I
nd
o
ne
s
ia
n J
o
urna
l o
f
E
lect
rica
l En
g
ineering
a
nd
Co
m
pu
t
er
Science
Vo
l.
3
7
,
No
.
3
,
Ma
r
ch
20
2
5
,
p
p
.
1
6
6
1
~
1
671
I
SS
N:
2
502
-
4
7
52
,
DOI
: 1
0
.
1
1
5
9
1
/ijee
cs
.v
3
7
.
i
3
.
pp
1
6
6
1
-
1
6
7
1
1661
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ee
cs
.
ia
esco
r
e.
co
m
A
n
o
v
el
da
tas
et
a
nd pa
rt
-
of
-
spee
ch
t
a
g
g
ing
appro
a
ch f
o
r
enha
ncing
sentim
ent
a
na
ly
sis
in
K
a
nna
da
Su
nil
M
ug
a
lih
a
lli
E
s
hwa
ra
pp
a
1,
3
,
Vina
y
Sh
iv
a
s
ub
ra
m
a
ny
a
n
2
1
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
E
n
g
i
n
e
e
r
i
n
g
,
P
ES
C
o
l
l
e
g
e
o
f
E
n
g
i
n
e
e
r
i
n
g
,
V
i
s
v
e
sv
a
r
a
y
a
Te
c
h
n
o
l
o
g
i
c
a
l
U
n
i
v
e
r
s
i
t
y
,
B
e
l
a
g
a
v
i
,
I
n
d
i
a
2
D
e
p
a
r
t
me
n
t
o
f
I
n
f
o
r
mat
i
o
n
S
c
i
e
n
c
e
a
n
d
E
n
g
i
n
e
e
r
i
n
g
,
P
ES C
o
l
l
e
g
e
o
f
E
n
g
i
n
e
e
r
i
n
g
,
V
i
s
v
e
s
v
a
r
a
y
a
T
e
c
h
n
o
l
o
g
i
c
a
l
U
n
i
v
e
r
si
t
y
,
B
e
l
a
g
a
v
i
,
I
n
d
i
a
3
S
o
f
t
w
a
r
e
E
n
g
i
n
e
e
r
,
W
i
p
r
o
L
i
mi
t
e
d
,
B
a
n
g
a
l
o
r
e
,
I
n
d
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Ap
r
16
,
2
0
2
4
R
ev
is
ed
Sep
11
,
2
0
2
4
Acc
ep
ted
Oct
7
,
2
0
2
4
Th
e
p
r
o
b
lem
a
d
d
re
ss
e
d
in
th
is
re
se
a
rc
h
is
th
e
li
m
it
e
d
a
v
a
il
a
b
il
it
y
o
f
lab
e
l
led
d
a
tas
e
ts
a
n
d
e
ffe
c
ti
v
e
se
n
ti
m
e
n
t
a
n
a
ly
sis
t
o
o
ls
f
o
r
t
h
e
Ka
n
n
a
d
a
lan
g
u
a
g
e
.
Ex
isti
n
g
c
h
a
ll
e
n
g
e
s
in
c
l
u
d
e
l
in
g
u
isti
c
v
a
riatio
n
s,
c
u
lt
u
ra
l
d
iv
e
rsiti
e
s,
a
n
d
th
e
a
b
se
n
c
e
o
f
c
o
m
p
re
h
e
n
si
v
e
d
a
ta
se
ts
d
e
sig
n
e
d
s
p
e
c
ifi
c
a
ll
y
fo
r
se
n
ti
m
e
n
t
a
n
a
ly
sis
in
Ka
n
n
a
d
a
.
Th
is
re
se
a
rc
h
a
ims
to
e
n
h
a
n
c
e
se
n
ti
m
e
n
t
a
n
a
ly
sis
c
a
p
a
b
il
it
ies
fo
r
th
e
Ka
n
n
a
d
a
lan
g
u
a
g
e
,
a
d
d
re
ss
in
g
c
h
a
ll
e
n
g
e
s
p
o
se
d
b
y
li
n
g
u
isti
c
v
a
riatio
n
s
a
n
d
li
m
it
e
d
l
a
b
e
ll
e
d
d
a
tas
e
ts.
A
n
o
v
e
l
Ka
n
n
a
d
a
d
a
tas
e
t
d
e
riv
e
d
fr
o
m
S
e
m
Ev
a
l
2
0
1
4
t
a
sk
4
wa
s
c
re
a
ted
u
sin
g
a
c
o
n
v
e
rsio
n
p
ro
c
e
ss
.
Th
e
d
a
tas
e
t
wa
s
p
ro
c
e
ss
e
d
u
si
n
g
p
a
rt
-
of
-
s
p
e
e
c
h
tag
g
in
g
,
a
n
d
a
s
p
e
c
ialize
d
m
o
d
e
l
c
a
ll
e
d
K
-
BER
T
(Ka
n
n
a
d
a
b
i
d
irec
ti
o
n
a
l
e
n
c
o
d
e
r
re
p
re
se
n
tat
io
n
s
fr
o
m
tran
sfo
rm
e
rs
)
wa
s
in
tro
d
u
c
e
d
a
n
d
imp
lem
e
n
ted
u
sin
g
P
y
th
o
n
with
in
th
e
An
a
c
o
n
d
a
e
n
v
ir
o
n
m
e
n
t.
P
e
rf
o
rm
a
n
c
e
e
v
a
lu
a
ti
o
n
re
su
lt
s
sh
o
wc
a
se
d
K
-
BERT
'
s
su
p
e
rio
rit
y
o
v
e
r
trad
i
ti
o
n
a
l
m
a
c
h
in
e
lea
rn
in
g
(
ML
)
a
lg
o
rit
h
m
s
a
n
d
t
h
e
BERT
m
o
d
e
l,
a
c
h
ie
v
in
g
a
n
a
c
c
u
ra
c
y
o
f
0
.
9
8
,
p
re
c
isio
n
o
f
0
.
9
7
,
re
c
a
ll
o
f
0
.
9
7
,
a
n
d
F
-
s
c
o
re
o
f
0
.
9
8
in
se
n
ti
m
e
n
t
c
las
sifica
ti
o
n
f
o
r
Ka
n
n
a
d
a
tex
t
d
a
ta.
T
h
is
wo
r
k
c
o
n
tri
b
u
tes
a
u
n
iq
u
e
Ka
n
n
a
d
a
d
a
tas
e
t,
in
tr
o
d
u
c
e
s
t
h
e
K
-
BERT
m
o
d
e
l
sp
e
c
ifi
c
a
ll
y
d
e
sig
n
e
d
f
o
r
Ka
n
n
a
d
a
se
n
ti
m
e
n
t
a
n
a
ly
sis,
a
n
d
e
m
p
h
a
siz
e
s
th
e
imp
o
rtan
c
e
o
f
c
o
ll
a
b
o
ra
ti
v
e
e
ffo
rts
in
a
d
v
a
n
c
i
n
g
n
a
tu
ra
l
lan
g
u
a
g
e
p
r
o
c
e
ss
in
g
(
NLP
)
re
se
a
rc
h
fo
r
m
u
lt
il
i
n
g
u
a
l
e
n
v
iro
n
m
e
n
ts.
K
ey
w
o
r
d
s
:
Kan
n
ad
a
K
-
B
E
R
T
m
o
d
el
Natu
r
al
lan
g
u
ag
e
p
r
o
ce
s
s
in
g
Sem
E
v
al
2
0
1
4
t
ask
4
Sen
tim
en
t a
n
aly
s
is
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
:
Su
n
il M
u
g
alih
alli E
s
h
war
ap
p
a
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
an
d
E
n
g
in
ee
r
in
g
,
PES Co
l
leg
e
o
f
E
n
g
in
ee
r
in
g
Vis
v
esv
ar
ay
a
T
ec
h
n
o
l
o
g
ical
Un
iv
er
s
ity
B
elag
av
i
-
5
9
0
0
1
8
,
I
n
d
ia
E
m
ail: su
n
i.m
g
h
alli@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
I
n
d
ia
is
k
n
o
wn
f
o
r
its
r
ich
cu
ltu
r
al
d
iv
er
s
ity
,
an
d
th
is
d
iv
er
s
ity
ex
ten
d
s
to
th
e
lan
g
u
ag
es
s
p
o
k
en
ac
r
o
s
s
th
e
co
u
n
tr
y
.
T
h
e
r
e
ar
e
m
o
r
e
th
an
1
,
6
0
0
d
if
f
e
r
en
t
d
ialec
ts
an
d
lan
g
u
a
g
es
s
p
o
k
en
in
I
n
d
ia,
ea
c
h
co
n
tr
ib
u
tin
g
to
its
u
n
iq
u
e
h
e
r
itag
e.
E
v
er
y
s
tate
in
I
n
d
i
a
h
as
its
o
wn
lan
g
u
ag
e,
ad
d
i
n
g
to
th
e
co
u
n
tr
y
'
s
lin
g
u
is
tic
d
iv
er
s
ity
.
T
h
is
d
iv
e
r
s
ity
is
a
r
ef
lectio
n
o
f
I
n
d
ia'
s
cu
ltu
r
al,
h
is
to
r
ical,
a
n
d
g
eo
g
r
ap
h
ical
r
ich
n
ess
.
W
h
en
tr
av
elin
g
th
r
o
u
g
h
s
tates
in
I
n
d
ia,
it
ca
n
b
e
n
o
ticed
th
at
ea
ch
s
tate
p
r
ed
o
m
in
an
tly
s
p
ea
k
s
a
d
if
f
er
en
t
lan
g
u
ag
e.
Fo
r
ex
am
p
le
,
Hin
d
i
is
wid
ely
s
p
o
k
en
i
n
s
tates
lik
e
Uttar
Pra
d
esh
an
d
B
ih
ar
,
wh
ile
T
am
il
is
th
e
p
r
im
ar
y
la
n
g
u
a
g
e
in
T
a
m
il
Nad
u
,
T
elu
g
u
in
An
d
h
r
a
Pra
d
esh
an
d
Ka
n
n
ad
a
i
n
Kar
n
at
ak
a.
T
h
is
lin
g
u
is
tic
v
ar
iatio
n
p
o
s
es
a
ch
allen
g
e
wh
en
an
aly
s
i
n
g
co
m
m
en
ts
,
r
ev
i
ews,
an
d
s
en
tim
en
ts
ex
p
r
ess
e
d
b
y
p
eo
p
le
in
th
eir
n
ativ
e
lan
g
u
a
g
es
[
1
]
.
Peo
p
le
in
I
n
d
ia
ex
p
r
ess
th
eir
o
p
in
i
o
n
s
,
g
iv
e
f
ee
d
b
ac
k
,
a
n
d
wr
it
e
r
ev
iews
in
th
eir
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
In
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
3
7
,
No
.
3
,
Ma
r
ch
20
2
5
:
1
661
-
1
6
7
1
1662
p
r
ef
er
r
e
d
la
n
g
u
a
g
es
in
v
ar
io
u
s
s
o
cial
m
ed
ia
p
latf
o
r
m
s
,
a
n
d
v
id
eo
s
h
ar
i
n
g
we
b
s
ites
[
2
]
.
T
h
i
s
d
iv
er
s
ity
m
a
k
es
it
co
m
p
lex
to
a
n
aly
s
e
s
en
tim
en
t
s
ac
cu
r
ately
.
Fo
r
ex
a
m
p
le,
a
p
o
s
itiv
e
co
m
m
en
t
in
Kan
n
ad
a
m
ay
h
av
e
d
if
f
er
e
n
t
cu
ltu
r
al
m
ea
n
in
g
s
co
m
p
a
r
ed
to
a
p
o
s
itiv
e
co
m
m
en
t
in
Hi
n
d
i
o
r
T
am
il.
T
h
is
d
iv
e
r
s
ity
ad
d
s
co
m
p
lex
ity
to
s
en
tim
en
t a
n
aly
s
is
,
esp
ec
ially
in
a
m
u
ltil
in
g
u
al
e
n
v
ir
o
n
m
en
t
lik
e
I
n
d
ia.
Sen
tim
en
t
an
aly
s
is
,
also
k
n
o
wn
as
o
p
in
io
n
m
i
n
in
g
,
is
a
m
et
h
o
d
u
s
ed
to
ex
tr
ac
t
s
u
b
jectiv
e
in
f
o
r
m
atio
n
f
r
o
m
tex
t
[
3
]
.
I
t
h
elp
s
u
n
d
er
s
tan
d
s
en
tim
en
ts
,
em
o
tio
n
s
,
attitu
d
es,
an
d
o
p
in
io
n
s
ex
p
r
ess
ed
b
y
in
d
iv
id
u
als.
Natu
r
al
lan
g
u
ag
e
p
r
o
ce
s
s
in
g
(
NL
P)
alg
o
r
ith
m
s
[
4
]
,
[
5
]
a
n
d
m
ac
h
in
e
lear
n
in
g
(
ML
)
tech
n
iq
u
es
[
6
]
,
[
7
]
in
r
ec
en
t
y
ea
r
s
ar
e
wid
ely
u
s
ed
f
o
r
s
en
tim
en
t
a
n
aly
s
is
to
ca
teg
o
r
ize
tex
t
d
ata
in
to
p
o
s
itiv
e,
n
eg
ativ
e,
o
r
n
eu
tr
al
s
en
tim
en
ts
.
Sen
tim
en
t
an
aly
s
is
i
s
cr
u
cial
in
to
d
ay
'
s
d
ig
ital
wo
r
ld
wh
er
e
v
ast
a
m
o
u
n
ts
o
f
tex
t
d
ata
a
r
e
g
e
n
e
r
a
t
e
d
d
a
i
l
y
o
n
s
o
c
i
a
l
m
e
d
i
a
,
e
-
c
o
m
m
e
r
c
e
p
l
a
t
f
o
r
m
s
,
n
e
w
s
w
e
b
s
i
t
es
,
a
n
d
c
u
s
t
o
m
e
r
f
e
e
d
b
a
c
k
f
o
r
u
m
s
[
8
]
.
B
u
s
in
ess
e
s
,
o
r
g
an
izatio
n
s
,
an
d
g
o
v
er
n
m
en
ts
u
s
e
s
en
tim
en
t
an
aly
s
is
to
u
n
d
e
r
s
tan
d
p
u
b
lic
o
p
in
io
n
,
ass
ess
cu
s
to
m
er
s
atis
f
ac
tio
n
,
m
o
n
ito
r
b
r
an
d
p
er
ce
p
tio
n
,
an
d
m
ak
e
d
ata
-
d
r
iv
e
n
d
ec
is
io
n
s
.
Ho
wev
er
,
an
aly
s
in
g
s
en
tim
en
ts
in
d
if
f
e
r
en
t
lan
g
u
a
g
es
is
ch
allen
g
in
g
d
u
e
to
lin
g
u
is
tic
v
ar
iatio
n
s
,
cu
ltu
r
al
d
iv
e
r
s
ities
,
an
d
co
m
p
lex
s
en
ten
ce
s
tr
u
ctu
r
es
[
9
]
.
ML
a
n
d
d
ee
p
lear
n
in
g
(
DL
)
alg
o
r
i
th
m
s
p
lay
a
s
ig
n
if
ican
t
r
o
le
i
n
ad
d
r
ess
in
g
th
ese
ch
allen
g
es
b
y
a
u
to
m
atin
g
s
en
tim
en
t
an
aly
s
is
ac
r
o
s
s
m
u
ltip
le
lan
g
u
ag
es
[
1
0
]
.
T
h
es
e
tech
n
iq
u
es
lear
n
lin
g
u
is
tic
p
atter
n
s
an
d
s
em
an
ti
c
s
tr
u
ctu
r
es to
im
p
r
o
v
e
th
e
ac
c
u
r
ac
y
an
d
ef
f
icien
c
y
o
f
s
en
tim
en
t a
n
aly
s
is
.
Desp
ite
ad
v
an
ce
m
en
ts
,
s
en
tim
en
t
an
aly
s
is
f
o
r
lan
g
u
a
g
es
lik
e
Kan
n
ad
a
f
ac
es
s
p
ec
if
ic
ch
alle
n
g
es.
T
h
e
lack
o
f
lab
elled
d
atasets
co
n
ta
in
in
g
asp
ec
ts
an
d
s
en
tim
en
ts
i
n
Kan
n
a
d
a
h
i
n
d
er
s
th
e
d
e
v
elo
p
m
en
t
o
f
ac
c
u
r
ate
s
en
tim
en
t
an
aly
s
is
to
o
ls
[
1
1
]
.
Ad
d
itio
n
ally
,
lin
g
u
is
tic
im
p
o
r
tan
ce
,
s
en
tim
en
t
e
x
p
r
es
s
io
n
s
,
an
d
cu
ltu
r
al
r
ef
er
en
ce
s
u
n
iq
u
e
to
Kan
n
ad
a
p
o
s
e
d
if
f
icu
lties
d
u
r
in
g
s
en
tim
en
t
an
aly
s
is
[
1
2
]
.
T
o
ad
d
r
ess
th
e
af
o
r
em
en
tio
n
e
d
ch
allen
g
es,
th
is
s
tu
d
y
tak
es
a
p
r
o
ac
tiv
e
ap
p
r
o
ac
h
b
y
in
tr
o
d
u
cin
g
a
n
o
v
el
Kan
n
a
d
a
d
atas
et
d
er
iv
ed
f
r
o
m
th
e
Sem
E
v
al
2
0
1
4
T
ask
4
d
ataset.
T
h
e
d
ataset
is
p
r
ep
ar
ed
b
y
c
o
n
v
er
s
io
n
p
r
o
ce
s
s
wh
ich
in
v
o
lv
es
tr
an
s
latin
g
th
e
Sem
E
v
al
d
ataset
f
r
o
m
E
n
g
lis
h
to
Kan
n
a
d
a.
T
h
is
s
tr
ateg
ic
s
tep
is
tak
en
b
ec
au
s
e
th
e
Se
m
E
v
al
2
0
1
4
T
ask
4
d
ataset
o
f
f
er
s
a
s
u
b
s
tan
tial
n
u
m
b
er
o
f
asp
ec
ts
an
d
s
en
tim
en
ts
,
m
ak
in
g
it
well
-
s
u
ited
f
o
r
ev
alu
atio
n
p
u
r
p
o
s
es.
Ad
d
itio
n
ally
,
th
is
co
n
v
er
s
io
n
s
ig
n
if
ican
tly
r
ed
u
ce
s
th
e
tim
e
an
d
ef
f
o
r
t
r
eq
u
ir
e
d
f
o
r
lab
elin
g
d
ata,
s
en
tim
en
ts
,
an
d
asp
ec
ts
in
Kan
n
ad
a.
Su
b
s
eq
u
en
tly
,
th
e
f
ea
tu
r
es
ex
tr
ac
t
ed
f
r
o
m
th
is
n
ewly
cr
ea
ted
K
an
n
ad
a
d
ataset
ar
e
p
r
o
ce
s
s
ed
u
s
in
g
th
e
p
ar
t
-
of
-
s
p
ee
ch
(
Po
S)
t
ag
g
in
g
m
eth
o
d
.
Po
S
t
ag
g
in
g
is
u
tili
ze
d
to
id
en
t
if
y
an
d
ca
teg
o
r
iz
e
th
e
g
r
am
m
atica
l
co
m
p
o
n
en
ts
o
f
th
e
tex
t,
wh
ich
is
cr
u
ci
al
f
o
r
ac
cu
r
ate
f
ea
tu
r
e
ex
tr
a
ctio
n
in
Kan
n
ad
a.
Mo
r
eo
v
er
,
a
n
o
v
el
m
o
d
el
s
p
ec
if
ically
d
esig
n
ed
f
o
r
Kan
n
ad
a,
ter
m
ed
K
-
B
E
R
T
(
Kan
n
ad
aB
E
R
T
)
,
is
in
tr
o
d
u
ce
d
in
th
is
s
tu
d
y
.
K
-
B
E
R
T
is
d
es
ig
n
ed
to
ef
f
ec
tiv
ely
class
if
y
th
e
ex
tr
ac
ted
f
ea
t
u
r
es
f
r
o
m
th
e
Kan
n
ad
a
d
ataset.
T
h
is
m
o
d
el
lev
er
ag
es th
e
ad
v
a
n
ce
m
en
ts
in
B
E
R
T
(
b
id
ir
ec
tio
n
al
en
co
d
er
r
e
p
r
esen
tatio
n
s
f
r
o
m
tr
an
s
f
o
r
m
e
r
s
)
,
a
s
tate
-
of
-
th
e
-
ar
t
lan
g
u
a
g
e
m
o
d
el
in
NL
P,
to
en
h
a
n
ce
th
e
ac
cu
r
ac
y
an
d
p
e
r
f
o
r
m
an
ce
o
f
s
en
tim
e
n
t
an
al
y
s
is
in
Kan
n
ad
a.
T
h
e
m
an
u
s
cr
ip
t
is
s
tr
u
ctu
r
ed
as
f
o
llo
ws
to
p
r
o
v
id
e
a
clea
r
an
d
s
y
s
tem
atic
p
r
esen
tati
o
n
o
f
th
e
r
esear
ch
f
in
d
in
g
s
.
Sectio
n
2
d
elv
es
in
to
t
h
e
liter
atu
r
e
s
u
r
v
ey
,
wh
e
r
e
e
x
is
tin
g
s
tu
d
ies
an
d
m
et
h
o
d
o
lo
g
ies
r
elate
d
to
s
en
tim
en
t
an
al
y
s
is
an
d
o
p
in
io
n
m
in
i
n
g
in
Kan
n
ad
a
ar
e
d
is
cu
s
s
ed
.
Mo
v
in
g
o
n
to
s
ec
tio
n
3
,
th
e
p
r
o
ce
s
s
o
f
d
ataset
p
r
ep
ar
atio
n
,
p
r
ep
r
o
ce
s
s
in
g
tec
h
n
iq
u
es,
f
ea
tu
r
e
ex
tr
ac
tio
n
m
eth
o
d
s
,
an
d
t
h
e
d
e
v
elo
p
m
e
n
t
o
f
th
e
K
-
B
E
R
T
class
if
ier
ar
e
elab
o
r
ated
u
p
o
n
.
Sectio
n
4
is
d
e
d
icate
d
to
a
n
aly
s
in
g
a
n
d
c
o
m
p
ar
in
g
t
h
e
f
ea
t
u
r
es
ex
tr
ac
ted
f
r
o
m
th
e
Po
S tag
g
i
n
g
m
eth
o
d
an
d
t
h
e
K
-
B
E
R
T
class
if
ier
.
Var
io
u
s
class
if
ier
s
ar
e
em
p
lo
y
ed
a
n
d
th
ei
r
p
er
f
o
r
m
an
ce
is
ev
alu
ated
b
ase
d
o
n
th
e
ex
tr
ac
te
d
f
ea
tu
r
es,
p
r
o
v
id
in
g
in
s
ig
h
ts
in
to
th
e
ef
f
ec
t
iv
en
ess
o
f
d
if
f
er
en
t
class
i
f
icatio
n
tech
n
iq
u
es
f
o
r
s
en
tim
en
t
an
aly
s
is
in
Kan
n
a
d
a.
L
astl
y
,
s
ec
tio
n
5
e
n
ca
p
s
u
late
s
th
e
co
n
clu
s
io
n
o
f
th
e
wo
r
k
,
s
u
m
m
ar
izi
n
g
th
e
k
ey
f
in
d
i
n
g
s
,
co
n
tr
i
b
u
tio
n
s
,
an
d
im
p
licatio
n
s
o
f
th
e
s
tu
d
y
.
T
h
is
s
ec
tio
n
also
d
is
cu
s
s
es
p
o
ten
tial
ar
ea
s
f
o
r
f
u
tu
r
e
r
esear
ch
an
d
h
ig
h
lig
h
t
s
th
e
s
ig
n
if
ican
ce
o
f
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
in
ad
v
an
cin
g
s
en
tim
en
t a
n
aly
s
is
ca
p
ab
ilit
ies f
o
r
Kan
n
ad
a
lan
g
u
ag
e
p
r
o
ce
s
s
in
g
.
2.
L
I
T
E
R
AT
U
RE
SU
RVE
Y
T
h
e
liter
atu
r
e
s
u
r
v
ey
en
ca
p
s
u
lates
a
m
u
ltifa
ce
ted
ex
p
lo
r
ati
o
n
in
to
s
en
tim
en
t
a
n
aly
s
is
an
d
o
p
i
n
io
n
m
in
in
g
,
with
a
s
p
ec
if
ic
f
o
cu
s
o
n
lan
g
u
ag
es su
ch
as Ka
n
n
a
d
a
,
T
am
il,
E
n
g
lis
h
,
a
n
d
th
eir
c
o
d
e
-
m
ix
ed
v
a
r
iatio
n
s
.
E
ac
h
o
f
th
e
r
ef
e
r
en
ce
d
s
tu
d
ie
s
ad
d
s
a
u
n
iq
u
e
p
er
s
p
ec
tiv
e
b
y
in
tr
o
d
u
cin
g
d
is
tin
ct
m
eth
o
d
o
lo
g
ies,
tech
n
iq
u
es,
an
d
f
in
d
in
g
s
,
th
er
eb
y
en
r
ic
h
in
g
th
e
b
r
o
a
d
er
d
is
co
u
r
s
e
o
f
c
o
m
p
u
tatio
n
al
lin
g
u
is
tics
an
d
p
av
in
g
th
e
way
f
o
r
ad
v
an
ce
m
e
n
ts
in
s
en
tim
en
t
an
aly
s
is
r
esear
ch
.
B
eg
in
n
in
g
with
[
1
3
]
,
r
esear
ch
er
s
in
tr
o
d
u
ce
d
an
in
n
o
v
ativ
e
h
y
b
r
id
a
p
p
r
o
ac
h
ca
lled
SAEK
C
S,
wh
ich
u
tili
ze
s
s
tate
-
of
-
th
e
-
ar
t
DL
ap
p
r
o
ac
h
es
wh
ich
in
clu
d
e
bi
-
d
ir
ec
tio
n
al
lo
n
g
-
s
h
o
r
t
ter
m
-
m
em
o
r
y
(
L
S
T
M)
an
d
co
n
v
o
lu
tio
n
al
-
n
eu
r
a
l
-
n
etwo
r
k
(
C
NN)
f
o
r
th
e
p
u
r
p
o
s
e
o
f
an
aly
zin
g
s
en
tim
en
ts
o
n
E
n
g
lis
h
-
Kan
n
ad
a
co
d
ed
-
s
witch
ed
tex
t
d
ataset.
T
h
e
ex
p
er
im
en
ts
p
r
ese
n
ted
in
th
e
s
tu
d
y
d
em
o
n
s
tr
ated
a
n
o
ta
b
le
ac
cu
r
a
cy
s
co
r
e
o
f
7
7
.
6
%
alo
n
g
with
an
o
v
e
r
all
co
v
e
r
ag
e
-
r
ate
o
f
6
9
.
6
%.
T
h
ese
r
esu
lts
h
ig
h
lig
h
t
t
h
e
ef
f
ec
tiv
en
ess
o
f
DL
tech
n
iq
u
es
in
ef
f
ec
tiv
el
y
p
r
o
ce
s
s
in
g
co
d
e
-
s
witch
ed
lin
g
u
is
tic
in
f
o
r
m
atio
n
.
Mo
v
in
g
o
n
to
[
1
4
]
,
r
esear
ch
e
r
s
s
et
o
u
t
to
d
is
co
v
er
an
d
class
if
y
d
if
f
er
e
n
t
p
o
in
ts
o
f
v
iew
co
n
v
ey
ed
in
Ka
n
n
ad
a
tex
t.
T
h
e
r
esear
ch
er
s
u
tili
ze
d
a
r
an
g
e
o
f
m
eth
o
d
s
,
n
am
ely
d
ec
is
io
n
tr
ee
(
DT
)
,
Naiv
e
B
ay
e
s
(
NB
)
,
an
d
n
eg
ato
r
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
A
n
o
ve
l d
a
ta
s
et
a
n
d
p
a
r
t
-
of
-
s
p
ee
ch
ta
g
g
in
g
a
p
p
r
o
a
ch
fo
r
…
(
S
u
n
il Mu
g
a
lih
a
lli E
s
h
w
a
r
a
p
p
a
)
1663
ap
p
r
o
ac
h
.
T
h
ese
m
eth
o
d
o
l
o
g
i
es
y
ield
ed
s
ig
n
if
ican
t
ac
cu
r
ac
y
lev
els
o
f
8
5
%,
6
5
%
,
an
d
5
3
%
r
esp
ec
tiv
ely
.
T
h
is
r
esear
ch
h
ig
h
lig
h
t
s
th
e
co
m
p
l
ex
p
r
o
ce
d
u
r
e
o
f
an
aly
s
in
g
o
p
in
io
n
s
in
s
ettin
g
s
with
a
wid
e
v
ar
iety
o
f
lan
g
u
ag
es.
A
t
h
o
r
o
u
g
h
e
v
a
l
u
a
t
i
o
n
o
f
K
a
n
n
a
d
a
-
l
a
n
g
u
a
g
e
I
M
D
B
r
e
v
i
e
w
s
o
b
t
a
i
n
e
d
f
r
o
m
r
e
l
i
a
b
l
e
s
o
u
r
c
e
s
w
a
s
c
a
r
r
i
e
d
o
u
t
i
n
[
1
5
]
.
T
h
e
r
esear
ch
er
s
ac
h
iev
e
d
8
9
%
r
ate
o
f
ac
cu
r
ac
y
b
y
s
u
g
g
esti
n
g
an
en
s
em
b
le
class
if
ier
m
eth
o
d
th
at
u
s
es
v
a
r
io
u
s
v
ec
to
r
izatio
n
alg
o
r
ith
m
s
.
Acc
o
r
d
in
g
to
th
is
wo
r
k
,
f
o
r
h
a
n
d
l
in
g
task
s
o
f
s
en
tim
en
t
an
aly
s
is
ac
r
o
s
s
n
u
m
er
o
u
s
f
i
e
l
d
s
,
r
o
b
u
s
t
c
a
t
e
g
o
r
i
z
a
t
i
o
n
a
p
p
r
o
a
c
h
e
s
a
r
e
c
r
u
c
i
a
l
.
I
n
a
c
o
m
p
a
r
a
b
l
e
m
a
n
n
e
r
t
o
[
1
5
]
,
t
h
e
s
t
u
d
y
c
o
n
d
u
c
t
e
d
b
y
[
1
6
]
ex
p
lo
r
ed
th
e
f
ield
o
f
s
en
tim
e
n
t
ev
alu
atio
n
ac
r
o
s
s
v
ar
io
u
s
lan
g
u
ag
es,
s
u
ch
as
Kan
n
ad
a
,
Hin
d
i
an
d
E
n
g
lis
h
lan
g
u
ag
es.
B
y
em
p
l
o
y
in
g
a
C
NN
co
m
b
in
e
d
with
L
STM
f
r
a
m
ewo
r
k
,
th
e
i
n
v
esti
g
atio
n
s
u
c
ce
s
s
f
u
lly
o
b
tain
ed
o
u
tco
m
es
th
at
o
u
tp
er
f
o
r
m
e
d
estab
lis
h
ed
ap
p
r
o
ac
h
es.
T
h
is
h
ig
h
lig
h
ts
th
e
p
r
o
m
is
in
g
ca
p
ab
ilit
ies
o
f
s
o
p
h
is
ticated
n
eu
r
al
-
n
etwo
r
k
s
tr
u
ctu
r
es in
ef
f
ec
tiv
el
y
ca
p
tu
r
in
g
in
tr
icate
s
en
tim
en
t r
elatio
n
s
h
ip
s
.
Sh
an
m
u
g
av
ad
iv
el
et
a
l.
[
1
7
]
tack
led
t
h
e
c
h
allen
g
in
g
task
o
f
id
e
n
tify
in
g
o
f
f
en
s
iv
e
wo
r
d
s
an
d
p
er
f
o
r
m
in
g
s
en
tim
en
t
ev
alu
ati
o
n
o
n
co
d
e
-
m
ix
e
d
in
f
o
r
m
atio
n
th
at
in
clu
d
ed
b
o
th
E
n
g
lis
h
a
n
d
T
am
il
lan
g
u
ag
e.
B
y
u
tili
zin
g
ad
v
an
ce
d
DL
an
d
ML
tech
n
iq
u
es,
s
p
ec
if
ically
em
p
lo
y
in
g
m
o
d
els
lik
e
R
o
B
E
R
T
a
an
d
B
E
R
T
,
r
esear
ch
er
s
wer
e
ab
le
to
s
h
o
wca
s
e
s
ig
n
if
ican
t
ac
h
iev
em
en
ts
in
th
e
f
ield
.
No
tab
ly
,
th
ey
ac
h
iev
ed
ac
cu
r
ac
y
lev
els
o
f
6
5
%
f
o
r
s
en
tim
en
t
e
v
alu
atio
n
an
d
7
9
%
f
o
r
in
ap
p
r
o
p
r
iate
lan
g
u
a
g
e
r
ec
o
g
n
itio
n
.
Am
o
n
g
th
e
v
a
r
io
u
s
a
p
p
r
o
a
c
h
e
s
t
e
s
t
e
d
,
t
h
e
a
d
a
p
t
e
r
-
B
E
R
T
a
p
p
r
o
a
c
h
p
r
o
v
e
d
t
o
b
e
e
f
f
i
c
i
e
n
t
i
n
a
c
h
i
e
v
i
n
g
t
h
e
s
e
r
e
s
u
l
t
s
.
C
h
u
n
d
i
e
t
a
l
.
[
1
8
]
,
p
r
esen
ted
an
in
n
o
v
ativ
e
lex
ico
n
-
b
ased
ap
p
r
o
ac
h
ca
lled
NB
L
ex
f
o
r
ac
cu
r
ately
p
r
ed
ictin
g
s
en
tim
en
ts
in
co
d
e
-
s
witch
ed
tex
t
wr
itten
in
Ka
n
n
ad
a
an
d
E
n
g
lis
h
lan
g
u
ag
e
s
.
T
h
e
a
p
p
r
o
a
ch
u
tili
ze
d
le
x
ico
n
s
,
wh
ich
is
a
co
llectio
n
o
f
wo
r
d
s
an
d
t
h
ei
r
ass
o
ciate
d
s
en
tim
en
ts
,
to
an
aly
s
e
th
e
tex
t
an
d
d
eter
m
in
e
th
e
s
en
tim
en
ts
ex
p
r
ess
ed
with
in
it.
T
h
is
m
eth
o
d
d
em
o
n
s
tr
ated
th
e
im
p
o
r
ta
n
ce
o
f
s
en
tim
en
t
ev
alu
atio
n
a
p
p
r
o
ac
h
es
u
tili
zin
g
lex
ico
n
s
in
m
u
l
tili
n
g
u
al
co
n
te
x
ts
b
y
o
u
t
p
er
f
o
r
m
in
g
s
tan
d
ar
d
m
eth
o
d
s
s
u
ch
as
B
i
-
L
STM
a
n
d
NB
with
r
esp
ec
t
to
tr
u
e
p
o
s
itiv
e
(
T
P)
r
ate
an
d
ac
cu
r
ac
y
.
R
o
y
[
1
9
]
h
av
e
f
o
cu
s
ed
o
n
th
e
d
if
f
ic
u
lties
o
f
an
al
y
s
in
g
s
en
tim
en
t
in
lan
g
u
ag
es
with
lim
ited
r
eso
u
r
ce
s
s
u
ch
as
Ma
lay
alam
an
d
Kan
n
ad
a.
T
o
tack
le
th
ese
d
if
f
icu
lties
,
th
ey
h
av
e
p
r
o
p
o
s
ed
an
en
s
em
b
le
ap
p
r
o
ac
h
.
T
h
e
af
o
r
em
en
tio
n
e
d
ap
p
r
o
ac
h
d
em
o
n
s
tr
ated
o
u
ts
tan
d
i
n
g
F1
-
s
co
r
es
wh
e
n
ap
p
lied
to
co
d
e
-
m
i
x
ed
lan
g
u
a
g
es,
th
er
eb
y
em
p
h
asizin
g
th
e
ef
f
icac
y
o
f
en
s
em
b
le
m
eth
o
d
s
f
o
r
ad
d
r
ess
in
g
th
e
lim
ita
tio
n
s
p
o
s
ed
b
y
in
s
u
f
f
icien
t
d
ata
av
ailab
ilit
y
.
C
h
u
n
d
i
et
a
l.
[
2
0
]
,
r
ec
o
n
s
id
er
e
d
th
e
task
o
f
an
aly
zin
g
s
en
tim
en
t
in
Kan
n
ad
a
-
E
n
g
lis
h
co
d
e
-
s
witch
ed
tex
t,
wh
ich
was
p
r
esen
ted
i
n
[
1
8
]
.
T
h
ey
em
p
lo
y
ed
th
e
NB
L
ex
ap
p
r
o
ac
h
[
1
8
]
an
d
d
em
o
n
s
tr
ated
th
at
th
eir
a
p
p
r
o
ac
h
a
ch
ie
v
ed
h
ig
h
er
ac
cu
r
ac
y
an
d
F1
-
s
co
r
e
in
co
m
p
ar
is
o
n
with
p
r
io
r
ap
p
r
o
ac
h
es.
T
h
e
af
o
r
em
en
tio
n
e
d
s
tatem
en
t
h
ig
h
li
g
h
ts
th
e
o
n
g
o
in
g
d
ev
elo
p
m
en
t
an
d
im
p
r
o
v
em
en
t
o
f
s
en
tim
en
t
ev
alu
atio
n
m
eth
o
d
s
with
in
co
d
e
-
s
witch
ed
l
an
g
u
ag
e
c
o
n
tex
ts
.
C
h
u
n
d
i
et
a
l.
[
2
1
]
,
u
tili
ze
d
a
ch
ar
ac
ter
-
lev
el
n
-
g
r
am
s
m
eth
o
d
to
ef
f
ec
tiv
ely
d
etec
t
co
d
e
-
s
witch
ed
an
d
m
o
n
o
lin
g
u
al
c
o
n
ten
t
in
E
n
g
lis
h
-
Kan
n
ad
a
o
n
lin
e
s
o
cial
n
etw
o
r
k
in
g
d
ata.
T
h
e
r
esu
lts
o
b
t
ain
ed
f
r
o
m
th
is
m
eth
o
d
h
av
e
s
h
o
wn
a
n
o
ta
b
le
im
p
r
o
v
em
e
n
t
in
F1
-
s
co
r
e
a
n
d
ac
cu
r
ac
y
wh
en
co
m
p
ar
e
d
to
c
o
n
v
e
n
tio
n
al
ML
ap
p
r
o
ac
h
es.
T
h
is
wo
r
k
h
ig
h
lig
h
ted
th
e
s
ig
n
if
ican
ce
o
f
em
p
lo
y
in
g
c
o
n
tex
t
-
awa
r
e
f
ea
tu
r
e
-
ex
tr
ac
tio
n
a
p
p
r
o
ac
h
in
o
r
d
er
to
ac
h
iev
e
b
etter
p
er
f
o
r
m
an
ce
.
Fin
ally
,
t
h
e
s
tu
d
y
co
n
d
u
cted
b
y
[
2
2
]
f
o
cu
s
ed
o
n
th
e
ap
p
licatio
n
o
f
s
en
tim
en
t
an
aly
s
is
tech
n
iq
u
e
s
to
an
aly
s
e
C
OVI
D
-
1
9
in
f
o
r
m
atio
n
co
n
tain
in
g
th
e
Kan
n
ad
a
lan
g
u
ag
e.
T
h
e
r
esear
ch
er
s
u
tili
ze
d
v
ar
io
u
s
ML
an
d
en
s
em
b
le
a
p
p
r
o
ac
h
to
ac
h
iev
e
th
eir
o
b
jectiv
es.
T
h
e
f
in
d
in
g
s
o
f
th
e
r
esear
ch
r
ev
ea
led
ac
cu
r
ac
y
s
co
r
es
th
at
v
a
r
ied
b
etwe
en
6
6
%
an
d
6
9
%,
th
er
eb
y
h
ig
h
lig
h
tin
g
t
h
e
v
er
s
atility
o
f
m
eth
o
d
s
f
o
r
s
en
tim
en
t a
n
aly
s
is
in
ef
f
ec
tiv
e
ly
an
aly
s
in
g
v
a
r
io
u
s
f
ield
s
an
d
d
atasets
.
Fro
m
th
e
ab
o
v
e
a
n
aly
s
is
o
f
th
e
v
ar
io
u
s
s
tu
d
ies
in
th
e
f
ield
o
f
s
en
tim
en
t
an
aly
s
is
r
ev
ea
ls
a
co
m
m
o
n
ch
allen
g
e:
th
e
ab
s
en
ce
o
f
a
co
m
p
r
eh
e
n
s
iv
e
an
d
s
tan
d
ar
d
i
ze
d
d
ataset
s
p
ec
if
ically
d
esig
n
ed
f
o
r
Kan
n
ad
a
s
en
tim
en
t
an
aly
s
is
.
Desp
ite
th
e
ad
v
an
ce
m
e
n
ts
in
s
en
tim
en
t
an
aly
s
is
tech
n
iq
u
es
an
d
th
e
em
er
g
en
ce
o
f
s
o
p
h
is
ticated
m
o
d
els
an
d
m
et
h
o
d
o
lo
g
ies,
r
esear
ch
er
s
co
n
s
i
s
ten
tly
en
co
u
n
ter
lim
itatio
n
s
d
u
e
to
th
e
lack
o
f
a
r
o
b
u
s
t
d
ataset
th
at
ac
cu
r
ately
r
ep
r
esen
ts
th
e
n
u
an
ce
s
o
f
s
en
tim
en
t
in
th
e
Kan
n
ad
a
lan
g
u
ag
e.
T
h
e
s
tu
d
ies
d
is
cu
s
s
ed
ea
r
lier
h
ig
h
lig
h
t
t
h
e
in
n
o
v
ativ
e
a
p
p
r
o
ac
h
es
an
d
te
ch
n
iq
u
es
r
esear
ch
er
s
h
a
v
e
em
p
lo
y
ed
to
o
v
er
c
o
m
e
th
is
is
s
u
e.
Fo
r
in
s
tan
ce
,
s
o
m
e
s
tu
d
ies
r
eso
r
t
to
cr
ea
tin
g
t
h
eir
o
wn
d
atasets
,
o
f
ten
b
y
t
r
an
s
latin
g
ex
is
tin
g
d
atasets
f
r
o
m
o
th
er
lan
g
u
a
g
es
to
Kan
n
ad
a
u
s
in
g
to
o
ls
lik
e
Go
o
g
le
T
r
a
n
s
late.
Ho
wev
er
,
t
h
is
ap
p
r
o
ac
h
m
ay
in
tr
o
d
u
ce
c
h
allen
g
es
r
elate
d
t
o
th
e
ac
cu
r
ac
y
a
n
d
au
th
e
n
ticity
o
f
s
en
tim
en
t
lab
els,
as
m
ac
h
in
e
tr
an
s
latio
n
m
ay
n
o
t
alwa
y
s
ca
p
tu
r
e
th
e
s
u
b
tleties
o
f
s
en
tim
en
t
ex
p
r
ess
io
n
s
in
Kan
n
ad
a.
Oth
er
s
tu
d
ies
lev
er
ag
e
en
s
em
b
le
tech
n
iq
u
es,
d
ee
p
lear
n
in
g
m
o
d
els,
an
d
lex
ico
n
-
b
ased
ap
p
r
o
ac
h
es
to
en
h
an
ce
s
en
tim
en
t
an
aly
s
is
ac
cu
r
ac
y
d
esp
ite
th
e
d
ata
s
ca
r
city
.
T
h
e
s
e
ap
p
r
o
a
ch
es
o
f
ten
in
v
o
l
v
e
a
co
m
b
in
atio
n
o
f
f
ea
tu
r
e
e
x
tr
a
ctio
n
,
Po
S
ta
g
g
in
g
,
an
d
em
o
tio
n
p
r
ed
ictio
n
m
eth
o
d
o
lo
g
ies
to
in
f
e
r
s
en
tim
en
t
f
r
o
m
lim
ited
d
atasets
.
Desp
it
e
th
ese
in
n
o
v
ativ
e
s
tr
ateg
ies,
th
e
lack
o
f
a
s
tan
d
ar
d
ized
an
d
wid
ely
ac
ce
p
ted
d
ataset
f
o
r
Kan
n
a
d
a
s
en
tim
en
t
an
aly
s
is
r
em
ain
s
a
s
ig
n
if
ican
t
b
o
ttlen
ec
k
in
th
e
f
ield
.
A
r
eliab
le
d
ataset
wo
u
ld
n
o
t
o
n
ly
f
ac
ilit
ate
m
o
r
e
ac
cu
r
ate
s
en
tim
en
t
an
aly
s
is
b
u
t
also
en
ab
le
r
esea
r
ch
er
s
to
b
en
c
h
m
ar
k
an
d
c
o
m
p
ar
e
d
if
f
e
r
en
t
m
o
d
els
an
d
tec
h
n
iq
u
es
ef
f
ec
tiv
el
y
.
I
n
co
n
clu
s
io
n
,
wh
ile
ad
v
an
c
em
en
ts
in
s
en
tim
en
t
an
aly
s
is
m
eth
o
d
o
lo
g
ies
ar
e
p
r
o
m
is
in
g
,
th
e
f
ield
wo
u
ld
g
r
ea
tly
b
e
n
ef
it
f
r
o
m
th
e
d
ev
e
lo
p
m
en
t
a
n
d
a
d
o
p
tio
n
o
f
a
s
t
an
d
ar
d
ized
Kan
n
a
d
a
s
en
tim
e
n
t
an
aly
s
is
d
ataset.
C
o
llab
o
r
ativ
e
ef
f
o
r
ts
to
war
d
s
d
ataset
cr
ea
tio
n
,
v
alid
ati
o
n
,
a
n
d
s
h
ar
in
g
ar
e
ess
en
tial
to
d
r
i
v
e
f
u
r
th
er
p
r
o
g
r
ess
an
d
in
n
o
v
atio
n
in
Kan
n
a
d
a
s
en
tim
en
t a
n
aly
s
is
r
esear
ch
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
In
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
3
7
,
No
.
3
,
Ma
r
ch
20
2
5
:
1
661
-
1
6
7
1
1664
3.
M
E
T
H
O
D
B
ased
o
n
th
e
liter
atu
r
e
s
u
r
v
e
y
co
n
d
u
cted
,
it
is
ev
id
en
t
th
at
a
m
ajo
r
ity
o
f
th
e
e
x
is
tin
g
wo
r
k
s
in
s
en
tim
en
t
an
aly
s
is
f
o
r
Kan
n
ad
a
l
an
g
u
ag
e
h
av
e
m
ai
n
ly
f
o
c
u
s
s
ed
o
n
cr
ea
tin
g
th
eir
o
wn
d
ata
s
ets
f
o
r
ev
alu
atio
n
p
u
r
p
o
s
es.
T
h
er
e
is
a
n
o
tab
le
s
ca
r
city
o
f
s
tan
d
ar
d
ize
d
d
atasets
s
p
ec
if
ically
d
esig
n
ed
f
o
r
Kan
n
ad
a
lan
g
u
ag
e
s
en
tim
en
t
an
aly
s
is
,
esp
ec
ially
th
o
s
e
th
at
co
m
e
eq
u
ip
p
e
d
with
s
en
tim
en
t
lab
els.
C
o
n
s
e
q
u
en
tly
,
th
is
s
tu
d
y
u
n
d
er
tak
es
th
e
task
o
f
p
r
ep
ar
i
n
g
a
d
ataset
th
at
alr
ea
d
y
in
c
o
r
p
o
r
ates
s
en
tim
en
t
lab
els
f
o
r
an
aly
tic
al
p
u
r
p
o
s
es.
Hen
ce
,
d
r
awin
g
in
s
p
ir
atio
n
f
r
o
m
[
1
5
]
,
wh
er
ein
th
ey
tr
an
s
la
ted
th
e
I
MBD
d
ataset
f
r
o
m
E
n
g
lis
h
to
Ka
n
n
ad
a
u
s
in
g
Go
o
g
le
T
r
an
s
late,
th
is
wo
r
k
ad
o
p
ts
a
s
im
ilar
m
eth
o
d
o
lo
g
y
.
Sp
ec
if
ically
,
it
tr
a
n
s
lates
th
e
Sem
E
v
al
2
0
1
4
t
ask
4
d
ataset
[
2
3
]
f
r
o
m
E
n
g
lis
h
to
Kan
n
ad
a.
T
h
e
c
h
o
ice
o
f
Sem
E
v
al
2
0
1
4
t
ask
4
d
ataset
is
m
o
tiv
ated
b
y
its
in
clu
s
io
n
o
f
lab
elled
asp
ec
t
wo
r
d
s
,
asp
ec
t
ca
teg
o
r
ies
(
s
en
ti
m
en
ts
)
,
an
d
p
o
lar
ity
,
wh
ic
h
p
r
o
v
e
in
s
tr
u
m
e
n
tal
in
ev
alu
atin
g
th
e
e
f
f
ec
tiv
en
ess
o
f
th
is
wo
r
k
w
h
en
co
m
p
a
r
ed
with
s
tan
d
ar
d
d
atasets
.
T
h
e
o
v
e
r
all
ar
ch
itectu
r
e
o
f
th
is
s
tu
d
y
is
p
r
esen
ted
in
Fi
g
u
r
e
1
.
I
n
itially
,
th
e
Sem
E
v
a
l
2
0
1
4
t
ask
4
d
ataset
s
er
v
es
as
th
e
f
o
u
n
d
atio
n
.
Su
b
s
eq
u
en
tly
,
E
n
g
lis
h
d
ata
u
n
d
er
g
o
es
tr
an
s
latio
n
t
o
Ka
n
n
ad
a
t
h
r
o
u
g
h
G
o
o
g
le
T
r
an
s
late.
T
h
e
r
esu
ltan
t
Kan
n
ad
a
r
aw
d
ata
th
en
u
n
d
e
r
g
o
es
p
r
e
p
r
o
ce
s
s
in
g
to
attain
clea
n
d
ata.
Fr
o
m
th
is
clea
n
d
ata,
f
ea
tu
r
es
ar
e
ex
tr
ac
ted
u
s
in
g
Po
S
tag
g
i
n
g
.
T
h
ese
ex
tr
ac
ted
f
ea
tu
r
es
ar
e
s
u
b
s
eq
u
en
tly
em
p
lo
y
e
d
in
tr
a
in
in
g
th
e
class
if
ier
m
o
d
el
an
d
th
e
ev
al
u
atio
n
is
d
o
n
e.
Fig
u
r
e
1
.
Pro
p
o
s
ed
a
r
c
h
itectu
r
e
3
.
1
.
P
re
pa
r
a
t
io
n o
f
d
a
t
a
s
et
I
n
th
e
in
itial
p
h
ase,
th
e
f
o
c
u
s
is
o
n
th
e
d
ataset,
p
ar
ticu
lar
ly
th
e
Sem
E
v
al
2
0
1
4
t
ask
4
d
ata
s
et,
wh
ich
co
m
p
r
is
es
r
aw
tex
t
en
co
m
p
a
s
s
in
g
d
iv
er
s
e
r
ev
iews
p
er
tain
in
g
to
lap
to
p
s
an
d
r
estau
r
a
n
t
s
.
T
h
i
s
d
ataset
i
s
p
ar
ticu
lar
ly
v
alu
a
b
le
as
it
i
n
clu
d
es
asp
ec
t
ter
m
s
an
d
p
o
lar
ities
co
r
r
esp
o
n
d
in
g
to
ea
ch
r
ev
iew,
th
er
eb
y
f
ac
ilit
atin
g
asp
ec
t
-
b
ased
s
en
tim
en
t
class
if
icatio
n
.
Su
b
s
eq
u
en
tly
,
th
e
E
n
g
lis
h
r
aw
tex
t
u
n
d
er
g
o
es
tr
an
s
latio
n
in
to
Kan
n
ad
a
r
aw
tex
t
u
s
in
g
Go
o
g
le
T
r
an
s
late.
An
ex
am
p
l
e
o
f
s
o
m
e
tex
t
is
g
iv
en
in
T
a
b
le
1
.
Fo
llo
win
g
th
e
tr
an
s
latio
n
p
r
o
ce
s
s
,
th
e
r
aw
d
a
ta
p
r
o
ce
ed
s
to
p
r
ep
r
o
ce
s
s
in
g
,
as e
lab
o
r
ated
in
th
e
s
u
b
s
eq
u
en
t sectio
n
.
T
ab
le
1
.
E
n
g
lis
h
to
Kan
n
a
d
a
t
ex
t
S
L.
N
o
En
g
l
i
sh
r
e
v
i
e
w
K
a
n
n
a
d
a
r
e
v
i
e
w
1
O
t
h
e
r
t
h
a
n
n
o
t
b
e
i
n
g
a
f
a
n
o
f
c
l
i
c
k
p
a
d
s
(
i
n
d
u
s
t
r
y
s
t
a
n
d
a
r
d
t
h
e
se
d
a
y
s)
a
n
d
t
h
e
l
o
u
sy
i
n
t
e
r
n
a
l
s
p
e
a
k
e
r
s
,
i
t
's
h
a
r
d
f
o
r
m
e
t
o
f
i
n
d
t
h
i
n
g
s
a
b
o
u
t
t
h
i
s
n
o
t
e
b
o
o
k
I
d
o
n
'
t
l
i
k
e
,
e
sp
e
c
i
a
l
l
y
c
o
n
si
d
e
r
i
n
g
t
h
e
$
3
5
0
p
r
i
c
e
t
a
g
.
K
l
i
k
p
y
ā
ḍ
g
a
ḷ
a
(
ī
d
i
n
a
g
a
ḷ
a
l
l
i
u
d
y
a
m
a
d
a
p
r
a
m
ā
ṇ
i
t
a
)
m
a
t
t
u
a
sa
h
y
a
v
ā
d
a
ā
n
t
a
ri
k
a
sp
ī
k
a
r
g
a
ḷ
a
a
b
h
i
m
ā
n
i
y
ā
g
i
r
a
d
e
b
ē
re
,
n
ā
n
u
i
ṣṭ
a
p
a
ḍ
a
d
a
ī
n
ō
ṭ
b
u
k
b
a
g
g
e
v
i
ṣ
a
y
a
g
a
ḷ
a
n
n
u
h
u
ḍ
u
k
a
l
u
n
a
n
a
g
e
k
a
ṣ
ṭ
a
v
ā
g
u
t
t
a
d
e
,
v
i
ś
ē
ṣ
a
v
ā
g
i
$
3
5
0
b
e
l
e
y
a
n
n
u
p
a
r
i
g
a
ṇ
i
s
i
.
2
N
o
i
n
s
t
a
l
l
a
t
i
o
n
d
i
s
k
(
D
V
D
)
i
s
i
n
c
l
u
d
e
d
.
Y
ā
v
u
d
ē
a
n
u
s
t
h
ā
p
a
n
ā
ḍ
i
s
k
(
ḍ
i
v
i
ḍ
i
)
o
ḷ
a
g
o
ṇ
ḍ
i
l
l
a
.
3
W
o
r
k
s
w
e
l
l
,
a
n
d
I
a
m
e
x
t
r
e
m
e
l
y
h
a
p
p
y
t
o
b
e
b
a
c
k
t
o
a
n
a
p
p
l
e
O
S
.
U
t
t
a
m
a
v
ā
g
i
k
ā
ry
a
n
i
rv
a
h
i
s
u
t
t
a
d
e
m
a
t
t
u
ā
p
a
l
ō
'
e
s
g
e
h
i
n
t
i
ru
g
a
l
u
n
a
n
a
g
e
t
u
m
b
ā
s
a
n
t
ō
ṣa
v
ā
g
i
d
e
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
A
n
o
ve
l d
a
ta
s
et
a
n
d
p
a
r
t
-
of
-
s
p
ee
ch
ta
g
g
in
g
a
p
p
r
o
a
ch
fo
r
…
(
S
u
n
il Mu
g
a
lih
a
lli E
s
h
w
a
r
a
p
p
a
)
1665
3
.
2
.
P
re
-
p
ro
ce
s
s
ing
I
n
NL
P,
p
r
ep
r
o
ce
s
s
in
g
is
a
cr
u
cial
s
tep
f
o
r
an
y
r
aw
tex
t.
T
h
er
ef
o
r
e,
in
th
e
s
ec
o
n
d
s
tag
e
o
f
th
is
s
tu
d
y
,
th
e
Kan
n
a
d
a
r
aw
tex
t
u
n
d
er
g
o
es
p
r
ep
r
o
ce
s
s
in
g
.
I
n
itially
,
t
o
k
en
izatio
n
is
p
er
f
o
r
m
e
d
to
s
eg
m
en
t
ea
c
h
wo
r
d
with
in
th
e
r
e
v
iew
s
en
ten
ce
.
Su
b
s
eq
u
en
tly
,
b
r
ac
k
ets,
s
y
m
b
o
l
s
,
h
y
p
h
en
s
,
in
v
er
ted
co
m
m
as,
an
d
o
th
er
s
y
m
b
o
ls
w
er
e
r
em
o
v
ed
p
o
s
t
-
to
k
en
izati
o
n
ex
ce
p
t
f
u
ll
s
to
p
,
ex
clam
atio
n
m
ar
k
a
n
d
co
m
m
as.
Fo
llo
w
in
g
th
e
p
u
n
ct
u
atio
n
r
em
o
v
al,
th
e
tex
t
u
n
d
er
g
o
es
s
tem
m
in
g
an
d
lem
m
atiza
tio
n
p
r
o
ce
s
s
es
to
g
et
th
e
in
ten
d
ed
m
ea
n
in
g
o
f
wo
r
d
s
.
Su
b
s
eq
u
en
tly
,
a
s
to
p
wo
r
d
s
lib
r
ar
y
is
co
n
s
tr
u
cted
t
o
f
ilter
o
u
t
co
m
m
o
n
ly
u
s
ed
wo
r
d
s
th
at
co
n
tr
ib
u
te
m
in
i
m
al
m
ea
n
in
g
f
u
l
in
f
o
r
m
atio
n
.
Up
o
n
co
m
p
letio
n
o
f
th
is
co
m
p
r
e
h
en
s
iv
e
p
r
ep
r
o
ce
s
s
in
g
p
ip
eli
n
e,
a
clea
n
tex
t
is
o
b
tain
ed
,
co
n
tain
in
g
m
ea
n
in
g
f
u
l
wo
r
d
s
ex
tr
ac
ted
f
r
o
m
t
h
e
o
r
ig
in
al
r
e
v
iew
s
en
ten
ce
s
.
Su
b
s
eq
u
en
tly
,
th
e
f
ea
tu
r
e
ex
tr
ac
tio
n
p
r
o
ce
s
s
is
in
itiated
,
wh
ich
is
elab
o
r
ated
u
p
o
n
in
d
etail
in
th
e
s
u
b
s
eq
u
en
t
s
ec
tio
n
o
f
th
is
wo
r
k
.
3
.
3
.
F
e
a
t
ure
e
x
t
r
a
ct
io
n
I
n
th
is
s
tu
d
y
,
th
e
f
ea
tu
r
e
e
x
tr
ac
tio
n
p
r
o
ce
s
s
f
r
o
m
th
e
cle
an
r
ev
iews,
c
o
m
p
r
is
in
g
wo
r
d
s
in
ea
ch
s
en
ten
ce
,
em
p
l
o
y
s
th
e
Po
S
t
ag
g
in
g
a
p
p
r
o
ac
h
.
T
h
e
p
u
r
p
o
s
e
o
f
Po
S
tag
g
in
g
in
th
is
w
o
r
k
is
two
f
o
ld
:
t
o
co
m
p
r
eh
e
n
d
th
e
g
r
am
m
atica
l
s
tr
u
ctu
r
e
o
f
r
ev
iew
s
en
ten
ce
s
an
d
to
d
is
am
b
ig
u
ate
wo
r
d
s
with
m
u
ltip
le
m
ea
n
in
g
s
.
T
h
e
u
tili
za
tio
n
o
f
Po
S
tag
g
in
g
aid
s
in
g
ain
i
n
g
in
s
ig
h
ts
in
to
t
h
e
s
y
n
tactic
co
m
p
o
s
itio
n
o
f
r
ev
iew
s
en
ten
ce
s
an
d
ai
d
s
in
r
eso
lv
in
g
am
b
ig
u
ity
with
in
wo
r
d
s
.
I
n
th
is
s
tu
d
y
,
th
e
T
r
ig
r
a
m
s
'
n
'
T
ag
s
(
T
n
T
)
m
o
d
el
is
em
p
lo
y
ed
as
Po
S
tag
g
e
r
s
.
T
h
e
T
n
T
m
eth
o
d
was
in
itially
in
tr
o
d
u
ce
d
b
y
[
2
4
]
,
in
w
h
ich
th
e
r
esear
ch
er
attem
p
ted
to
d
esig
n
ate
a
s
u
ita
b
le
lab
el
o
r
tag
b
y
co
m
p
u
tin
g
th
e
lik
elih
o
o
d
s
o
f
p
o
ten
tial
tag
s
f
o
r
ev
er
y
p
h
r
ase.
T
h
e
T
n
T
m
eth
o
d
s
er
v
es
as
a
v
ar
ian
t
o
f
s
ec
o
n
d
-
o
r
d
er
Ma
r
k
o
v
ap
p
r
o
ac
h
th
at
in
teg
r
at
es
m
u
ltip
le
n
-
g
r
a
m
m
o
d
els,
in
clu
d
in
g
tr
ig
r
a
m
,
b
i
g
r
am
an
d
u
n
i
g
r
am
,
with
th
e
g
o
al
to
d
eter
m
in
e
th
e
m
o
s
t
ap
p
r
o
p
r
iate
tag
f
o
r
a
g
iv
en
w
o
r
d
.
Acc
o
r
d
in
g
to
th
e
r
esear
ch
co
n
d
u
cted
b
y
[
2
5
]
,
th
e
p
r
o
ce
s
s
o
f
g
en
e
r
atin
g
a
s
e
q
u
en
ce
o
f
Po
S
tag
s
r
ep
r
esen
ted
as
1
,
…
,
f
r
o
m
th
e
s
p
ec
if
ic
s
eq
u
en
ce
o
f
p
h
r
ases
/wo
r
d
s
r
ep
r
esen
ted
as
1
,
…
,
ca
n
b
e
ac
h
ie
v
ed
b
y
u
tili
zin
g
(
1
)
.
Mo
r
eo
v
er
,
th
e
u
n
ig
r
am
s
,
b
ig
r
am
s
an
d
tr
ig
r
am
s
f
o
r
a
g
iv
e
n
s
en
t
en
ce
ar
e
co
n
v
er
te
d
u
s
in
g
(
2
)
to
(
4
)
.
1
,
…
,
[
∏
(
|
−
1
,
−
2
)
(
|
)
=
1
]
(
+
1
|
)
(
1
)
=
(
)
=
(
)
(
2
)
=
(
|
−
1
)
=
(
−
1
)
(
−
1
)
(
3
)
=
(
|
−
2
−
1
)
=
(
−
2
−
1
)
(
−
2
−
1
)
(
4
)
W
h
er
e,
(
)
d
en
o
tes
th
e
f
r
eq
u
e
n
c
y
o
f
th
e
o
c
cu
r
r
e
n
ce
f
o
r
th
e
w
o
r
d
,
wh
ile
d
en
o
tes
th
e
o
v
e
r
all
wo
r
d
s
p
r
esen
t
in
tr
ain
in
g
d
a
ta.
Fu
r
th
er
,
f
o
r
d
eter
m
in
in
g
th
e
lik
el
ih
o
o
d
o
f
a
p
ar
ticu
lar
Po
S
ta
g
o
cc
u
r
r
i
n
g
af
ter
a
s
p
ec
if
ic
Po
S tag
in
a
s
eq
u
en
ce
,
th
e
(
5
)
is
u
tili
ze
d
.
(
|
−
1
)
=
(
,
)
(
)
(
5
)
T
h
e
(
5
)
r
e
p
r
esen
ts
th
e
p
r
in
cip
les
o
f
co
n
d
itio
n
al
p
r
o
b
ab
ilit
y
,
wh
er
e
th
e
n
u
m
er
ato
r
(
(
,
)
)
r
ep
r
esen
ts
th
e
co
u
n
t
o
f
tim
e
s
th
e
wo
r
d
(
)
is
a
s
s
o
ciate
d
w
i
th
th
e
Po
S
tag
(
)
in
th
e
d
atas
et,
an
d
th
e
d
en
o
m
in
at
o
r
(
)
r
ep
r
esen
ts
th
e
to
tal
co
u
n
t o
f
o
cc
u
r
r
e
n
ce
s
o
f
th
e
wo
r
d
(
)
in
th
e
d
ataset.
T
h
is
co
n
d
itio
n
al
p
r
o
b
a
b
ilit
y
ca
lcu
latio
n
aid
s
in
th
e
ap
p
r
o
ac
h
b
y
p
r
o
v
id
i
n
g
a
m
ec
h
an
is
m
to
esti
m
ate
th
e
p
r
o
b
ab
ilit
y
o
f
Po
S
tag
s
eq
u
en
ce
s
,
wh
ich
is
cr
u
cial
f
o
r
d
eter
m
in
i
n
g
th
e
m
o
s
t
p
r
o
b
a
b
le
tag
s
eq
u
en
ce
f
o
r
ea
ch
wo
r
d
in
a
g
i
v
en
s
en
ten
ce
.
Fu
r
th
er
m
o
r
e,
it
is
wo
r
th
n
o
ti
n
g
th
at
t
h
e
c
o
m
p
u
tatio
n
o
f
tr
i
g
r
am
p
r
o
b
ab
ilit
y
em
p
lo
y
in
g
th
e
(
1
)
u
tili
zin
g
th
e
p
r
ep
ar
e
d
d
ataset
is
n
o
t
en
tire
ly
h
elp
f
u
l
b
ec
au
s
e
o
f
th
e
is
s
u
e
o
f
lim
ited
in
f
o
r
m
atio
n
.
C
o
n
s
eq
u
en
tly
,
th
e
in
s
u
f
f
icien
t
f
r
eq
u
en
cy
o
f
o
cc
u
r
r
en
ce
s
o
f
e
v
er
y
tr
i
g
r
am
p
r
ev
e
n
ts
th
e
r
elia
b
le
co
m
p
u
tati
o
n
o
f
its
p
r
o
b
ab
ilit
y
.
I
n
ad
d
itio
n
,
a
s
s
ig
n
in
g
a
p
r
o
b
ab
ilit
y
o
f
ze
r
o
f
o
r
a
p
a
r
ticu
lar
tr
i
g
r
am
ca
n
h
a
v
e
u
n
in
ten
d
ed
co
n
s
eq
u
en
ce
s
,
as
it
im
p
lies
th
at
th
e
ass
o
ciate
d
tr
ig
r
am
was
n
o
t
p
r
e
v
io
u
s
ly
o
b
s
er
v
e
d
in
t
h
e
co
llectio
n
o
f
d
ata.
T
h
u
s
,
it
is
n
o
t
f
ea
s
ib
le
to
ca
te
g
o
r
ize
v
ar
i
o
u
s
s
eq
u
en
ce
s
ca
r
r
y
in
g
a
ze
r
o
p
r
o
b
ab
ilit
y
b
ec
a
u
s
e
th
e
p
o
s
s
ib
ilit
y
o
f
a
wh
o
le
s
eq
u
en
ce
is
d
eter
m
in
e
d
to
ze
r
o
wh
e
n
ev
er
its
em
p
lo
y
m
en
t
is
r
eq
u
ir
ed
f
o
r
a
n
en
tir
ely
n
o
v
el
s
eq
u
e
n
ce
.
T
h
er
ef
o
r
e,
th
e
u
tili
za
tio
n
o
f
a
n
o
r
m
alizin
g
v
ar
iab
le
th
at
i
n
co
r
p
o
r
ate
s
th
e
lin
ea
r
in
ter
p
o
latio
n
o
f
tr
i
g
r
am
s
,
b
ig
r
am
s
an
d
u
n
ig
r
am
s
h
as
b
e
en
f
o
u
n
d
to
y
ield
th
e
m
o
s
t
f
a
v
o
u
r
a
b
le
r
esu
lts
in
th
e
m
o
d
el
f
o
r
th
is
wo
r
k
.
C
o
n
s
eq
u
en
tly
,
th
e
e
v
alu
atio
n
o
f
th
e
tr
ig
r
am
p
r
o
b
a
b
ilit
y
u
s
in
g
th
e
n
o
r
m
alizin
g
v
ar
iab
le
is
c
o
n
d
u
cte
d
u
s
in
g
(
6
).
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
In
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
3
7
,
No
.
3
,
Ma
r
ch
20
2
5
:
1
661
-
1
6
7
1
1666
(
|
−
2
−
1
)
=
1
(
)
+
2
(
|
2
)
+
3
(
|
−
2
−
1
)
(
6
)
W
h
er
e
th
e
s
u
m
o
f
1
,
2
,
an
d
3
is
eq
u
al
t
o
1
,
i.e
.
,
1
+
2
+
3
=
1
.
W
ith
in
th
e
s
co
p
e
o
f
th
is
s
tu
d
y
,
it
is
im
p
o
r
ta
n
t
to
n
o
te
t
h
at
th
e
v
alu
es
o
f
s
r
em
ain
u
n
a
f
f
ec
ted
b
y
th
e
s
p
ec
if
ic
tr
ig
r
a
m
b
ein
g
an
al
y
ze
d
.
T
h
is
is
d
u
e
to
th
e
im
p
lem
en
ta
tio
n
o
f
a
co
n
tex
t
-
in
d
e
p
en
d
en
t
lin
ea
r
-
in
ter
p
o
latio
n
ap
p
r
o
ac
h
.
T
h
e
u
tili
za
tio
n
o
f
th
is
ap
p
r
o
ac
h
f
ac
ilit
ates
th
e
attain
m
en
t
o
f
s
u
p
er
io
r
r
esu
lts
co
m
p
ar
e
d
to
th
e
p
r
ev
ailin
g
co
n
tex
t
-
d
ep
en
d
en
t
m
eth
o
d
o
l
o
g
y
.
B
ec
au
s
e
o
f
th
e
lim
ited
in
f
o
r
m
atio
n
is
s
u
e,
it
i
s
n
o
t
f
ea
s
ib
le
to
d
eter
m
in
e
an
in
d
ep
e
n
d
en
t
s
et
o
f
s
f
o
r
ev
er
y
tr
ig
r
am
.
T
h
e
r
ef
o
r
e
,
th
e
tr
ig
r
am
s
ar
e
o
r
g
a
n
ized
b
a
s
ed
o
n
th
eir
f
r
eq
u
en
cies,
an
d
c
o
r
r
esp
o
n
d
in
g
s
ets
o
f
s
ar
e
co
m
p
u
ted
f
o
r
ev
er
y
c
ateg
o
r
y
.
Acc
o
r
d
i
n
g
t
o
o
u
r
c
u
r
r
en
t
u
n
d
er
s
tan
d
in
g
,
t
h
er
e
h
as
b
ee
n
n
o
p
r
ev
io
u
s
r
esear
ch
th
at
h
as
ex
p
lo
r
ed
th
e
u
s
e
o
f
f
r
eq
u
en
c
y
v
ar
iatio
n
cl
ass
if
icatio
n
f
o
r
in
ter
p
o
latio
n
o
f
lin
ea
r
f
r
eq
u
e
n
cies
in
P
o
S
tag
g
in
g
.
Hen
ce
,
th
e
n
u
m
er
ical
v
alu
es
f
o
r
t
h
e
v
ar
iab
le
s
1
,
2
,
an
d
3
ar
e
d
eter
m
in
e
d
u
s
in
g
th
e
p
r
o
ce
s
s
o
f
d
elete
d
-
in
ter
p
o
latio
n
.
T
h
is
m
eth
o
d
h
elp
s
to
r
em
o
v
e
all
tr
ig
r
am
s
f
r
o
m
tr
ain
in
g
-
s
et
in
a
s
eq
u
en
tial
f
ash
io
n
an
d
f
i
n
d
s
th
e
b
est
p
o
s
s
ib
le
v
a
lu
es
f
o
r
th
e
s
th
r
o
u
g
h
ev
er
y
s
in
g
le
o
n
e
o
f
th
e
r
em
ain
i
n
g
n
-
g
r
am
s
ac
r
o
s
s
all
s
ets.
F
in
d
in
g
th
e
co
u
n
t
o
f
f
r
eq
u
en
cy
o
f
u
n
ig
r
am
s
,
b
ig
r
a
m
s
,
an
d
tr
ig
r
am
s
allo
ws
o
n
e
to
co
m
p
u
tatio
n
ally
ef
f
icien
tly
co
n
s
tr
u
ct
t
h
e
weig
h
ts
h
av
in
g
a
tim
e
c
o
m
p
le
x
ity
th
at
is
lin
ea
r
with
th
e
to
tal
n
u
m
b
er
o
f
d
is
tin
ct
tr
ig
r
am
s
.
3
.
4
.
Cla
s
s
if
ier
I
n
th
is
wo
r
k
,
f
o
r
class
if
icatio
n
,
a
class
if
ier
ca
lled
Kan
n
a
d
aBER
T
(
K
-
B
E
R
T
)
is
p
r
esen
ted
.
T
h
e
B
E
R
T
f
r
am
ewo
r
k
co
n
s
is
ts
o
f
a
m
u
lti
-
lay
er
b
id
ir
ec
tio
n
al
-
tr
an
s
f
o
r
m
er
-
en
co
d
er
[
2
6
]
.
T
h
e
p
u
r
p
o
s
e
o
f
th
is
f
r
am
ewo
r
k
is
to
p
r
etr
ai
n
d
e
ep
b
id
ir
ec
tio
n
al
-
r
ep
r
esen
tatio
n
s
u
s
in
g
u
n
lab
ell
ed
p
h
r
ases
/tex
t/wo
r
d
s
b
y
co
n
d
itio
n
in
g
b
o
th
r
ig
h
t
an
d
lef
t
b
ac
k
g
r
o
u
n
d
ac
r
o
s
s
ev
er
y
lay
er
[
2
6
]
.
B
E
R
T
is
f
r
eq
u
en
tly
u
tili
ze
d
to
f
in
d
a
v
ec
to
r
r
ep
r
esen
tatio
n
f
o
r
ev
er
y
wo
r
d
with
in
a
p
h
r
ase.
T
h
e
s
tan
d
ar
d
B
E
R
T
f
r
am
ewo
r
k
in
itially
r
ec
eiv
es
in
p
u
t
in
th
e
f
o
r
m
o
f
s
en
ten
ce
s
,
wh
ich
ar
e
b
r
o
k
en
d
o
w
n
b
y
a
s
p
ec
if
ic
to
k
en
k
n
o
w
n
as
s
ep
ar
ato
r
(
SEP)
.
T
h
e
in
itial
in
p
u
t
s
eq
u
en
ce
to
k
en
is
co
m
m
o
n
l
y
r
ef
e
r
r
ed
as
class
i
f
icatio
n
(
C
L
S)
to
k
e
n
.
Fo
r
ta
s
k
s
in
v
o
lv
in
g
class
if
icatio
n
,
ev
er
y
wo
r
d
o
f
th
e
s
en
ten
ce
is
r
ep
r
esen
ted
b
y
th
e
last
h
id
d
en
s
ta
te
th
at
co
r
r
esp
o
n
d
s
with
th
e
C
L
S
to
k
en
.
No
ta
b
ly
,
B
E
R
T
alr
ea
d
y
in
co
r
p
o
r
ates
to
k
e
n
izatio
n
p
r
ep
r
o
ce
s
s
in
g
b
y
d
e
f
au
lt.
T
h
e
B
E
R
T
to
k
en
izer
em
p
lo
y
s
a
t
o
k
en
izatio
n
p
r
o
ce
s
s
th
at
in
v
o
lv
es
d
i
v
id
in
g
th
e
s
en
ten
c
e
in
to
i
n
d
iv
id
u
al
to
k
en
s
.
Ad
d
i
tio
n
ally
,
it
s
tr
ateg
ically
p
lace
s
th
e
u
n
iq
u
e
to
k
en
s
C
L
S
an
d
SEP
in
th
eir
r
esp
ec
ti
v
e
p
o
s
itio
n
s
with
in
th
e
to
k
e
n
ized
s
eq
u
en
ce
.
B
y
co
n
s
id
er
i
n
g
th
e
s
tan
d
ar
d
B
E
R
T
f
r
am
ewo
r
k
,
th
is
wo
r
k
p
r
esen
ts
th
e
K
-
B
E
R
T
s
im
ilar
to
th
e
B
E
R
T
f
r
am
ewo
r
k
.
I
n
s
tead
o
f
th
e
p
ass
in
g
t
h
e
co
m
p
lete
s
en
ten
ce
as
in
p
u
t
w
h
ich
g
o
es
f
o
r
p
r
ep
r
o
ce
s
s
in
g
a
n
d
to
k
e
n
izatio
n
,
th
e
K
-
B
E
R
T
m
o
d
el
co
n
s
id
er
s
th
e
tr
ig
r
am
s
as
in
p
u
t.
T
h
e
p
r
o
p
o
s
ed
K
-
B
E
R
T
m
o
d
el
is
s
h
o
wn
in
Fig
u
r
e
2
.
I
n
th
is
K
-
B
E
R
T
m
o
d
el
th
e
f
u
n
ctio
n
o
f
SEP
is
to
s
ep
ar
ate
ea
c
h
tr
ig
r
am
an
d
th
e
f
u
n
ctio
n
o
f
C
L
S
is
to
class
if
y
ea
c
h
tr
ig
r
am
s
o
th
at
a
m
ea
n
in
g
f
u
l
class
if
icatio
n
f
o
r
a
g
i
v
en
s
en
ten
ce
is
ac
h
iev
e
d
.
Als
o
,
in
t
h
is
wo
r
k
o
n
e
-
h
o
t
en
c
o
d
in
g
tr
ig
r
am
s
is
u
s
ed
f
o
r
co
n
v
er
tin
g
ea
ch
tr
ig
r
am
in
to
a
h
ig
h
-
d
im
e
n
s
io
n
al
v
ec
to
r
w
h
er
e
o
n
ly
th
e
elem
en
t
co
r
r
es
p
o
n
d
in
g
to
s
p
ec
if
ic
tr
ig
r
am
is
1
,
a
n
d
all
o
th
e
r
s
ar
e
0
.
Fig
u
r
e
2
.
Pro
p
o
s
ed
K
-
B
E
R
T
m
o
d
el
+
+
+
+
+
+
+
+
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
A
n
o
ve
l d
a
ta
s
et
a
n
d
p
a
r
t
-
of
-
s
p
ee
ch
ta
g
g
in
g
a
p
p
r
o
a
ch
fo
r
…
(
S
u
n
il Mu
g
a
lih
a
lli E
s
h
w
a
r
a
p
p
a
)
1667
I
n
th
e
s
tan
d
ar
d
B
E
R
T
f
r
am
e
wo
r
k
,
f
o
r
ex
tr
ac
tin
g
P
o
S
ta
g
s
,
th
e
C
L
S
is
u
tili
ze
d
to
en
co
d
e
th
e
in
p
u
t
s
en
ten
ce
as
it
d
o
es
f
o
r
class
if
icatio
n
-
task
s
,
wh
er
ea
s
in
th
e
K
-
B
E
R
T
,
ev
er
y
in
p
u
t
to
k
en
is
s
en
t
th
r
o
u
g
h
th
e
s
am
e
f
u
lly
-
co
n
n
ec
ted
class
if
icatio
n
lay
er
s
f
o
r
ex
tr
ac
tin
g
P
o
S
t
ag
s
.
Fu
r
th
e
r
m
o
r
e
,
it
is
im
p
o
r
tan
t
t
o
n
o
te
th
at
th
e
u
tili
za
tio
n
o
f
th
e
wo
r
d
-
p
iece
to
k
en
izer
n
ec
ess
itates
th
e
estab
lis
h
m
en
t
o
f
a
clea
r
r
elatio
n
s
h
ip
am
o
n
g
s
u
b
-
wo
r
d
s
o
r
wo
r
d
-
p
iece
s
an
d
th
e
ir
r
esp
ec
tiv
e
lab
els.
C
o
n
ce
r
n
i
n
g
th
e
wo
r
d
-
lev
el
to
k
en
izer
,
th
er
e
ex
is
ts
a
d
ir
ec
t
m
ap
p
in
g
am
o
n
g
th
e
in
p
u
t
to
k
en
s
an
d
th
ei
r
r
esp
ec
tiv
e
lab
els.
Nev
er
th
eless
,
wh
en
em
p
l
o
y
in
g
a
wo
r
d
-
p
iece
to
k
en
izer
lik
e
th
e
B
E
R
T
to
k
en
izer
,
it
is
p
o
s
s
ib
le
f
o
r
ev
er
y
in
d
iv
id
u
al
wo
r
d
to
b
e
d
iv
i
d
ed
in
to
n
u
m
er
o
u
s
to
k
en
s
.
I
t
is
im
p
e
r
ativ
e
f
o
r
est
ab
lis
h
in
g
a
“to
k
en
-
m
a
p
p
in
g
”
ap
p
r
o
ac
h
th
at
m
a
p
s
wo
r
d
-
p
iec
es
to
co
r
r
esp
o
n
d
in
g
lab
els.
I
n
s
tan
d
ar
d
B
E
R
T
f
r
a
m
ewo
r
k
[
1
]
,
th
e
ch
o
ice
was
m
ad
e
to
u
s
e
th
e
d
e
p
ictio
n
o
f
t
h
e
in
itial
s
u
b
-
to
k
en
to
b
e
th
e
in
p
u
t
f
o
r
th
e
n
e
x
t
lay
er
.
T
h
is
d
ec
is
io
n
was
m
ad
e
wit
h
th
e
in
ten
tio
n
o
f
n
e
g
lectin
g
th
e
d
ep
ictio
n
o
f
th
e
r
em
ain
in
g
s
u
b
-
t
o
k
en
s
.
Fro
m
a
p
r
ac
tical
s
tan
d
p
o
in
t,
th
e
im
p
l
em
en
tatio
n
o
f
th
is
ap
p
r
o
ac
h
i
n
v
o
lv
es
allo
ca
tin
g
th
e
wo
r
d
-
lab
el
f
o
r
th
e
in
itial
s
u
b
-
wo
r
d
,
wh
i
le
all
o
ca
tin
g
a
n
i
m
ag
in
ar
y
lab
el
“X”
f
o
r
th
e
r
e
m
ain
in
g
s
u
b
-
wo
r
d
s
.
Du
r
in
g
th
e
co
m
p
u
tatio
n
o
f
t
h
e
lo
s
s
-
f
u
n
ctio
n
,
th
e
“X”
la
b
els
ass
o
ciate
d
with
th
e
s
u
b
-
to
k
e
n
s
ar
e
n
e
g
lecte
d
.
I
n
ad
d
itio
n
,
it is
p
o
s
s
ib
le
to
allo
ca
te
th
e
lab
el
o
f
a
s
in
g
le
wo
r
d
t
o
d
eter
m
in
e
th
e
d
ep
ictio
n
o
f
t
h
e
f
in
al
wo
r
d
-
p
iece
.
Alter
n
ativ
ely
,
th
e
wo
r
d
-
lab
el
co
u
ld
b
e
e
x
ten
d
e
d
ac
r
o
s
s
all
s
u
b
-
wo
r
d
s
,
an
d
s
u
b
s
eq
u
en
tly
,
a
m
ea
n
d
ep
ictio
n
o
f
th
e
wo
r
d
-
p
iece
s
ca
n
b
e
d
eter
m
in
ed
.
Hen
ce
,
in
K
-
B
E
R
T
,
th
is
wo
r
k
h
as
ch
o
s
en
to
u
tili
ze
t
h
e
in
itial
wo
r
d
-
p
iece
d
ep
ic
tio
n
.
Ho
wev
e
r
,
it
is
ac
k
n
o
wled
g
ed
th
at
th
er
e
a
r
e
ad
d
iti
o
n
al
m
ap
p
in
g
m
eth
o
d
s
th
at
co
u
ld
b
e
ex
p
l
o
r
ed
in
f
u
tu
r
e
r
esear
c
h
.
Up
o
n
co
m
p
le
tio
n
o
f
th
e
K
-
B
E
R
T
b
lo
ck
,
th
e
r
esu
ltin
g
o
u
t
p
u
t
is
s
u
b
s
eq
u
e
n
tly
p
ass
ed
th
r
o
u
g
h
a
d
en
s
e
lay
er
a
n
d
th
e
n
class
if
icatio
n
lay
er
an
d
th
en
th
e
o
u
tp
u
t is ac
h
iev
ed
.
T
h
is
wo
r
k
h
as
o
p
ted
to
em
p
lo
y
th
e
f
r
ee
zin
g
B
E
R
T
ap
p
r
o
ac
h
,
wh
er
ein
th
e
en
tire
B
E
R
T
ar
ch
itectu
r
e
r
em
ain
s
f
ix
ed
,
a
n
d
o
n
ly
u
n
t
r
a
in
ed
lay
er
s
an
d
n
eu
r
o
n
s
at
th
e
en
d
ar
e
ad
d
ed
.
Su
b
s
eq
u
en
tl
y
,
a
n
ew
m
o
d
el
is
tr
ain
ed
in
s
u
ch
a
way
t
h
at
o
n
ly
th
e
weig
h
ts
o
f
th
e
n
ewly
ad
d
e
d
lay
e
r
s
ar
e
u
p
d
ated
d
u
r
in
g
tr
ai
n
in
g
.
T
h
is
ap
p
r
o
ac
h
e
n
s
u
r
es
th
at
t
h
e
c
o
r
e
B
E
R
T
lay
er
s
r
em
ain
u
n
ch
a
n
g
ed
wh
ile
f
in
e
-
tu
n
in
g
t
h
e
m
o
d
el.
Fu
r
th
er
m
o
r
e,
th
is
s
tu
d
y
h
as
ex
ten
d
ed
th
e
ex
is
tin
g
B
E
R
T
f
r
am
ewo
r
k
,
as
d
ep
icted
in
Fig
u
r
e
2
,
b
y
in
co
r
p
o
r
atin
g
a
class
if
icatio
n
lay
er
an
d
a
d
en
s
e
lay
er
.
T
h
e
p
r
im
ar
y
o
b
jecti
v
e
o
f
th
is
m
o
d
if
icatio
n
is
to
en
ab
le
th
e
m
o
d
el
to
g
en
er
ate
tag
s
eq
u
en
ce
s
f
o
r
in
p
u
t
s
en
ten
ce
s
.
T
h
is
is
ac
h
iev
ed
th
r
o
u
g
h
th
e
u
tili
za
tio
n
o
f
th
e
So
f
tMa
x
ac
tiv
atio
n
f
u
n
ctio
n
,
wh
ich
f
ac
ilit
ates
th
e
g
en
er
atio
n
o
f
p
r
o
b
ab
ilit
y
d
is
tr
ib
u
tio
n
s
o
v
er
th
e
o
u
tp
u
t
class
es.
T
o
ad
d
r
ess
th
e
r
is
k
o
f
o
v
e
r
f
itti
n
g
,
a
d
r
o
p
o
u
t
n
o
r
m
aliza
tio
n
tec
h
n
iq
u
e
h
as
b
ee
n
ap
p
lied
s
p
ec
if
ically
o
n
th
e
d
en
s
e
lay
er
.
Fu
r
th
er
,
th
e
r
esu
lts
o
f
th
e
K
-
B
E
R
T
m
o
d
el
ar
e
ev
alu
ated
an
d
co
m
p
ar
ed
with
o
th
er
cla
s
s
if
ier
s
wh
ich
ar
e
d
is
cu
s
s
ed
in
th
e
n
ex
t sectio
n
.
4.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
e
K
-
B
E
R
T
m
o
d
el
was
im
p
lem
en
ted
o
n
a
s
y
s
tem
r
u
n
n
i
n
g
th
e
W
in
d
o
ws
1
1
o
p
e
r
atin
g
s
y
s
tem
,
eq
u
ip
p
e
d
with
1
6
GB
o
f
R
AM
an
d
an
NVI
DI
A
GeFo
r
ce
GT
X
1
6
5
0
g
r
ap
h
ics
c
ar
d
.
T
h
e
im
p
lem
en
tatio
n
was
ca
r
r
ied
o
u
t
u
s
in
g
Py
th
o
n
p
r
o
g
r
am
m
in
g
lan
g
u
ag
e
with
in
t
h
e
An
ac
o
n
d
a
en
v
ir
o
n
m
e
n
t.
Py
th
o
n
p
r
o
v
id
e
d
a
r
o
b
u
s
t
f
r
am
ewo
r
k
f
o
r
ML
an
d
NL
P
task
s
,
m
ak
in
g
it
well
-
s
u
ited
f
o
r
im
p
lem
en
tin
g
co
m
p
lex
m
o
d
els
lik
e
K
-
B
E
R
T
.
Fo
r
ev
alu
atin
g
th
e
p
e
r
f
o
r
m
an
ce
o
f
th
e
class
if
icatio
n
m
o
d
e
l,
v
ar
io
u
s
p
er
f
o
r
m
a
n
ce
m
etr
ics
wer
e
em
p
l
o
y
ed
,
in
clu
d
in
g
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all,
a
n
d
F
-
s
co
r
e
,
i.e
.
,
(
7
)
to
(
1
0
)
r
esp
ec
tiv
ely
.
T
h
ese
m
etr
ics
p
r
o
v
id
e
a
co
m
p
r
eh
e
n
s
iv
e
ass
ess
m
en
t
o
f
th
e
m
o
d
el'
s
ab
ilit
y
to
co
r
r
ec
t
ly
class
if
y
s
en
tim
en
t
i
n
th
e
i
n
p
u
t
te
x
t
d
ata.
T
h
e
p
er
f
o
r
m
an
ce
m
etr
ics ar
e
e
v
alu
ated
as f
o
llo
ws
:
=
+
+
+
+
(
7
)
=
+
(
8
)
=
+
(
9
)
−
=
2
×
×
+
(
1
0
)
W
h
er
e,
T
P
r
ep
r
esen
ts
tr
u
e
-
p
o
s
itiv
e
,
FP
r
ep
r
esen
ts
f
alse
-
p
o
s
itiv
e
,
T
N
r
ep
r
esen
ts
tr
u
e
-
n
eg
ativ
e
an
d
FN
r
e
p
r
e
s
e
n
t
s
f
al
s
e
-
n
e
g
at
i
v
e
.
F
u
r
th
e
r
,
f
r
o
m
t
h
e
d
a
t
a
s
e
t
,
a
s
m
a
l
l
p
a
r
t
o
f
t
h
e
t
e
s
ti
n
g
e
x
a
m
p
l
e
s
is
p
r
e
s
e
n
t
e
d
i
n
T
a
b
l
e
2
.
T
h
e
T
ab
le
3
p
r
esen
ts
th
e
r
esu
l
ts
ac
h
iev
ed
b
y
t
h
e
K
-
B
E
R
T
m
o
d
el
f
o
r
th
e
test
in
g
ex
a
m
p
les
co
n
s
is
tin
g
o
f
wo
r
d
s
o
r
to
k
en
s
alo
n
g
with
th
eir
co
r
r
esp
o
n
d
in
g
tag
s
.
E
ac
h
to
k
en
in
th
e
d
ataset
is
tag
g
ed
with
a
s
p
ec
if
ic
P
O
S
tag
,
p
r
o
v
id
i
n
g
in
f
o
r
m
atio
n
ab
o
u
t
its
g
r
am
m
atica
l
f
u
n
ctio
n
in
a
s
en
ten
ce
.
T
h
e
tag
s
in
clu
d
e
NN
(
n
o
u
n
)
,
NNP
(
p
r
o
p
e
r
n
o
u
n
)
,
PR
P
(
pr
o
n
o
u
n
)
,
C
C
(
co
n
ju
n
ctio
n
)
,
DE
M
(
d
e
m
o
n
s
tr
ativ
e
)
,
VM
(
v
er
b
f
in
ite
)
,
J
J
(
ad
jectiv
e
)
,
R
B
(
ad
v
er
b
)
,
QC
(
ca
r
d
i
n
al
)
,
I
NT
F
(
in
ten
s
if
ier
)
,
an
d
SYM
(
s
y
m
b
o
l
)
.
Fo
r
in
s
tan
ce
,
i
n
th
e
f
i
r
s
t
r
o
w,
"Bū
ṭ"
is
tag
g
ed
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
In
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
3
7
,
No
.
3
,
Ma
r
ch
20
2
5
:
1
661
-
1
6
7
1
1668
as
NNP,
"
s
am
ay
av
u
"
as
NN,
"a
ti"
as
I
NT
F,
an
d
s
o
o
n
.
Similar
ly
,
o
th
er
r
o
ws
co
n
tain
to
k
e
n
s
alo
n
g
with
th
eir
r
esp
ec
tiv
e
P
O
S tag
s
,
p
r
o
v
id
in
g
a
s
tr
u
ctu
r
ed
r
ep
r
esen
tatio
n
o
f
th
e
lin
g
u
is
tic
elem
en
ts
p
r
ese
n
t in
t
h
e
d
ataset.
T
ab
le
2
.
T
esti
n
g
e
x
am
p
les
S
L.
n
o
En
g
l
i
sh
K
a
n
n
a
d
a
1
“
B
o
o
t
t
i
m
e
i
s
su
p
e
r
f
a
st
,
a
r
o
u
n
d
a
n
y
w
h
e
r
e
f
r
o
m
3
5
sec
o
n
d
s
t
o
1
mi
n
u
t
e
.
”
Bū
ṭ
sa
m
a
y
a
v
u
a
t
i
v
ē
g
a
v
ā
g
i
ru
t
t
a
d
e
,
s
u
m
ā
ru
3
5
s
e
k
e
ṇ
ḍ
u
g
a
ḷ
i
n
d
a
1
n
i
m
i
ṣa
d
a
v
a
re
g
e
.
2
“
t
e
c
h
su
p
p
o
r
t
w
o
u
l
d
n
o
t
f
i
x
t
h
e
p
r
o
b
l
e
m u
n
l
e
ss I
b
o
u
g
h
t
y
o
u
r
p
l
a
n
f
o
r
$
1
5
0
p
l
u
s.
”
N
ā
n
u
n
i
m
'
m
a
y
ō
j
a
n
e
y
a
n
n
u
$
1
5
0
p
l
a
s
g
e
k
h
a
r
ī
d
i
s
a
d
a
h
o
ra
t
u
ṭ
e
k
b
e
m
b
a
l
a
v
u
s
a
m
a
sy
e
y
a
n
n
u
p
a
r
i
h
a
r
i
su
v
u
d
i
l
l
a
.
3
“
b
u
t
i
n
r
e
s
u
me
t
h
i
s
c
o
m
p
u
t
e
r
r
o
c
k
s!
”
Ād
a
re
p
u
n
a
r
ā
ra
m
b
h
a
d
a
l
l
i
ī
k
a
m
p
y
ū
ṭ
a
r rā
k
!
4
“
S
e
t
u
p
w
a
s
e
a
sy
.
”
H
o
n
d
i
s
u
v
u
d
u
su
l
a
b
h
a
v
ā
y
i
t
u
.
5
“
D
i
d
n
o
t
e
n
j
o
y
t
h
e
n
e
w
W
i
n
d
o
w
s
8
a
n
d
t
o
u
c
h
scre
e
n
f
u
n
c
t
i
o
n
s.
”
H
o
sa
v
i
ṇ
ḍ
ō
s
8
m
a
t
t
u
ṭ
a
c
skr
ī
n
k
ā
ry
a
g
a
ḷ
a
n
n
u
ā
n
a
n
d
i
s
a
l
i
l
l
a
.
T
ab
le
3
.
W
o
r
d
-
t
ag
s
S
L.
n
o
W
o
r
d
/
-
t
ag
1
B
ū
ṭ
-
NNP
sama
y
a
v
u
-
NN
a
t
i
-
I
N
TF
v
ē
g
a
v
ā
g
i
r
u
t
t
a
d
e
-
VM
,
-
S
Y
M
su
mār
u
-
JJ
35
-
QC
sek
e
ṇ
ḍ
u
g
a
ḷ
i
n
d
a
-
NN
1
-
QC
n
i
m
i
ṣa
d
a
v
a
r
e
g
e
-
NN
2
N
ā
n
u
-
P
R
P
n
i
m'm
a
-
P
R
P
y
ō
j
a
n
e
y
a
n
n
u
-
NN
$
1
5
0
-
QC
p
l
a
sg
e
-
NN
k
h
a
r
ī
d
i
s
a
d
a
-
NN
h
o
r
a
t
u
-
RB
ṭ
e
k
-
NN
b
e
m
b
a
l
a
v
u
-
NN
sama
s
y
e
y
a
n
n
u
-
NN
p
a
r
i
h
a
r
i
su
v
u
d
i
l
l
a
.
-
NN
3
Ā
d
a
r
e
-
CC
p
u
n
a
r
ā
r
a
m
b
h
a
d
a
l
l
i
-
NN
ī
-
D
EM
k
a
m
p
y
ū
ṭ
a
r
-
NN
r
ā
k
-
NNP
!
-
S
Y
M
4
H
o
n
d
i
s
u
v
u
d
u
-
VM
su
l
a
b
h
a
v
ā
y
i
t
u
-
VM
.
–
S
Y
M
5
H
o
sa
-
JJ
v
i
ṇ
ḍ
ō
s
-
NNP
8
-
QC
mat
t
u
-
CC
ṭ
a
c
s
k
r
ī
n
-
NNP
k
ā
r
y
a
g
a
ḷ
a
n
n
u
-
NN
ā
n
a
n
d
i
s
a
l
i
l
l
a
-
VM
.
-
S
Y
M
T
h
e
T
ab
le
4
p
r
esen
ts
a
P
O
S
T
ag
s
et
g
en
er
ated
b
y
th
e
p
r
o
p
o
s
ed
K
-
B
E
R
T
m
o
d
el,
p
r
o
v
id
in
g
a
s
tr
u
ctu
r
ed
r
e
p
r
esen
tatio
n
o
f
li
n
g
u
is
tic
elem
en
ts
with
th
eir
r
esp
ec
tiv
e
tag
s
an
d
d
escr
ip
tio
n
s
.
E
ac
h
r
o
w
in
th
e
tab
le
co
r
r
esp
o
n
d
s
to
a
s
p
ec
if
i
c
wo
r
d
o
r
to
k
en
alo
n
g
with
it
s
P
O
S
tag
an
d
d
escr
ip
tio
n
.
T
h
e
tag
s
in
clu
d
e
NN,
NNP,
PR
P,
C
C
,
DE
M,
VM
,
J
J
,
R
B
,
Q
C
,
I
NT
F,
an
d
SYM.
Fo
r
in
s
tan
ce
,
th
e
wo
r
d
"Sam
ay
av
u
"
is
tag
g
ed
as
NN,
"Bū
ṭ"
a
s
NN
P,
"Nā
n
u
"
as
PR
P,
"Ād
ar
e"
as
C
C
,
"Ī
"
a
s
DE
M,
"Vē
g
av
āg
ir
u
ttad
e"
as
VM
,
"Su
m
ār
u
"
as
J
J
,
"h
o
r
atu
"
as
R
B
,
an
d
s
o
o
n
.
T
h
ese
tag
s
an
d
d
escr
ip
tio
n
s
p
r
o
v
id
e
v
alu
ab
le
in
s
ig
h
ts
in
to
th
e
g
r
am
m
atica
l
r
o
les
an
d
f
u
n
ctio
n
s
o
f
th
e
wo
r
d
s
o
r
t
o
k
en
s
with
in
th
e
d
ataset,
f
ac
ilit
atin
g
lin
g
u
is
tic
an
aly
s
is
an
d
NL
P task
s
.
T
ab
le
4
.
Po
S T
ag
s
et
g
e
n
er
ated
b
y
p
r
o
p
o
s
ed
K
-
B
E
R
T
S
L.
n
o
Ta
g
D
e
scri
p
t
i
o
n
W
o
r
d
1
NN
N
o
u
n
S
a
m
a
y
a
v
u
,
se
k
e
ṇ
ḍ
u
g
a
ḷ
i
n
d
a
,
n
i
m
i
ṣ
a
d
a
v
a
r
e
g
e
,
y
ō
j
a
n
e
y
a
n
n
u
,
p
l
a
s
g
e
,
k
h
a
rī
d
i
s
a
d
a
,
ṭ
e
k
,
b
e
m
b
a
l
a
v
u
,
s
a
m
a
s
y
e
y
a
n
n
u
,
p
a
r
i
h
a
ri
s
u
v
u
d
i
l
l
a
,
p
u
n
a
rā
r
a
m
b
h
a
d
a
l
l
i
,
k
a
m
p
y
ū
ṭ
a
r
,
k
ā
ry
a
g
a
ḷ
a
n
n
u
2
NNP
P
r
o
p
e
r
N
o
u
n
Bū
ṭ
,
r
ā
k
,
v
i
ṇ
ḍ
ō
s,
ṭ
a
c
sk
rī
n
3
P
R
P
P
r
o
n
o
u
n
N
ā
n
u
,
n
i
m
'
m
a
4
CC
C
o
n
j
u
n
c
t
i
o
n
Ād
a
re,
m
a
t
t
u
5
D
EM
D
e
mo
n
st
r
a
t
i
v
e
Ī
,
6
VM
V
e
r
b
F
i
n
i
t
e
Vē
g
a
v
ā
g
i
r
u
t
t
a
d
e
,
H
o
n
d
i
su
v
u
d
u
,
s
u
l
a
b
h
a
v
ā
y
i
t
u
,
ā
n
a
n
d
i
s
a
l
i
l
l
a
7
JJ
A
d
j
e
c
t
i
v
e
S
u
m
ā
r
u
,
H
o
sa
8
RB
A
d
v
e
r
b
h
o
r
a
t
u
9
QC
C
a
r
d
i
n
a
l
3
5
,
1
,
$
1
5
0
,
8
10
I
N
TF
I
n
t
e
n
s
i
f
i
e
r
At
i
11
S
Y
M
S
y
mb
o
l
,
!
.
T
h
e
r
esu
lts
p
r
esen
ted
in
T
ab
le
5
s
h
o
w
th
e
p
er
f
o
r
m
a
n
ce
ev
alu
atio
n
m
etr
ics,
in
clu
d
in
g
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all,
an
d
F
-
s
co
r
e,
f
o
r
v
ar
i
o
u
s
ML
m
o
d
els
u
s
ed
in
s
en
tim
en
t
an
al
y
s
is
.
E
x
tr
em
e
g
r
ad
ie
n
t
b
o
o
s
tin
g
(
XGBo
o
s
t
)
ac
h
iev
ed
an
ac
cu
r
ac
y
o
f
0
.
6
8
,
p
r
ec
is
io
n
o
f
0
.
6
7
,
r
ec
all
o
f
0
.
6
7
,
an
d
F
-
s
co
r
e
o
f
0
.
6
9
.
L
o
g
is
tic
r
eg
r
ess
io
n
(
L
R
)
d
em
o
n
s
tr
ated
s
im
ilar
p
er
f
o
r
m
an
ce
with
an
a
cc
u
r
ac
y
o
f
0
.
6
7
,
p
r
ec
is
io
n
o
f
0
.
6
8
,
r
ec
all
o
f
0
.
6
4
,
an
d
F
-
s
co
r
e
o
f
0
.
6
8
.
R
an
d
o
m
f
o
r
est
(
R
F)
h
ad
an
ac
cu
r
a
cy
o
f
0
.
6
6
,
p
r
ec
is
io
n
o
f
0
.
6
5
,
r
ec
all
o
f
0
.
6
8
,
an
d
F
-
s
co
r
e
o
f
0
.
6
2
.
Ad
aBo
o
s
t
an
d
g
r
ad
ien
t
b
o
o
s
tin
g
e
x
h
ib
ited
co
m
p
ar
ab
le
r
esu
lts
with
ac
cu
r
ac
y
s
co
r
es
o
f
0
.
6
7
an
d
0
.
6
9
,
r
esp
ec
tiv
ely
,
alo
n
g
with
p
r
ec
is
io
n
,
r
ec
all,
an
d
F
-
sc
o
r
e
v
alu
es
ar
o
u
n
d
0
.
6
5
-
0
.
6
8
.
T
h
e
v
o
tin
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
A
n
o
ve
l d
a
ta
s
et
a
n
d
p
a
r
t
-
of
-
s
p
ee
ch
ta
g
g
in
g
a
p
p
r
o
a
ch
fo
r
…
(
S
u
n
il Mu
g
a
lih
a
lli E
s
h
w
a
r
a
p
p
a
)
1669
en
s
em
b
le
m
o
d
el
s
h
o
wed
lo
we
r
p
er
f
o
r
m
an
ce
with
an
ac
cu
r
ac
y
o
f
0
.
5
8
an
d
p
r
ec
is
io
n
o
f
0
.
6
2
,
b
u
t
h
ig
h
er
r
ec
all
an
d
F
-
s
co
r
e
v
alu
es
at
0
.
6
8
an
d
0
.
6
4
,
r
esp
ec
tiv
ely
.
I
n
co
n
tr
a
s
t,
th
e
B
E
R
T
m
o
d
el
ac
h
iev
ed
s
ig
n
if
ican
tly
h
ig
h
e
r
p
er
f
o
r
m
an
ce
with
an
ac
cu
r
ac
y
o
f
0
.
8
1
,
p
r
ec
is
io
n
o
f
0
.
7
9
,
r
ec
all
o
f
0
.
8
,
an
d
F
-
s
co
r
e
o
f
0
.
8
2
.
No
ta
b
ly
,
th
e
p
r
o
p
o
s
ed
K
-
B
E
R
T
m
o
d
el
o
u
t
p
er
f
o
r
m
ed
all
o
th
er
m
o
d
els,
s
h
o
wca
s
in
g
ex
ce
p
tio
n
al
r
esu
lts
with
an
ac
c
u
r
ac
y
o
f
0
.
9
8
,
p
r
ec
is
io
n
o
f
0
.
9
7
,
r
ec
all
o
f
0
.
9
7
,
an
d
F
-
s
co
r
e
o
f
0
.
9
8
.
T
h
ese
f
in
d
in
g
s
h
ig
h
lig
h
t
th
e
s
u
p
er
io
r
p
e
r
f
o
r
m
an
ce
o
f
K
-
B
E
R
T
in
s
en
tim
en
t
an
al
y
s
is
ta
s
k
s
,
em
p
h
asizin
g
its
ef
f
ec
tiv
en
ess
in
ac
cu
r
ately
clas
s
i
f
y
in
g
s
en
tim
en
t
in
tex
t
d
ata
co
m
p
ar
ed
to
tr
ad
iti
o
n
al
ML
alg
o
r
ith
m
s
an
d
ev
e
n
th
e
B
E
R
T
m
o
d
el.
T
h
e
r
esu
lts
ar
e
g
r
ap
h
ically
s
h
o
wn
in
Fig
u
r
e
3
.
T
ab
le
5
.
Per
f
o
r
m
an
ce
e
v
alu
ati
o
n
M
o
d
e
l
s
A
c
c
u
r
a
c
y
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F
-
S
c
o
r
e
X
G
B
o
o
st
0
.
6
8
0
.
6
7
0
.
6
7
0
.
6
9
LR
0
.
6
7
0
.
6
8
0
.
6
4
0
.
6
8
RF
0
.
6
6
0
.
6
5
0
.
6
8
0
.
6
2
A
d
a
B
o
o
st
0
.
6
7
0
.
6
8
0
.
6
5
0
.
6
4
G
r
a
d
i
e
n
t
0
.
6
9
0
.
6
6
0
.
6
4
0
.
6
3
V
o
t
i
n
g
0
.
5
8
0
.
6
2
0
.
6
8
0
.
6
4
B
ER
T
0
.
8
1
0
.
7
9
0
.
8
0
.
8
2
K
-
B
E
R
T
0
.
9
8
0
.
9
7
0
.
9
7
0
.
9
8
Fig
u
r
e
3
.
Per
f
o
r
m
an
c
e
e
v
alu
at
io
n
5.
CO
NCLU
SI
O
N
I
n
c
o
n
clu
s
io
n
,
th
is
wo
r
k
h
a
s
m
ad
e
s
ig
n
i
f
ican
t
im
p
r
o
v
e
m
en
ts
in
a
d
v
an
ci
n
g
s
en
tim
en
t
an
aly
s
is
ca
p
ab
ilit
ies
f
o
r
th
e
Ka
n
n
ad
a
l
an
g
u
ag
e
.
B
y
in
tr
o
d
u
cin
g
th
e
K
-
B
E
R
T
m
o
d
el
an
d
le
v
er
ag
in
g
ad
v
an
ce
d
ML
an
d
NL
P
tech
n
iq
u
es,
we
h
av
e
a
d
d
r
ess
ed
th
e
ch
allen
g
es
p
o
s
ed
b
y
lin
g
u
is
tic
v
ar
iatio
n
s
,
c
u
lt
u
r
al
n
u
a
n
ce
s
,
an
d
lim
ited
lab
elled
d
atasets
.
T
h
e
p
er
f
o
r
m
an
ce
ev
alu
atio
n
r
esu
lts
d
em
o
n
s
tr
ate
th
at
th
e
K
-
B
E
R
T
m
o
d
el
o
u
tp
er
f
o
r
m
s
tr
ad
itio
n
al
ML
a
lg
o
r
ith
m
s
,
in
clu
d
i
n
g
XGBo
o
s
t,
LR
,
RF
,
Ad
a
B
o
o
s
t,
an
d
g
r
ad
ien
t
b
o
o
s
tin
g
,
as
well
as
th
e
B
E
R
T
m
o
d
el.
W
it
h
an
ex
ce
p
tio
n
al
ac
cu
r
ac
y
o
f
0
.
9
8
,
p
r
ec
is
io
n
o
f
0
.
9
7
,
r
ec
all
o
f
0
.
9
7
,
a
n
d
F
-
s
co
r
e
o
f
0
.
9
8
,
th
e
K
-
B
E
R
T
m
o
d
el
s
h
o
wca
s
es
its
ef
f
ec
tiv
en
ess
in
ac
cu
r
ately
class
if
y
in
g
s
en
tim
en
t
in
Ka
n
n
ad
a
tex
t
d
ata.
T
h
is
wo
r
k
n
o
t
o
n
ly
c
o
n
tr
ib
u
tes
a
n
o
v
el
d
ataset
d
er
iv
e
d
f
r
o
m
Sem
E
v
al
2
0
1
4
t
a
s
k
4
f
o
r
Kan
n
ad
a
s
en
tim
en
t
an
aly
s
is
b
u
t
also
i
n
tr
o
d
u
ce
s
a
r
o
b
u
s
t
m
o
d
el
s
p
ec
if
ically
d
esig
n
ed
f
o
r
Kan
n
a
d
a,
p
av
in
g
th
e
way
f
o
r
f
u
r
th
er
ad
v
a
n
ce
m
en
ts
in
NL
P
r
esear
ch
an
d
ap
p
licatio
n
s
d
esig
n
ed
f
o
r
t
h
e
lin
g
u
is
tic
d
iv
er
s
ity
o
f
I
n
d
ia.
C
o
llab
o
r
ativ
e
ef
f
o
r
ts
to
war
d
s
d
ataset
cr
ea
tio
n
,
m
o
d
el
d
ev
elo
p
m
en
t,
an
d
ev
alu
atio
n
m
eth
o
d
o
l
o
g
ies
ar
e
ess
en
tial f
o
r
en
h
an
cin
g
s
en
tim
en
t a
n
aly
s
is
ca
p
ab
ilit
ies
in
m
u
ltil
in
g
u
al
en
v
ir
o
n
m
e
n
ts
an
d
d
r
iv
in
g
in
n
o
v
atio
n
in
co
m
p
u
tatio
n
al
lin
g
u
is
tics
.
Fo
r
f
u
tu
r
e
w
o
r
k
,
t
h
e
B
E
R
T
m
o
d
el
ca
n
b
e
f
u
r
th
er
en
h
a
n
ce
d
f
o
r
ac
h
iev
in
g
b
ette
r
r
esu
lts
an
d
co
m
p
a
r
ed
with
o
t
h
er
d
atasets
.
0
0
.
1
0
.
2
0
.
3
0
.
4
0
.
5
0
.
6
0
.
7
0
.
8
0
.
9
1
Ac
c
u
r
a
c
y
Pre
c
isio
n
Re
c
a
ll
F-S
c
o
r
e
(
%
)
X
G
Bo
o
s
t
LR
RF
A
da
Bo
o
s
t
G
r
a
di
e
nt
V
o
ti
ng
BE
R
T
K
-BE
R
T
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
In
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
3
7
,
No
.
3
,
Ma
r
ch
20
2
5
:
1
661
-
1
6
7
1
1670
RE
F
E
R
E
NC
E
S
[
1
]
K
.
R
.
M
a
b
o
k
e
l
a
,
T.
C
e
l
i
k
,
a
n
d
M
.
R
a
b
o
r
i
f
e
,
“
M
u
l
t
i
l
i
n
g
u
a
l
se
n
t
i
me
n
t
a
n
a
l
y
si
s
f
o
r
u
n
d
e
r
-
r
e
so
u
r
c
e
d
l
a
n
g
u
a
g
e
s
:
a
s
y
st
e
ma
t
i
c
r
e
v
i
e
w
o
f
t
h
e
l
a
n
d
s
c
a
p
e
,
”
I
EE
E
A
c
c
e
ss
,
v
o
l
.
1
1
,
p
p
.
1
5
9
9
6
–
1
6
0
2
0
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
2
.
3
2
2
4
1
3
6
.
[
2
]
M
.
Z.
A
n
sar
i
,
M
.
B
.
A
z
i
z
,
M
.
O
.
S
i
d
d
i
q
u
i
,
H
.
M
e
h
r
a
,
a
n
d
K
.
P
.
S
i
n
g
h
,
“
A
n
a
l
y
s
i
s
o
f
p
o
l
i
t
i
c
a
l
se
n
t
i
m
e
n
t
o
r
i
e
n
t
a
t
i
o
n
s
o
n
t
w
i
t
t
e
r
,
”
Pro
c
e
d
i
a
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
1
6
7
,
p
p
.
1
8
2
1
–
1
8
2
8
,
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
p
r
o
c
s.
2
0
2
0
.
0
3
.
2
0
1
.
[
3
]
M
.
K
a
u
r
,
K
.
Jo
s
h
i
,
B
.
G
o
y
a
l
,
a
n
d
A
.
D
o
g
r
a
,
“
A
n
a
p
p
r
o
a
c
h
t
o
p
e
r
f
o
r
m
s
e
n
t
i
me
n
t
a
n
a
l
y
si
s
u
s
i
n
g
d
a
t
a
m
i
n
i
n
g
a
l
g
o
r
i
t
h
ms
,
”
i
n
2
0
2
3
2
n
d
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
Ed
g
e
C
o
m
p
u
t
i
n
g
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
(
I
C
EC
AA)
,
Ju
l
.
2
0
2
3
,
p
p
.
8
0
3
–
8
0
8
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
EC
A
A
5
8
1
0
4
.
2
0
2
3
.
1
0
2
1
2
4
0
4
.
[
4
]
J.
R
.
Ji
m
,
M
.
A
.
R
.
Ta
l
u
k
d
e
r
,
P
.
M
a
l
a
k
a
r
,
M
.
M
.
K
a
b
i
r
,
K
.
N
u
r
,
a
n
d
M
.
F
.
M
r
i
d
h
a
,
“
R
e
c
e
n
t
a
d
v
a
n
c
e
me
n
t
s
a
n
d
c
h
a
l
l
e
n
g
e
s
o
f
N
LP
-
b
a
s
e
d
se
n
t
i
me
n
t
a
n
a
l
y
si
s
:
a
s
t
a
t
e
-
of
-
t
h
e
-
a
r
t
r
e
v
i
e
w
,
”
N
a
t
u
r
a
l
L
a
n
g
u
a
g
e
Pr
o
c
e
ssi
n
g
J
o
u
rn
a
l
,
v
o
l
.
6
,
p
.
1
0
0
0
5
9
,
M
a
r
.
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
n
l
p
.
2
0
2
4
.
1
0
0
0
5
9
.
[
5
]
P
.
N
a
n
d
w
a
n
i
a
n
d
R
.
V
e
r
ma,
“
A
r
e
v
i
e
w
o
n
se
n
t
i
me
n
t
a
n
a
l
y
s
i
s
a
n
d
e
m
o
t
i
o
n
d
e
t
e
c
t
i
o
n
f
r
o
m
t
e
x
t
,
”
S
o
c
i
a
l
N
e
t
w
o
rk
An
a
l
y
si
s
a
n
d
Mi
n
i
n
g
,
v
o
l
.
1
1
,
n
o
.
1
,
p
.
8
1
,
D
e
c
.
2
0
2
1
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
3
2
7
8
-
0
2
1
-
0
0
7
7
6
-
6.
[
6
]
V
.
J
o
sh
i
,
S
.
P
a
t
e
l
,
R
.
A
g
a
r
w
a
l
,
a
n
d
H
.
A
r
o
r
a
,
“
S
e
n
t
i
m
e
n
t
s
a
n
a
l
y
s
i
s
u
si
n
g
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
m
s,”
i
n
2
0
2
3
S
e
c
o
n
d
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
El
e
c
t
r
o
n
i
c
s
a
n
d
R
e
n
e
w
a
b
l
e
S
y
s
t
e
m
s
(
I
C
EAR
S
)
,
M
a
r
.
2
0
2
3
,
p
p
.
1
4
2
5
–
1
4
2
9
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
EA
R
S
5
6
3
9
2
.
2
0
2
3
.
1
0
0
8
5
4
3
2
.
[
7
]
M
.
A
r
u
m
u
g
a
m,
S
.
S
R
,
a
n
d
C
.
Ja
y
a
n
t
h
i
,
“
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
f
o
r
se
n
t
i
men
t
a
n
a
l
y
s
i
s
u
t
i
l
i
z
i
n
g
s
o
c
i
a
l
m
e
d
i
a
,
”
i
n
2
0
2
3
2
n
d
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
E
d
g
e
C
o
m
p
u
t
i
n
g
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
(
I
C
E
C
AA)
,
J
u
l
.
2
0
2
3
,
p
p
.
5
2
3
–
5
3
0
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
EC
A
A
5
8
1
0
4
.
2
0
2
3
.
1
0
2
1
2
1
3
5
.
[
8
]
M
.
R
o
d
r
í
g
u
e
z
-
I
b
á
n
e
z
,
A
.
C
a
sá
n
e
z
-
V
e
n
t
u
r
a
,
F
.
C
a
st
e
j
ó
n
-
M
a
t
e
o
s,
a
n
d
P
.
-
M
.
C
u
e
n
c
a
-
J
i
mé
n
e
z
,
“
A
r
e
v
i
e
w
o
n
s
e
n
t
i
m
e
n
t
a
n
a
l
y
si
s
f
r
o
m
so
c
i
a
l
m
e
d
i
a
p
l
a
t
f
o
r
ms,”
Ex
p
e
rt
S
y
st
e
m
s
w
i
t
h
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
2
2
3
,
p
.
1
1
9
8
6
2
,
A
u
g
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
e
sw
a
.
2
0
2
3
.
1
1
9
8
6
2
.
[
9
]
G
.
M
a
n
i
a
s,
A
.
M
a
v
r
o
g
i
o
r
g
o
u
,
A
.
K
i
o
u
r
t
i
s,
C
.
S
y
m
v
o
u
l
i
d
i
s
,
a
n
d
D
.
K
y
r
i
a
z
i
s
,
“
M
u
l
t
i
l
i
n
g
u
a
l
t
e
x
t
c
a
t
e
g
o
r
i
z
a
t
i
o
n
a
n
d
se
n
t
i
me
n
t
a
n
a
l
y
si
s
:
a
c
o
mp
a
r
a
t
i
v
e
a
n
a
l
y
s
i
s
o
f
t
h
e
u
t
i
l
i
z
a
t
i
o
n
o
f
mu
l
t
i
l
i
n
g
u
a
l
a
p
p
r
o
a
c
h
e
s
f
o
r
c
l
a
ss
i
f
y
i
n
g
t
w
i
t
t
e
r
d
a
t
a
,
”
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
5
,
n
o
.
2
9
,
p
p
.
2
1
4
1
5
–
2
1
4
3
1
,
O
c
t
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
7
/
s
0
0
5
2
1
-
023
-
0
8
6
2
9
-
3.
[
1
0
]
V
.
U
maran
i
,
A
.
J
u
l
i
a
n
,
a
n
d
J.
D
e
e
p
a
,
“
S
e
n
t
i
m
e
n
t
a
n
a
l
y
si
s
u
si
n
g
v
a
r
i
o
u
s
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
a
n
d
d
e
e
p
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s,
”
J
o
u
r
n
a
l
o
f
t
h
e
N
i
g
e
r
i
a
n
S
o
c
i
e
t
y
o
f
P
h
y
s
i
c
a
l
S
c
i
e
n
c
e
s
,
p
p
.
3
8
5
–
3
9
4
,
N
o
v
.
2
0
2
1
,
d
o
i
:
1
0
.
4
6
4
8
1
/
j
n
sp
s
.
2
0
2
1
.
3
0
8
.
[
1
1
]
Y
.
G
a
r
a
n
i
,
S
.
Jo
s
h
i
,
a
n
d
S
.
K
u
l
k
a
r
n
i
,
“
O
f
f
e
n
si
v
e
se
n
t
i
m
e
n
t
d
e
t
e
c
t
i
o
n
w
i
t
h
C
h
a
t
G
P
T
a
n
d
o
t
h
e
r
t
r
a
n
sf
o
r
m
e
r
s
i
n
K
a
n
n
a
d
a
,
”
i
n
2
0
2
3
I
EEE
2
n
d
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
D
a
t
a
,
D
e
c
i
si
o
n
a
n
d
S
y
s
t
e
m
s
(
I
C
D
D
S
)
,
D
e
c
.
2
0
2
3
,
p
p
.
1
–
6
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
D
D
S
5
9
1
3
7
.
2
0
2
3
.
1
0
4
3
4
6
8
4
.
[
1
2
]
B
.
R
.
C
h
a
k
r
a
v
a
r
t
h
i
e
t
a
l
.
,
“
D
r
a
v
i
d
i
a
n
C
o
d
e
M
i
x
:
se
n
t
i
me
n
t
a
n
a
l
y
si
s
a
n
d
o
f
f
e
n
si
v
e
l
a
n
g
u
a
g
e
i
d
e
n
t
i
f
i
c
a
t
i
o
n
d
a
t
a
set
f
o
r
D
r
a
v
i
d
i
a
n
l
a
n
g
u
a
g
e
s
i
n
c
o
d
e
-
m
i
x
e
d
t
e
x
t
,
”
L
a
n
g
u
a
g
e
Re
s
o
u
rc
e
s
a
n
d
E
v
a
l
u
a
t
i
o
n
,
v
o
l
.
5
6
,
n
o
.
3
,
p
p
.
7
6
5
–
8
0
6
,
S
e
p
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
0
5
7
9
-
0
2
2
-
0
9
5
8
3
-
7.
[
1
3
]
R
.
C
h
u
n
d
i
,
V
.
R
.
H
u
l
i
p
a
l
l
e
d
,
a
n
d
J
.
.
S
i
m
h
a
,
“
S
A
E
K
C
S
:
s
e
n
t
i
m
e
n
t
a
n
a
l
y
s
i
s
f
o
r
E
n
g
l
i
sh
–
K
a
n
n
a
d
a
c
o
d
e
sw
i
t
c
h
t
e
x
t
u
s
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s,”
i
n
2
0
2
0
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
S
m
a
rt
T
e
c
h
n
o
l
o
g
i
e
s
i
n
C
o
m
p
u
t
i
n
g
,
El
e
c
t
ri
c
a
l
a
n
d
El
e
c
t
ro
n
i
c
s
(
I
C
S
T
C
EE)
,
O
c
t
.
2
0
2
0
,
p
p
.
3
2
7
–
3
3
1
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
S
TC
EE
4
9
6
3
7
.
2
0
2
0
.
9
2
7
7
0
3
0
.
[
1
4
]
P
.
R
a
n
j
i
t
h
a
a
n
d
K
.
N
.
B
h
a
n
u
,
“
I
mp
r
o
v
e
d
se
n
t
i
me
n
t
a
n
a
l
y
si
s
f
o
r
d
r
a
v
i
d
i
a
n
l
a
n
g
u
a
g
e
-
k
a
n
n
a
d
a
u
s
i
n
g
d
i
c
i
si
o
n
t
r
e
e
a
l
g
o
r
i
t
h
m
w
i
t
h
e
f
f
i
c
i
e
n
t
d
a
t
a
d
i
c
t
i
o
n
a
r
Y
,
”
I
O
P
C
o
n
f
e
re
n
c
e
S
e
ri
e
s:
M
a
t
e
ri
a
l
s
S
c
i
e
n
c
e
a
n
d
En
g
i
n
e
e
ri
n
g
,
v
o
l
.
1
1
2
3
,
n
o
.
1
,
p
.
0
1
2
0
3
9
,
A
p
r
.
2
0
2
1
,
d
o
i
:
1
0
.
1
0
8
8
/
1
7
5
7
-
8
9
9
X
/
1
1
2
3
/
1
/
0
1
2
0
3
9
.
[
1
5
]
M
.
E.
S
u
n
i
l
a
n
d
S
.
V
i
n
a
y
,
“
K
a
n
n
a
d
a
sen
t
i
m
e
n
t
a
n
a
l
y
si
s
u
s
i
n
g
v
e
c
t
o
r
i
z
a
t
i
o
n
a
n
d
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
,
”
i
n
S
e
n
t
i
m
e
n
t
a
l
A
n
a
l
y
s
i
s
a
n
d
D
e
e
p
L
e
a
rn
i
n
g
:
Pro
c
e
e
d
i
n
g
s
o
f
I
C
S
A
D
L
2
0
2
1
,
2
0
2
2
,
p
p
.
6
7
7
–
6
8
9
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
9
8
1
-
16
-
5
1
5
7
-
1
_
5
3
.
[
1
6
]
S
.
S
h
e
t
t
y
e
t
a
l
.
,
“
S
e
n
t
i
m
e
n
t
a
n
a
l
y
si
s
o
f
t
w
i
t
t
e
r
p
o
st
s
i
n
En
g
l
i
s
h
,
K
a
n
n
a
d
a
a
n
d
H
i
n
d
i
l
a
n
g
u
a
g
e
s
,
”
i
n
Re
c
e
n
t
A
d
v
a
n
c
e
s
i
n
Ar
t
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
a
n
d
D
a
t
a
En
g
i
n
e
e
r
i
n
g
:
S
e
l
e
c
t
P
ro
c
e
e
d
i
n
g
s
o
f
A
I
D
E
2
0
2
0
,
2
0
2
2
,
p
p
.
3
6
1
–
3
7
5
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
981
-
16
-
3342
-
3
_
2
9
.
[
1
7
]
K
.
S
h
a
n
m
u
g
a
v
a
d
i
v
e
l
,
V
.
E.
S
a
t
h
i
s
h
k
u
mar,
S
.
R
a
j
a
,
T
.
B
.
Li
n
g
a
i
a
h
,
S
.
N
e
e
l
a
k
a
n
d
a
n
,
a
n
d
M
.
S
u
b
r
a
ma
n
i
a
n
,
“
D
e
e
p
l
e
a
r
n
i
n
g
b
a
se
d
sen
t
i
m
e
n
t
a
n
a
l
y
si
s
a
n
d
o
f
f
e
n
s
i
v
e
l
a
n
g
u
a
g
e
i
d
e
n
t
i
f
i
c
a
t
i
o
n
o
n
m
u
l
t
i
l
i
n
g
u
a
l
c
o
d
e
-
m
i
x
e
d
d
a
t
a
,
”
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
5
5
7
,
D
e
c
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
3
8
/
s
4
1
5
9
8
-
022
-
2
6
0
9
2
-
3.
[
1
8
]
R
.
C
h
u
n
d
i
,
V
.
R
.
H
u
l
i
p
a
l
l
e
d
,
a
n
d
J
.
B
.
S
i
mh
a
,
“
N
B
Le
x
:
e
m
o
t
i
o
n
p
r
e
d
i
c
t
i
o
n
i
n
K
a
n
n
a
d
a
-
En
g
l
i
s
h
c
o
d
e
-
sw
i
t
c
h
t
e
x
t
u
si
n
g
n
a
ï
v
e
b
a
y
e
s
l
e
x
i
c
o
n
a
p
p
r
o
a
c
h
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
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
En
g
i
n
e
e
ri
n
g
(
I
J
E
C
E)
,
v
o
l
.
1
3
,
n
o
.
2
,
p
p
.
2
0
6
8
–
2
0
7
7
,
A
p
r
.
2
0
2
3
,
d
o
i
:
1
0
.
1
1
5
9
1
/
i
j
e
c
e
.
v
1
3
i
2
.
p
p
2
0
6
8
-
2
0
7
7
.
[
1
9
]
P
.
K
.
R
o
y
,
“
D
e
e
p
e
n
sem
b
l
e
n
e
t
w
o
r
k
f
o
r
se
n
t
i
me
n
t
a
n
a
l
y
si
s
i
n
b
i
-
l
i
n
g
u
a
l
l
o
w
-
r
e
so
u
r
c
e
l
a
n
g
u
a
g
e
s,
”
A
C
M
T
r
a
n
s
a
c
t
i
o
n
s
o
n
As
i
a
n
a
n
d
L
o
w
-
Re
s
o
u
r
c
e
L
a
n
g
u
a
g
e
I
n
f
o
rm
a
t
i
o
n
Pr
o
c
e
s
si
n
g
,
v
o
l
.
2
3
,
n
o
.
1
,
p
p
.
1
–
1
6
,
Ja
n
.
2
0
2
4
,
d
o
i
:
1
0
.
1
1
4
5
/
3
6
0
0
2
2
9
.
[
2
0
]
R
.
C
h
u
n
d
i
,
V
.
R
.
H
u
l
i
p
a
l
l
e
d
,
a
n
d
J
.
B
h
a
r
t
h
i
sh
S
i
m
h
a
,
“
L
e
x
i
c
o
n
-
b
a
se
d
se
n
t
i
men
t
a
n
a
l
y
si
s
f
o
r
K
a
n
n
a
d
a
-
E
n
g
l
i
s
h
c
o
d
e
-
sw
i
t
c
h
t
e
x
t
,
”
I
AES
I
n
t
e
rn
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
Art
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
(
I
J
-
AI
)
,
v
o
l
.
1
2
,
n
o
.
3
,
p
p
.
1
5
0
0
–
1
5
0
7
,
S
e
p
.
2
0
2
3
,
d
o
i
:
1
0
.
1
1
5
9
1
/
i
j
a
i
.
v
1
2
.
i
3
.
p
p
1
5
0
0
-
1
5
0
7
.
[
2
1
]
R
.
C
h
u
n
d
i
,
V
.
R
.
H
u
l
i
p
a
l
l
e
d
,
a
n
d
J.
B
.
S
i
mh
a
,
“
I
d
e
n
t
i
f
i
c
a
t
i
o
n
o
f
m
o
n
o
l
i
n
g
u
a
l
a
n
d
c
o
d
e
-
sw
i
t
c
h
i
n
f
o
r
m
a
t
i
o
n
f
r
o
m
En
g
l
i
sh
-
K
a
n
n
a
d
a
c
o
d
e
-
sw
i
t
c
h
d
a
t
a
,
”
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
r
i
c
a
l
a
n
d
C
o
m
p
u
t
e
r
En
g
i
n
e
e
ri
n
g
(
I
J
EC
E)
,
v
o
l
.
1
3
,
n
o
.
5
,
p
p
.
5
6
3
2
–
5
6
4
0
,
O
c
t
.
2
0
2
3
,
d
o
i
:
1
0
.
1
1
5
9
1
/
i
j
e
c
e
.
v
1
3
i
5
.
p
p
5
6
3
2
-
5
6
4
0
.
[
2
2
]
R
.
S
h
a
n
k
a
r
,
S
.
S
w
a
my
,
a
n
d
S
.
H
e
g
d
e
,
“
E
x
p
l
o
r
i
n
g
se
n
t
i
me
n
t
a
n
a
l
y
si
s
i
n
K
a
n
n
a
d
a
l
a
n
g
u
a
g
e
:
a
c
o
mp
r
e
h
e
n
s
i
v
e
st
u
d
y
o
n
C
O
V
I
D
-
19
d
a
t
a
u
si
n
g
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
a
n
d
e
n
sem
b
l
e
a
l
g
o
r
i
t
h
ms
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
I
n
t
e
l
l
i
g
e
n
t
S
y
s
t
e
m
s
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
i
n
En
g
i
n
e
e
ri
n
g
,
v
o
l
.
1
2
,
n
o
.
1
1
,
p
p
.
2
1
–
2
9
,
2
0
2
4
.
[
2
3
]
M
.
P
o
n
t
i
k
i
,
D
.
G
a
l
a
n
i
s,
J.
P
a
v
l
o
p
o
u
l
o
s,
H
.
P
a
p
a
g
e
o
r
g
i
o
u
,
I
.
A
n
d
r
o
u
t
so
p
o
u
l
o
s,
a
n
d
S
.
M
a
n
a
n
d
h
a
r
,
“
S
e
mE
v
a
l
-
2
0
1
4
Ta
sk
4
:
a
s
p
e
c
t
b
a
s
e
d
s
e
n
t
i
m
e
n
t
a
n
a
l
y
s
i
s,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
8
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
W
o
rks
h
o
p
o
n
S
e
m
a
n
t
i
c
E
v
a
l
u
a
t
i
o
n
(
S
e
m
E
v
a
l
2
0
1
4
)
,
2
0
1
4
,
p
p
.
27
–
3
5
,
d
o
i
:
1
0
.
3
1
1
5
/
v
1
/
S
1
4
-
2
0
0
4
.
[
2
4
]
T.
B
r
a
n
t
s,
“
T
n
T
-
a
st
a
t
i
st
i
c
a
l
p
a
r
t
-
of
-
sp
e
e
c
h
t
a
g
g
e
r
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
s
i
x
t
h
c
o
n
f
e
r
e
n
c
e
o
n
A
p
p
l
i
e
d
n
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
ro
c
e
ss
i
n
g
,
2
0
0
0
,
p
p
.
2
2
4
–
2
3
1
,
d
o
i
:
1
0
.
3
1
1
5
/
9
7
4
1
4
7
.
9
7
4
1
7
8
.
[
2
5
]
T.
H
a
r
i
y
a
n
t
i
,
S
.
A
i
d
a
,
a
n
d
H
.
K
a
me
d
a
,
“
S
a
m
a
w
a
l
a
n
g
u
a
g
e
p
a
r
t
o
f
sp
e
e
c
h
t
a
g
g
i
n
g
w
i
t
h
p
r
o
b
a
b
i
l
i
st
i
c
a
p
p
r
o
a
c
h
:
c
o
m
p
a
r
i
so
n
o
f
u
n
i
g
r
a
m,
H
M
M
a
n
d
T
n
T
m
o
d
e
l
s,
”
J
o
u
r
n
a
l
o
f
P
h
y
s
i
c
s:
C
o
n
f
e
re
n
c
e
S
e
ri
e
s
,
v
o
l
.
1
2
3
5
,
n
o
.
1
,
p
.
0
1
2
0
1
3
,
Ju
n
.
2
0
1
9
,
d
o
i
:
1
0
.
1
0
8
8
/
1
7
4
2
-
6
5
9
6
/
1
2
3
5
/
1
/
0
1
2
0
1
3
.
[
2
6
]
J.
D
e
v
l
i
n
,
M
.
-
W
.
C
h
a
n
g
,
K
.
L
e
e
,
a
n
d
K
.
T
o
u
t
a
n
o
v
a
,
“
B
ER
T
:
p
r
e
-
t
r
a
i
n
i
n
g
o
f
d
e
e
p
b
i
d
i
r
e
c
t
i
o
n
a
l
t
r
a
n
sf
o
r
m
e
r
s
f
o
r
l
a
n
g
u
a
g
e
u
n
d
e
r
s
t
a
n
d
i
n
g
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
2
0
1
9
C
o
n
f
e
r
e
n
c
e
o
f
t
h
e
N
o
rt
h
,
2
0
1
9
,
p
p
.
4
1
7
1
–
4
1
8
6
,
d
o
i
:
1
0
.
1
8
6
5
3
/
v
1
/
N
1
9
-
1
4
2
3
.
Evaluation Warning : The document was created with Spire.PDF for Python.