
r
e
is
n
o
e
x
p
l
i
c
i
t
w
e
i
g
h
t
i
n
g
o
f
t
h
e
o
p
in
i
o
n
w
o
r
d
s
,
w
h
i
c
h
c
a
u
s
es
s
o
m
e
w
o
r
d
s
t
o
b
e
c
l
ass
i
f
i
e
d
i
n
c
o
r
r
e
c
t
l
y
.
F
o
r
e
x
a
m
p
le
,
t
h
e
w
o
r
d
e
x
p
e
n
s
i
v
e
i
s
c
l
as
s
i
f
i
e
d
i
n
t
h
e
p
o
s
i
t
i
v
e
c
l
a
s
s
.
M
o
r
e
o
v
e
r
,
c
o
m
p
l
e
m
e
n
tar
y
p
h
r
a
s
e
s
t
h
a
t
h
a
v
e
n
o
t
b
ee
n
c
a
p
t
u
r
e
d
a
l
s
o
c
a
u
s
e
t
h
e
s
a
m
e
a
s
s
e
s
s
m
e
n
t
b
e
tw
e
e
n
s
i
n
g
l
e
a
n
d
c
o
m
p
l
e
m
e
n
t
a
r
y
o
p
in
i
o
n
w
o
r
d
s
.
F
o
r
e
x
a
m
p
l
e
,
e
x
p
e
n
s
i
v
e
a
n
d
v
e
r
y
e
x
p
e
n
s
i
v
e
.
T
h
e
r
e
f
o
r
e
,
w
e
c
o
m
b
i
n
e
d
B
E
R
T
w
it
h
a
h
y
b
r
i
d
l
e
x
i
c
o
n
a
n
d
o
p
i
n
i
o
n
d
e
g
r
e
e
w
e
i
g
h
t
i
n
g
t
o
o
v
e
r
c
o
m
e
t
h
e
s
e
p
r
o
b
l
e
m
s
.
3
.
7
.
2
.
H
y
brid
l
ex
ico
n
W
e
u
s
e
d
a
h
y
b
r
i
d
l
e
x
i
c
o
n
[
1
5
]
,
c
o
m
b
i
n
e
d
t
h
e
o
p
i
n
i
o
n
l
e
x
i
c
o
n
a
n
d
t
h
e
i
m
p
l
i
c
i
t
o
p
i
n
i
o
n
l
e
x
i
c
o
n
,
t
o
h
a
n
d
l
e
p
r
o
b
l
e
m
s
i
n
t
h
e
e
x
i
s
t
i
n
g
l
e
x
i
c
o
n
a
n
d
g
r
o
u
p
e
a
c
h
w
o
r
d
b
a
s
e
d
o
n
t
h
e
a
c
t
u
a
l
c
l
a
s
s
.
T
h
e
o
p
i
n
i
o
n
l
e
x
i
c
o
n
c
o
n
s
i
s
t
e
d
o
f
6
,
8
0
0
w
o
r
d
s
f
r
o
m
t
w
o
c
l
ass
e
s
o
f
p
o
s
i
ti
v
e
a
n
d
n
e
g
a
t
i
v
e
s
e
n
t
i
m
e
n
ts
.
T
h
e
n
,
t
h
e
i
m
p
l
i
ci
t
o
p
i
n
i
o
n
l
e
x
i
c
o
n
w
as
g
e
n
e
r
a
t
e
d
m
a
n
u
a
lly
u
s
i
n
g
t
h
e
p
r
o
p
o
s
e
d
t
e
x
t
e
x
t
r
ac
t
i
o
n
m
e
t
h
o
d
.
3
.
7
.
3
.
O
pin
io
n deg
re
e
weig
hting
W
e
m
a
n
u
a
l
l
y
a
d
d
e
d
w
e
i
g
h
ti
n
g
d
e
g
r
e
e
s
o
f
o
p
i
n
i
o
n
t
e
r
m
s
to
h
a
n
d
l
e
p
r
o
b
l
e
m
s
i
n
d
e
te
r
m
i
n
i
n
g
t
h
e
s
e
n
t
i
m
e
n
t
a
n
d
g
r
o
u
p
c
o
m
p
l
e
m
e
n
t
a
r
y
p
h
r
a
s
es
b
as
e
d
o
n
t
h
e
ac
t
u
a
l
o
p
i
n
i
o
n
t
e
r
m
cl
a
s
s
.
W
e
u
p
d
a
t
e
d
t
h
e
p
o
l
a
r
it
y
a
n
d
s
t
r
e
n
g
t
h
o
f
s
e
n
t
i
m
e
n
t
u
s
i
n
g
t
h
e
V
a
d
e
r
o
p
i
n
i
o
n
l
e
x
i
c
o
n
[
2
1
]
.
T
h
e
u
p
d
a
t
e
d
s
e
n
t
i
m
e
n
t
s
c
o
n
s
i
s
t
o
f
v
e
r
y
p
o
s
i
t
i
v
e
,
p
o
s
i
t
i
v
e
,
n
e
g
at
i
v
e
,
a
n
d
v
e
r
y
n
e
g
a
t
i
v
e
.
T
o
g
et
th
e
s
en
tim
en
t
p
o
lar
ity
v
alu
e
o
f
all
o
p
in
io
n
ter
m
s
,
w
e
d
eter
m
in
e
d
th
e
r
an
g
e
o
f
v
alu
es
f
r
o
m
+4
to
-
4
f
o
r
t
h
e
p
o
s
itiv
e
an
d
n
eg
ativ
e
class
es.
T
h
en
,
to
s
y
n
c
h
r
o
n
ize
th
e
s
en
tim
en
t
p
o
lar
ity
v
alu
e
o
f
a
n
o
p
in
io
n
ter
m
,
we
tr
an
s
f
o
r
m
e
d
th
e
p
o
lar
ity
s
en
tim
en
t
v
alu
e
f
r
o
m
+1
to
-
1
u
s
in
g
th
e
n
o
r
m
aliza
tio
n
f
u
n
ctio
n
=
√
2
+
,
wh
er
e
y
is
th
e
o
p
in
io
n
weig
h
t
v
alu
e
o
f
ex
is
tin
g
ter
m
s
.
Af
ter
o
b
tain
in
g
th
e
n
o
r
m
alize
d
v
alu
e
o
f
p
o
lar
ity
s
en
tim
en
t,
we
ca
lcu
l
ated
th
e
to
tal
v
al
u
e
o
f
s
en
tim
en
t
p
o
lar
ity
b
ased
o
n
a
d
v
e
r
b
p
air
s
u
s
in
g
th
e
eq
u
atio
n
′
=
×
,
wh
er
e
u
is
th
e
ad
v
er
b
d
eg
r
ee
v
a
r
iab
le
f
o
r
s
en
tim
en
t
p
o
lar
ity
p
air
ed
with
th
e
o
p
in
io
n
ter
m
.
Fo
r
th
is
r
ea
s
o
n
,
we
co
m
p
iled
a
lis
t
o
f
a
d
v
er
b
d
eg
r
e
e
v
ar
iab
le
v
al
u
es
th
at
ca
n
a
f
f
ec
t
th
e
p
o
lar
ity
o
f
s
en
tim
en
t,
as sh
o
wn
in
T
ab
le
5
.
T
ab
le
5
.
Valu
e
lis
t o
f
th
e
ad
v
e
r
b
d
eg
r
ee
v
ar
ia
b
le
f
o
r
s
en
tim
en
t p
o
lar
ity
A
d
v
e
r
b
d
e
g
r
e
e
v
a
r
i
a
b
l
e
f
o
r
s
e
n
t
i
me
n
t
p
o
l
a
r
i
t
y
Le
v
e
l
V
a
r
i
a
b
l
e
A
sp
e
c
t
t
e
r
m
t
a
r
g
e
t
V
a
l
u
e
Le
v
e
l
1
t
o
o
I
t
e
m o
f
f
o
o
d
t
a
st
e
t
y
p
e
-
1
Le
v
e
l
2
b
a
r
e
l
y
;
sc
a
rc
e
l
y
A
l
l
a
sp
e
c
t
t
e
r
ms
-
0
.
5
Le
v
e
l
3
a
l
m
o
st
;
q
u
i
t
e
;
s
o
m
e
w
h
a
t
;
e
n
o
u
g
h
;
l
i
t
t
l
e
;
l
e
s
s;
l
e
a
s
t
;
a
b
i
t
A
l
l
a
sp
e
c
t
t
e
r
ms
+
0
.
5
Le
v
e
l
4
o
t
h
e
r
a
d
v
e
r
b
s
t
h
a
t
d
o
n
o
t
e
x
i
s
t
a
t
l
e
v
e
l
s 1
,
2
,
a
n
d
3
;
n
o
i
n
t
e
n
si
f
i
e
r
A
l
l
a
sp
e
c
t
t
e
r
ms
+1
F
u
r
t
h
e
r
m
o
r
e
,
t
o
o
b
t
a
i
n
t
h
e
s
e
n
t
i
m
e
n
t
s
t
r
e
n
g
t
h
v
a
l
u
e
f
r
o
m
t
h
e
t
e
r
m
,
we
d
e
t
e
r
m
i
n
e
d
t
h
e
v
a
l
u
e
o
f
t
h
e
o
p
i
n
i
o
n
t
e
r
m
d
e
g
r
e
e
t
h
r
o
u
g
h
p
a
i
r
s
o
f
i
n
t
e
n
s
i
f
i
e
r
s
w
it
h
a
r
a
n
g
e
o
f
v
a
l
u
e
s
.
+
1
t
o
-
1
f
o
r
v
e
r
y
p
o
s
i
t
i
v
e
,
p
o
s
it
i
v
e,
n
e
g
a
t
i
v
e
,
a
n
d
v
e
r
y
n
e
g
a
t
i
v
e
.
T
h
e
t
o
t
a
l
v
a
l
u
e
o
f
t
h
e
s
e
n
t
i
m
e
n
t
v
a
l
u
e
f
o
r
t
h
e
s
t
r
e
n
g
t
h
o
f
a
n
o
p
i
n
i
o
n
t
e
r
m
w
a
s
o
b
t
a
i
n
e
d
b
y
t
h
e
e
q
u
a
t
i
o
n
′
=
×
|
′
|
,
w
h
e
r
e
v
i
s
t
h
e
i
n
te
n
s
i
f
i
e
r
v
a
r
i
a
b
le
f
o
r
t
h
e
s
t
r
e
n
g
t
h
o
f
t
h
e
s
e
n
t
im
e
n
t
p
a
i
r
e
d
w
it
h
t
h
e
o
p
i
n
i
o
n
t
e
r
m
.
T
a
b
l
e
6
s
h
o
w
s
a
l
i
s
t
o
f
a
d
v
e
r
b
d
e
g
r
e
e
v
a
r
i
a
b
l
e
v
a
l
u
e
s
u
s
e
d
t
o
d
e
t
e
r
m
i
n
e
t
h
e
s
e
n
t
i
m
e
n
t
p
o
l
a
r
i
t
y
o
f
t
h
e
o
p
i
n
i
o
n
t
e
r
m
s
.
H
e
r
e
,
e
s
p
e
ci
a
l
l
y
f
o
r
t
h
e
a
d
v
e
r
b
t
o
o
,
w
e
a
d
d
e
d
t
o
t
h
e
s
t
r
e
n
g
t
h
o
f
t
h
e
j
u
d
g
m
e
n
t
s
e
n
ti
m
e
n
t
i
f
t
h
e
a
d
v
er
b
w
a
s
p
ai
r
e
d
wi
t
h
a
n
o
p
i
n
i
o
n
te
r
m
i
t
e
m
,
w
h
i
c
h
is
a
t
y
p
e
o
f
f
o
o
d
t
a
s
t
e
(
i
.
e
.
,
s
we
e
t
a
n
d
s
a
l
t
)
.
T
a
b
l
e
7
s
h
o
w
s
t
h
e
l
is
t
o
f
a
d
v
e
r
b
d
e
g
r
e
e
v
a
r
i
a
b
l
e
v
a
l
u
e
s
u
s
e
d
t
o
d
e
t
e
r
m
i
n
e
t
h
e
s
en
t
i
m
e
n
t
s
t
r
e
n
g
t
h
o
f
o
p
i
n
i
o
n
t
e
r
m
s
.
T
ab
le
8
s
h
o
ws
th
e
r
esu
lt
o
f
s
en
tim
en
t
p
o
lar
ity
d
eter
m
in
atio
n
f
o
r
r
e
v
iew
[
1
0
0
1
]
.
B
ased
o
n
th
e
p
r
o
p
o
s
ed
r
u
le
alg
o
r
ith
m
,
th
e
ter
m
ve
r
y
r
u
mb
lin
g
is
ex
tr
ac
ted
as
th
e
o
p
i
n
io
n
ter
m
.
T
h
en
,
th
e
s
en
tim
en
t
s
co
r
e
m
ea
s
u
r
em
en
t
r
esu
lt
is
-
0
.
6
2
5
.
Fin
ally
,
a
s
co
r
e
o
f
-
0
.
6
2
5
is
class
if
ied
as
a
v
er
y
n
eg
ativ
e
s
en
tim
en
t.
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
A
s
p
ec
t
-
b
a
s
ed
s
en
timen
t a
n
a
l
ysis
:
n
a
tu
r
a
l la
n
g
u
a
g
e
u
n
d
ers
ta
n
d
in
g
fo
r
imp
licit
r
ev
iew
(
S
u
h
a
r
iya
n
to
)
6719
T
ab
le
6
.
Valu
e
lis
t o
f
th
e
ad
v
e
r
b
d
eg
r
ee
v
ar
ia
b
le
f
o
r
s
en
tim
en
t stre
n
g
th
Le
v
e
l
V
a
r
i
a
b
l
e
V
a
l
u
e
Le
v
e
l
1
t
o
o
;
v
e
ry;
a
l
o
t
,
l
o
t
s
,
e
x
t
r
e
m
e
l
y
;
h
o
rr
i
b
l
y
;
u
n
u
su
a
l
l
y
;
w
o
n
d
e
rf
u
l
l
y
;
d
e
e
p
l
y
;
a
b
so
l
u
t
e
l
y
;
f
u
l
l
y
;
h
a
r
d
l
y
;
t
e
rri
b
l
y
;
p
ret
t
y
;
re
a
l
l
y
;
i
n
sa
n
e
l
y
;
rema
r
k
a
b
l
y
;
g
re
a
t
l
y
;
h
i
g
h
l
y
;
m
o
st
;
m
u
c
h
;
i
n
t
e
n
se
l
y
;
s
t
ro
n
g
l
y
;
u
t
t
e
rl
y
;
s
o
;
a
m
a
zi
n
g
l
y
-
1
Le
v
e
l
2
A
v
a
r
i
a
b
l
e
mi
x
t
u
r
e
o
f
l
e
v
e
l
s
1
a
n
d
3
-
0
.
5
Le
v
e
l
3
a
l
m
o
st
;
c
o
m
p
l
e
t
e
l
y
;
b
a
re
l
y
;
q
u
i
t
e
;
s
o
m
e
w
h
a
t
;
f
a
i
r
l
y
;
i
n
c
re
d
i
b
l
y
;
e
n
o
u
g
h
;
l
a
rg
e
l
y
;
sc
a
rc
e
l
y
;
b
a
d
l
y
;
l
i
t
t
l
e
;
l
e
ss;
l
e
a
st
;
j
u
s
t
;
p
u
r
e
l
y
;
t
h
o
r
o
u
g
h
l
y
+
0
.
5
Le
v
e
l
4
N
o
i
n
t
e
n
si
f
i
e
r
v
a
r
i
a
b
l
e
+1
T
ab
le
7
.
F1
-
s
co
r
e
r
esu
lts
f
o
r
A
C
an
d
AB
SA
S
e
n
t
e
n
c
e
t
y
p
e
C
a
ses
o
f
a
s
p
e
c
t
a
n
d
o
p
i
n
i
o
n
t
e
r
ms
Ex
p
l
i
c
i
t
I
mp
l
i
c
i
t
EA
EO
I
A
EO
EA
I
O
I
A
I
O
AC
A
B
S
A
AC
A
B
S
A
AC
A
B
S
A
AC
A
B
S
A
S
i
mp
l
e
0
.
9
8
2
6
0
.
9
8
9
1
0
.
9
6
5
1
0
.
9
5
0
6
0
.
9
5
8
7
0
.
9
2
7
2
0
.
8
7
5
6
0
.
8
4
9
8
C
o
m
p
o
u
n
d
0
.
9
8
3
5
0
.
9
8
3
1
0
.
9
5
0
2
0
.
9
5
5
8
0
.
9
4
1
2
0
.
9
0
7
0
0
.
9
1
8
0
0
.
8
6
8
9
C
o
m
p
l
e
x
0
.
9
8
5
1
0
.
9
7
3
3
0
.
9
5
0
5
0
.
9
5
5
2
0
.
9
3
1
8
0
.
9
1
4
3
0
.
8
7
5
0
0
.
8
5
3
9
C
o
m
p
o
u
n
d
-
c
o
mp
l
e
x
0
.
9
7
8
0
0
.
9
7
2
5
0
.
9
4
8
5
0
.
9
5
3
8
0
.
9
1
7
6
0
.
9
0
3
2
0
.
8
4
9
3
0
.
8
2
9
3
A
v
e
r
a
g
e
0
.
9
8
2
3
0
.
9
7
9
5
0
.
9
5
3
6
0
.
9
5
3
9
0
.
9
3
7
3
0
.
9
1
2
9
0
.
8
7
9
5
0
.
8
5
0
5
T
ab
le
8
.
R
esu
lt o
f
s
en
tim
en
t p
o
lar
ity
d
eter
m
in
atio
n
I
D
R
e
v
.
I
D
S
e
n
t
.
R
e
v
i
e
w
Ex
t
r
a
c
t
e
d
o
p
i
n
i
o
n
t
e
r
m
S
e
n
t
i
me
n
t
sc
o
r
e
S
e
n
t
i
me
n
t
r
e
s
u
l
t
1
0
0
1
0
t
h
e
w
a
i
t
t
i
me
h
e
r
e
f
o
r
t
h
e
f
o
o
d
e
v
e
n
mo
r
e
m
a
k
e
s m
y
st
o
ma
c
h
v
e
r
y
r
u
m
b
l
i
n
g
v
e
r
y
r
u
m
b
l
i
n
g
-
0
.
6
2
5
v
e
r
y
n
e
g
a
t
i
v
e
4.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
I
n
th
is
s
ec
tio
n
,
th
e
p
er
f
o
r
m
a
n
ce
o
f
o
u
r
s
tu
d
y
is
d
escr
ib
e
d
.
First,
th
e
tex
t
ex
tr
ac
tio
n
m
eth
o
d
f
o
r
ex
p
licit
an
d
im
p
licit
r
e
v
iews
is
ex
p
lain
ed
.
T
h
e
n
,
th
e
asp
ec
t
ca
teg
o
r
izatio
n
a
n
d
th
e
asp
e
ct
-
b
ased
s
en
tim
en
t
an
aly
s
is
ar
e
also
ex
p
lain
ed
.
L
a
s
tly
,
th
e
r
esu
lts
ar
e
also
co
m
p
a
r
ed
with
o
th
e
r
m
eth
o
d
s
.
4
.
1
.
T
ex
t
ex
t
ra
ct
i
o
n r
esu
lt
T
h
e
tex
t
e
x
tr
ac
tio
n
ca
n
p
r
o
p
er
ly
ex
tr
ac
t
t
h
e
asp
ec
t
a
n
d
o
p
in
io
n
ter
m
s
,
wh
ich
ar
e
e
x
p
licit
an
d
im
p
licit,
in
f
o
u
r
s
en
ten
ce
ty
p
es.
T
h
e
av
er
ag
e
F1
-
s
co
r
es
f
o
r
ex
p
licit
r
ev
iew
ca
s
es
o
f
E
A
E
O
ar
e
9
6
.
3
3
%
f
o
r
ex
p
licit
asp
ec
t
ex
tr
ac
tio
n
an
d
9
6
.
8
3
%
f
o
r
ex
p
licit
o
p
i
n
io
n
ex
tr
ac
tio
n
.
T
h
e
a
v
er
ag
e
F1
-
s
co
r
es
f
o
r
im
p
licit
r
ev
iew
ca
s
es
o
f
I
AE
O
ar
e
8
9
.
2
3
%
f
o
r
im
p
licit
asp
ec
t
ex
tr
a
ctio
n
an
d
9
5
.
8
9
%
f
o
r
ex
p
licit
o
p
in
io
n
ex
tr
ac
tio
n
.
T
h
e
av
er
ag
e
F1
-
s
co
r
es
f
o
r
i
m
p
licit
r
ev
iew
ca
s
es
o
f
E
AI
O
ar
e
9
5
.
0
6
%
f
o
r
e
x
p
licit
asp
ec
t
ex
tr
ac
tio
n
an
d
8
2
.
0
7
%
f
o
r
im
p
licit
o
p
in
io
n
e
x
tr
ac
tio
n
.
T
h
e
av
er
a
g
e
F1
-
s
co
r
es
f
o
r
im
p
licit
r
ev
iew
ca
s
es
o
f
I
AI
O
ar
e
8
2
.
1
6
%
f
o
r
im
p
licit a
s
p
ec
t e
x
tr
ac
tio
n
a
n
d
7
8
.
7
2
% f
o
r
im
p
licit o
p
in
io
n
ex
tr
ac
tio
n
.
4
.
2
.
AC
a
nd
AB
SA
r
esu
lt
s
T
h
e
AC
an
d
AB
SA
r
e
s
u
lts
f
o
r
f
o
u
r
s
en
ten
ce
ty
p
es
ar
e
s
h
o
w
n
in
T
ab
le
8
.
T
ab
le
8
s
h
o
ws
th
at
th
e
AC
an
d
AB
SA
m
eth
o
d
s
ca
n
wo
r
k
p
r
o
p
er
ly
in
ex
p
licit
a
n
d
im
p
l
icit
s
en
ten
ce
s
.
T
h
e
av
er
a
g
e
F1
-
s
co
r
es
o
f
AC
f
o
r
f
o
u
r
s
en
ten
ce
t
y
p
es
ar
e
9
8
.
2
3
%
f
o
r
E
AE
O,
9
5
.
3
6
%
f
o
r
I
AE
O,
9
3
.
7
3
%
f
o
r
E
AI
O,
a
n
d
8
7
.
9
5
%
f
o
r
I
A
I
O.
T
h
en
,
t
h
e
av
e
r
ag
e
F1
-
s
co
r
es
o
f
AB
SA
f
o
r
f
o
u
r
s
en
ten
ce
t
y
p
es
ar
e
9
7
.
9
5
%
f
o
r
E
AE
O,
9
5
.
3
9
%
f
o
r
I
AE
O,
9
1
.
2
9
%
f
o
r
E
AI
O,
an
d
8
5
.
0
4
%
f
o
r
I
AI
O.
T
h
ese
r
esu
lts
s
h
o
w
th
at
tex
t
ex
tr
ac
tio
n
ca
n
h
elp
th
e
AC
m
eth
o
d
p
r
o
p
er
l
y
class
if
y
ea
c
h
r
ev
iew
'
s
asp
ec
t
ca
teg
o
r
y
.
M
o
r
eo
v
er
,
th
e
te
x
t
ex
tr
ac
tio
n
ca
n
h
elp
th
e
AB
SA
m
eth
o
d
p
r
o
p
er
l
y
class
if
y
ea
ch
r
e
v
iew's sen
tim
en
t p
o
lar
ity
.
T
h
e
co
m
p
ar
is
o
n
o
f
AB
SA
ca
s
es
wo
r
k
ed
o
n
f
r
o
m
s
ev
e
r
al
AB
SA
r
esear
ch
es
is
s
h
o
wn
i
n
T
ab
le
9
.
T
ab
le
9
s
h
o
ws
th
at
th
e
p
r
o
p
o
s
ed
m
eth
o
d
ca
n
wo
r
k
b
etter
t
h
an
p
r
ev
i
o
u
s
r
esear
ch
.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
ca
n
wo
r
k
o
n
s
en
ten
ce
ca
s
es
o
f
E
AE
O,
I
AE
O,
E
AI
O,
an
d
I
AI
O
with
f
o
u
r
p
o
s
itiv
e,
p
o
s
itiv
e,
n
eg
ativ
e,
an
d
v
e
r
y
n
eg
ativ
e
s
en
tim
en
t
class
es.
Me
an
wh
ile,
m
an
y
p
r
ev
io
u
s
r
esear
ch
es
s
till
wo
r
k
ed
o
n
s
en
ten
ce
ca
s
es
o
f
E
AE
O
an
d
I
AE
O
with
p
o
s
itiv
e
an
d
n
eg
ativ
e
s
en
tim
en
t
class
es.
Pre
v
io
u
s
r
esear
ch
also
wo
r
k
ed
o
n
th
e
s
en
te
n
ce
ca
s
e
o
f
E
AI
O
b
u
t
n
o
t
th
e
s
en
ten
ce
ca
s
e
o
f
I
AI
O.
Ho
wev
er
,
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
s
till
wo
r
k
s
o
n
r
estau
r
an
t
d
o
m
ain
r
ev
iews
an
d
d
o
es
n
ot
tr
y
o
th
e
r
d
o
m
ain
r
ev
iews.
T
h
e
r
esu
lts
s
till
h
av
e
f
alse
p
o
s
itiv
es
an
d
n
eg
ativ
es
b
ec
au
s
e
th
e
p
r
o
p
o
s
ed
m
eth
o
d
ca
n
n
o
t
ex
p
an
d
th
e
s
y
n
o
n
y
m
o
f
s
o
m
e
im
p
licit
ter
m
s
o
f
asp
ec
t
a
n
d
o
p
i
n
io
n
th
r
o
u
g
h
W
ik
ip
ed
ia
s
o
u
r
ce
p
ag
es.
Fo
r
i
n
s
tan
ce
,
th
e
r
ev
iew
to
ld
u
s
to
s
it a
n
ywh
ere,
a
n
d
w
h
en
w
e
s
a
t
,
h
e
s
a
id
th
a
t it
w
a
s
r
eser
ve
d
a
n
d
to
ld
u
s
to
mo
ve
,
wh
ich
th
e
im
p
licit
asp
ec
t
ter
m
s
h
o
u
ld
b
e
he
as
th
e
v
ar
iab
l
e
o
f
s
er
v
ice
asp
ec
t
ca
teg
o
r
y
,
a
n
d
th
e
im
p
licit
o
p
i
n
io
n
s
h
o
u
l
d
b
e
to
ld
u
s
to
s
it
a
n
ywh
ere,
h
e
s
a
id
it
w
a
s
r
ese
r
ve
d
,
to
ld
u
s
to
mo
ve
b
u
t th
e
p
r
o
p
o
s
ed
m
eth
o
d
s
till
ca
n
n
o
t
ex
tr
ac
t it
ac
c
u
r
ately
.
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.
14
,
No
.
6
,
Dec
em
b
e
r
20
24
:
6
7
1
1
-
6
7
2
2
6720
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
ca
n
well
tak
e
th
e
asp
ec
t
an
d
o
p
in
i
o
n
ter
m
s
f
o
r
th
e
ex
p
licit
a
n
d
im
p
licit
r
ev
iews
in
s
im
p
le,
co
m
p
o
u
n
d
,
co
m
p
lex
,
an
d
c
o
m
p
o
u
n
d
-
co
m
p
lex
s
en
ten
ce
s
.
T
h
e
r
esu
lts
o
f
th
e
asp
ec
t
an
d
o
p
in
io
n
te
r
m
s
ca
r
r
ied
o
u
t
b
y
th
e
p
r
o
p
o
s
ed
m
eth
o
d
wer
e
p
r
o
ce
s
s
ed
to
d
eter
m
i
n
e
th
e
a
s
p
ec
t
ca
teg
o
r
y
an
d
s
en
tim
en
t
p
o
lar
ity
ac
cu
r
ately
.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
s
u
cc
ess
f
u
lly
class
if
ied
th
e
o
p
in
i
o
n
ter
m
s
in
to
f
o
u
r
s
en
tim
en
t
s
tr
en
g
th
class
es
:
v
e
r
y
p
o
s
itiv
e,
p
o
s
itiv
e,
n
eg
ativ
e
,
an
d
v
er
y
n
e
g
ativ
e.
I
d
e
n
tify
i
n
g
p
air
s
o
f
asp
ec
ts
an
d
o
p
in
io
n
ter
m
s
th
at
ar
e
e
x
p
licit
an
d
im
p
licit
is
im
p
o
r
ta
n
t
b
ec
au
s
e
a
n
im
p
licit
o
p
in
io
n
ter
m
m
ay
h
av
e
a
tar
g
et
asp
ec
t
ca
teg
o
r
y
th
at
is
d
if
f
er
en
t
f
r
o
m
th
e
ex
is
tin
g
e
x
p
licit
asp
ec
t
ter
m
a
n
d
h
as
a
s
e
q
u
en
ce
d
is
tan
ce
th
at
is
clo
s
er
th
an
th
e
ac
tu
al
tar
g
et
asp
ec
t
ter
m
d
is
tan
ce
.
C
las
s
if
y
in
g
f
o
u
r
s
en
tim
en
t
class
es
in
to
f
o
u
r
asp
ec
t
ca
teg
o
r
ies
ca
n
also
p
r
o
v
id
e
m
o
r
e
co
n
v
in
cin
g
in
p
u
t
f
o
r
u
s
er
s
in
d
eter
m
in
in
g
t
h
e
d
esire
d
r
estau
r
an
t
ch
o
ice.
Ho
wev
er
,
f
u
r
th
er
r
esear
ch
s
till
n
ee
d
s
to
b
e
ca
r
r
ied
o
u
t
to
r
e
s
o
lv
e
th
e
p
r
o
b
lem
o
f
im
p
licit
o
p
in
io
n
ex
tr
ac
tio
n
,
wh
ich
ca
n
n
o
t
b
e
r
eso
lv
ed
in
t
h
is
r
esear
ch
.
I
n
ad
d
itio
n
,
we
h
o
p
e
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
an
d
r
esear
ch
r
esu
lts
ca
n
b
e
d
ev
elo
p
ed
a
n
d
u
s
ed
f
o
r
v
ar
io
u
s
d
o
m
ain
s
,
n
o
t ju
s
t r
estau
r
a
n
ts
.
T
ab
le
9
.
C
o
m
p
a
r
is
o
n
o
f
AB
S
A
ca
s
es wo
r
k
ed
o
n
R
e
se
a
r
c
h
e
s
R
e
v
i
e
w
c
a
ses
Ex
p
l
i
c
i
t
I
mp
l
i
c
i
t
EA
EO
I
A
EO
EA
I
O
I
A
I
O
A
sp
e
c
t
O
p
i
n
i
o
n
A
sp
e
c
t
O
p
i
n
i
o
n
A
sp
e
c
t
O
p
i
n
i
o
n
A
sp
e
c
t
O
p
i
n
i
o
n
C
a
d
i
l
h
a
c
[
2
]
✓
✓
✓
✓
✓
✓
╳
╳
S
c
h
o
u
t
e
n
[
2
4
]
✓
✓
✓
✓
✓
✓
╳
╳
P
o
r
i
a
[
2
5
]
✓
✓
✓
✓
✓
✓
╳
╳
C
h
e
n
[
3
]
✓
✓
✓
✓
✓
✓
╳
╳
Tu
b
i
s
h
a
t
[
2
6
]
✓
✓
✓
✓
╳
╳
╳
╳
P
r
o
p
o
se
d
m
e
t
h
o
d
✓
✓
✓
✓
✓
✓
✓
✓
4
.
3
.
E
v
a
lua
t
io
n
r
esu
lt
I
n
p
r
ev
io
u
s
AB
SA
s
tu
d
ies
f
o
r
r
estau
r
an
t
r
ev
iew
[
2
2
]
,
th
e
b
est
F1
-
s
co
r
e
is
0
.
7
9
4
0
,
a
n
d
th
e
b
est
ac
cu
r
ac
y
is
0
.
7
9
9
4
.
W
e
co
m
p
ar
ed
o
u
r
r
esu
lts
with
p
r
ev
i
o
u
s
s
tu
d
ies
th
at
u
s
ed
th
e
s
am
e
d
a
taset
[
2
2
]
.
T
ab
le
1
0
s
h
o
ws
th
e
co
m
p
ar
is
o
n
o
f
th
e
AB
SA
ev
alu
atio
n
r
esu
lts
i
n
r
estau
r
an
t
r
ev
iew
[
2
2
]
.
T
h
e
p
r
o
p
o
s
ed
AB
SA
m
eth
o
d
was p
r
o
v
e
n
to
wo
r
k
b
etter
an
d
m
o
r
e
ac
cu
r
ately
th
an
th
e
p
r
ev
io
u
s
s
tu
d
ies
[
2
]
,
[
3
]
,
[
2
2
]
,
[
2
4
]
,
[
2
5
]
wit
h
o
b
tain
ed
a
v
er
ag
e
F1
-
s
co
r
e
to
t
al
o
f
0
.
9
2
4
2
an
d
av
er
a
g
e
ac
cu
r
ac
y
to
tal
o
f
0
.
9
2
3
3
.
T
ab
le
10
.
C
o
m
p
ar
is
o
n
o
f
AB
SA e
v
alu
atio
n
r
esu
lt in
r
estau
r
a
n
t r
ev
iew
[
2
2
]
R
e
se
a
r
c
h
e
s
R
e
s
u
l
t
s
P
R
F
A
P
o
n
t
i
k
i
[
2
2
]
N
/
A
N
/
A
0
.
6
2
6
8
N
/
A
C
a
d
i
l
h
a
c
[
2
]
0
.
7
6
9
2
0
.
7
7
3
3
0
.
7
7
1
2
N
/
A
S
c
h
o
u
t
e
n
[
2
4
]
N
/
A
N
/
A
0
.
7
9
4
0
N
/
A
P
o
r
i
a
[
2
5
]
0
.
8
5
2
1
0
.
8
8
1
5
N
/
A
N
/
A
C
h
e
n
[
3
]
N
/
A
N
/
A
N
/
A
0
.
7
9
9
4
P
r
o
p
o
se
d
m
e
t
h
o
d
0
.
9
2
7
2
0
.
9
2
1
4
0
.
9
2
4
2
0
.
9
2
3
3
5.
CO
NCLU
SI
O
N
T
h
is
p
ap
er
p
r
o
p
o
s
ed
an
AB
SA
m
eth
o
d
:
n
atu
r
al
lan
g
u
ag
e
u
n
d
er
s
tan
d
i
n
g
f
o
r
im
p
licit
r
ev
iews.
W
e
b
u
ilt
a
s
et
o
f
r
u
les
as
th
e
NL
U
s
y
s
tem
th
at
ca
n
s
o
lv
e
th
e
p
r
o
b
lem
o
f
ex
tr
ac
tin
g
im
p
licit
r
ev
iews
u
s
in
g
tex
t
ex
tr
ac
tio
n
th
at
p
ay
s
atten
tio
n
to
d
if
f
er
en
t
ty
p
es
o
f
s
en
ten
ce
s
an
d
wo
r
d
s
.
T
h
e
tex
t
ex
tr
ac
t
io
n
ca
n
ac
cu
r
ately
r
etr
iev
e
p
ai
r
in
g
ter
m
s
o
f
asp
e
ct
an
d
o
p
in
i
o
n
,
wh
ich
ar
e
e
x
p
licit
an
d
im
p
licit.
T
h
e
tex
t
ex
tr
ac
tio
n
ca
n
also
s
ig
n
if
ican
tly
in
cr
ea
s
e
t
h
e
p
er
f
o
r
m
an
ce
s
o
f
th
e
AC
m
eth
o
d
with
an
a
v
er
ag
e
F1
-
s
co
r
e
o
f
0
.
9
3
8
2
an
d
th
e
AB
SA
m
eth
o
d
with
a
n
av
e
r
ag
e
F1
-
s
co
r
e
o
f
0
.
9
2
4
2
.
W
e
co
m
p
ar
ed
th
e
r
esu
lts
o
f
t
h
e
o
b
tain
ed
A
B
SA
ev
alu
atio
n
in
r
estau
r
an
t
r
ev
iews
with
th
o
s
e
o
f
p
r
ev
io
u
s
s
tu
d
ies.
T
h
e
o
b
t
ain
ed
F1
-
s
co
r
e
r
esu
lt
o
f
0
.
9
2
4
2
s
h
o
ws
th
at
th
e
p
r
o
p
o
s
ed
m
eth
o
d
ca
n
wo
r
k
b
e
tter
th
an
p
r
e
v
io
u
s
m
eth
o
d
s
.
Un
d
er
s
tan
d
in
g
th
e
im
p
licit
r
e
v
iews
is
im
p
o
r
tan
t
to
in
cr
ea
s
e
AC
an
d
A
B
SA
p
er
f
o
r
m
an
ce
s
.
W
e
h
o
p
e
th
is
p
r
o
p
o
s
ed
m
eth
o
d
ca
n
b
e
im
p
lem
en
ted
o
n
o
th
er
r
e
v
iew
d
atasets
in
f
u
tu
r
e
wo
r
k
.
T
h
e
d
ev
elo
p
m
en
t
o
f
d
atasets
f
o
r
s
en
ten
ce
ty
p
es
co
n
tain
in
g
v
ar
io
u
s
c
h
ar
ac
ter
is
tics
o
f
im
p
licit
o
p
in
io
n
s
is
s
till
n
ee
d
ed
.
Ap
ar
t
f
r
o
m
th
at,
we
also
h
o
p
e
th
at
th
is
p
r
o
p
o
s
ed
m
eth
o
d
ca
n
b
e
f
u
r
th
er
im
p
r
o
v
e
d
to
g
et
ev
e
n
b
etter
r
esu
lts
,
esp
ec
ially
in
th
e
ca
s
e
o
f
o
th
e
r
s
en
ten
ce
s
th
a
t w
e
h
av
e
n
o
t b
ee
n
a
b
le
to
d
o
well.
Evaluation Warning : The document was created with Spire.PDF for Python.