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
8
1
5
~1
829
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
8
1
5
-
1
8
2
9
1815
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ee
cs
.
ia
esco
r
e.
co
m
A review
bas
ed o
n sentime
ntal a
na
ly
sis
f
o
r H
in
di lan
g
ua
g
e
Va
ibh
a
v
P
ra
ka
s
h Va
s
a
ni,
As
ha
Am
bh
a
ik
a
r
D
e
p
a
r
t
me
n
t
o
f
C
o
mp
u
t
e
r
S
c
i
e
n
c
e
a
n
d
En
g
i
n
e
e
r
i
n
g
,
K
a
l
i
n
g
a
U
n
i
v
e
r
si
t
y
,
N
a
y
a
R
a
i
p
u
r
,
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
Feb
2
8
,
2
0
2
4
R
ev
is
ed
Oct
2
,
2024
Acc
ep
ted
Oct
7
,
2
0
2
4
Th
e
'
M
o
th
e
r
T
o
n
g
u
e
'
o
f
I
n
d
ia
is
c
o
n
sid
e
re
d
t
o
b
e
Hi
n
d
i
.
Th
e
d
e
m
a
n
d
fo
r
Hin
d
i
c
o
n
ten
t
rise
s
a
s
th
e
In
d
ian
p
o
p
u
lati
o
n
g
ro
ws
.
Hin
d
i
is
ty
p
i
c
a
ll
y
u
se
d
wh
e
n
I
n
d
ian
s
e
x
p
re
ss
th
e
ir
o
p
in
i
o
n
s
a
b
o
u
t
so
m
e
th
i
n
g
.
T
h
is
g
e
n
e
ra
tes
d
a
ta
in
Hin
d
i
t
h
a
t
m
a
y
late
r
b
e
a
n
a
l
y
se
d
.
S
e
n
ti
m
e
n
t
a
n
a
l
y
sis
(S
A)
is
a
c
a
t
e
g
o
ry
f
o
r
o
n
e
o
f
t
h
e
a
n
a
ly
se
s.
S
A
e
x
a
m
in
e
s
h
o
w
a
sp
e
a
k
e
r
o
r
a
u
th
o
r'
s
e
m
o
ti
o
n
s,
se
n
ti
m
e
n
ts,
a
n
d
a
tt
it
u
d
e
s
a
re
e
x
p
r
e
ss
e
d
in
a
p
a
rti
c
u
lar
tex
t.
S
A,
o
c
c
a
sio
n
a
ll
y
re
fe
rre
d
to
a
s
"
Op
in
io
n
M
in
in
g
,
"
is
a
so
rt
o
f
c
o
n
tex
t
u
a
l
m
in
in
g
t
h
a
t
u
se
s
a
lg
o
rit
h
m
ic
re
c
o
g
n
it
i
o
n
a
n
d
c
a
teg
o
riza
ti
o
n
t
o
d
e
tec
t
th
e
v
iew
p
o
in
ts
e
x
p
re
ss
e
d
in
a
tex
t
a
n
d
d
e
term
i
n
e
wh
e
th
e
r
th
e
y
a
re
p
o
siti
v
e
,
n
e
g
a
ti
v
e
,
o
r
n
e
u
tral.
W
o
rd
p
o
larity
,
t
h
a
t
S
A
g
iv
e
s,
e
n
a
b
les
u
s
to
a
ss
e
ss
th
e
tex
t'
s
imp
a
c
t
a
n
d
d
e
c
id
e
if
it
is
g
o
o
d
o
r
n
e
g
a
ti
v
e
.
It
is
p
o
ss
ib
le
to
imp
lem
e
n
t
S
A
u
sin
g
a
v
a
riety
o
f
m
e
th
o
d
s.
T
h
e
re
wa
s
a
lo
t
o
f
stu
d
y
o
n
S
A
i
n
t
h
e
En
g
li
sh
lan
g
u
a
g
e
,
b
u
t
n
o
t
a
s
m
u
c
h
f
o
r
Hi
n
d
i
.
In
t
h
is
e
ss
a
y
,
S
A
st
u
d
ies
re
g
a
r
d
in
g
th
e
Hin
d
i
lan
g
u
a
g
e
a
re
re
v
iew
e
d
a
n
d
a
n
a
ly
s
e
d
.
K
ey
w
o
r
d
s
:
Hin
d
i te
x
t
L
in
g
u
is
tic
Natu
r
al
lan
g
u
ag
e
p
r
o
ce
s
s
in
g
Op
in
io
n
m
in
i
n
g
Sen
tim
en
tal
an
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
:
Vaib
h
av
Pra
k
ash
Vasan
i
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
,
Kalin
g
a
Un
iv
er
s
ity
Nea
r
Ma
n
tr
alay
a,
Nay
a
R
aip
u
r
-
4
9
2
1
0
1
,
C
h
h
attis
g
ar
h
,
I
n
d
ia
E
m
ail: v
aib
h
av
.
p
r
ak
ash
@
k
alin
g
au
n
i
v
er
s
ity
.
ac
.
in
1.
I
NT
RO
D
UCT
I
O
N
I
n
n
atu
r
al
lan
g
u
a
g
e
p
r
o
ce
s
s
in
g
(
NL
P),
wh
ich
is
t
h
e
ac
t
o
f
ex
am
in
in
g
an
d
s
u
m
m
a
r
is
in
g
co
n
s
u
m
er
s
'
v
iews
an
d
f
ee
lin
g
s
as
co
n
v
ey
ed
in
a
p
h
r
ase,
s
en
tim
en
t
an
aly
s
is
(
SA)
[1
]
-
[
4
]
is
a
cr
u
cial
d
u
ty
.
I
t
is
em
p
lo
y
e
d
to
id
en
tify
wh
eth
er
a
tex
t'
s
u
n
d
er
ly
in
g
s
en
tim
en
t
is
n
e
u
tr
al
,
p
o
s
itiv
e,
o
r
n
eg
ativ
e.
SA
h
a
s
d
r
awn
in
ter
est
in
b
o
th
in
d
u
s
tr
y
an
d
ac
a
d
em
ics,
esp
ec
ially
wh
en
it
co
m
es
to
d
eter
m
in
in
g
co
n
s
u
m
er
h
a
p
p
in
e
s
s
with
g
o
o
d
s
an
d
s
er
v
ic
es.
An
en
o
r
m
o
u
s
am
o
u
n
t
o
f
m
a
r
k
etin
g
f
i
r
m
s
an
d
s
y
s
tem
s
f
o
r
s
u
g
g
esti
o
n
s
ar
e
wo
r
k
in
g
to
r
ec
o
g
n
ize
p
eo
p
le'
s
lik
es
an
d
d
is
lik
es
o
f
th
is
m
ater
ial.
T
h
e
f
o
u
r
th
lar
g
est
co
m
m
o
n
ly
u
tili
ze
d
lan
g
u
a
g
e
in
t
h
e
wo
r
ld
is
Hin
d
i.
T
h
e
ex
p
an
d
in
g
v
o
lu
m
e
o
f
u
s
er
-
g
en
er
ate
d
co
n
ten
t
o
n
th
e
in
ter
n
et
is
wh
at
p
r
o
m
p
ted
th
e
SA
r
esear
ch
.
A
lar
g
e
p
o
r
tio
n
o
f
th
e
ex
tan
t
r
esear
ch
o
n
th
is
to
p
ic
is
in
E
n
g
lis
h
.
L
ittl
e
ef
f
o
r
t
h
as
b
ee
n
d
ed
icate
d
to
SA in
th
e
Hin
d
i la
n
g
u
ag
e.
T
h
e
s
u
b
s
tan
ce
o
f
in
f
o
r
m
atio
n
in
H
in
d
i m
u
s
t b
e
ass
ess
ed
f
o
r
in
d
u
s
tr
ial
u
s
ag
e
.
I
n
d
ia
is
a
m
u
ltil
in
g
u
al
co
u
n
tr
y
with
a
wid
e
r
an
g
e
o
f
lin
g
u
i
s
tic
an
d
cu
ltu
r
al
tr
a
d
itio
n
s
.
T
h
er
e
ar
e
2
2
o
f
f
icial
lan
g
u
ag
es
in
th
e
co
u
n
tr
y
.
Ho
wev
er
,
ex
ce
p
t
a
f
ew
p
ap
er
s
s
u
ch
as
[
5
]
,
[
6
]
th
er
e
h
asn
'
t
b
ee
n
a
lo
t
o
f
s
tu
d
y
o
n
SA
in
I
n
d
ian
lan
g
u
ag
es.
T
h
ese
p
r
e
v
i
o
u
s
ef
f
o
r
ts
,
h
o
wev
er
,
d
o
n
o
t
h
an
d
le
f
in
e
-
g
r
ain
ed
SA
at
th
e
asp
ec
t le
v
el.
Sen
tim
en
tal
an
aly
s
is
p
r
o
v
id
e
d
two
asp
ec
ts
n
am
ely
o
p
in
i
o
n
-
b
ased
a
n
d
em
o
tio
n
al.
Fir
s
t,
o
p
in
io
n
m
in
in
g
d
e
p
en
d
s
u
p
o
n
tex
t
p
o
lar
ity
th
is
wo
u
ld
p
er
f
o
r
m
t
h
e
tex
t
class
if
icatio
n
in
to
p
o
s
itiv
e
an
d
n
eg
ativ
e
s
en
tim
en
ts
[
7
]
.
T
h
e
au
to
m
atic
class
if
icatio
n
o
f
th
e
p
o
lar
ity
o
f
s
en
tim
en
t
d
is
p
lay
ed
in
a
tw
ee
t
o
r
co
m
m
en
t
is
a
cr
u
cial
s
tag
e
in
th
e
a
n
aly
s
is
o
f
tex
ts
f
r
o
m
s
o
cial
m
e
d
ia.
SA
i
s
th
e
ter
m
f
o
r
th
is
f
ield
o
f
s
tu
d
y
th
r
o
u
g
h
o
p
in
io
n
m
in
in
g
an
d
SA
h
as
b
ec
o
m
e
ch
allen
g
in
g
an
d
in
ten
s
iv
e
ar
e
as
o
f
s
tu
d
y
f
o
r
b
o
t
h
h
ig
h
l
y
an
d
lo
w
-
r
eso
u
r
ce
d
lan
g
u
ag
es.
Ho
we
v
er
,
e
m
o
tio
n
s
ar
e
co
n
f
u
s
ed
with
t
h
eir
o
p
i
n
io
n
s
s
in
ce
th
ey
ar
e
ass
o
ciate
d
[
8
]
,
[
9
]
.
SA
i
n
Hin
d
i
is
p
r
o
b
lem
atic
f
o
r
a
v
ar
iety
o
f
r
ea
s
o
n
s
,
in
clu
d
in
g
th
o
s
e
lis
ted
:
i)
s
u
p
er
v
is
ed
m
ac
h
in
e
lear
n
i
n
g
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
815
-
1
8
2
9
1816
tech
n
iq
u
es
ca
n
n
o
t
b
e
a
p
p
lie
d
s
in
ce
th
er
e
ar
e
n
'
t
en
o
u
g
h
well
-
an
n
o
tated
s
tan
d
ar
d
co
r
p
o
r
a
,
ii)
Hin
d
i
is
a
lan
g
u
ag
e
with
lim
ited
r
eso
u
r
ce
s
;
n
o
ef
f
ec
tiv
e
p
a
r
s
er
s
o
r
t
ag
g
er
s
ex
is
t
f
o
r
it
,
iii)
t
h
e
Hi
n
d
i
Sen
ti
W
o
r
d
Net
(
HSW
N)
is
o
n
e
o
f
th
e
f
ew
s
o
u
r
ce
s
av
ailab
le
f
o
r
t
h
is
lan
g
u
ag
e.
T
h
er
e
ar
e
n
'
t
m
an
y
ad
jectiv
e
s
an
d
ad
v
er
b
s
in
it.
E
v
en
m
o
s
t
o
f
th
e
te
r
m
s
h
av
e
v
ar
y
in
g
i
n
f
lecte
d
f
o
r
m
s
.
Fu
r
th
er
m
o
r
e,
n
o
n
e
o
f
th
e
wo
r
d
'
s
in
f
lecte
d
f
o
r
m
s
ex
is
t.
HSW
N
i
s
b
u
ilt
b
y
co
n
s
u
m
in
g
th
e
Hin
d
i
W
o
r
d
Net
an
d
E
n
g
lis
h
Sen
tiW
o
r
d
Net
(
SW
N)
.
I
t
was
ass
u
m
ed
th
at
all
an
to
n
y
m
s
p
o
s
s
ess
ed
o
p
p
o
s
ite
p
o
lar
ity
o
f
a
wo
r
d
f
r
o
m
all
s
y
n
o
n
y
m
s
th
r
o
u
g
h
o
u
t
th
e
b
u
i
ld
in
g
o
f
th
is
Hin
d
i
liter
ar
y
r
eso
u
r
ce
.
T
h
is
ass
u
m
p
tio
n
d
is
r
eg
ar
d
ed
t
h
e
id
ea
o
f
w
o
r
d
in
ten
s
ity
as
a
p
o
lar
ity
in
d
i
ca
to
r
,
ev
e
n
th
o
u
g
h
wo
r
d
p
o
lar
ity
in
ten
s
ity
is
im
p
o
r
tan
t
wh
e
n
an
aly
s
in
g
v
iew
p
o
in
ts
,
an
d
(
iv
)
b
ec
au
s
e
o
f
la
n
g
u
ag
e
d
if
f
e
r
en
ce
s
,
ev
en
tr
an
s
latio
n
d
ictio
n
ar
ies
m
ay
n
o
t
co
n
tain
all
o
f
th
e
e
x
p
r
ess
io
n
s
.
C
o
n
tex
t
-
d
ep
e
n
d
en
t
wo
r
d
m
ap
p
in
g
is
a
ch
allen
g
in
g
,
er
r
o
r
-
p
r
o
n
e
p
r
o
ce
d
u
r
e
t
h
at
n
ec
ess
itates
m
an
u
al
lab
o
u
r
b
ec
au
s
e
t
h
e
s
am
e
ter
m
s
ca
n
b
e
em
p
lo
y
ed
in
s
ev
er
al
d
is
tin
ct
s
itu
atio
n
s
.
T
h
er
e
is
an
im
p
o
r
tan
t
lik
elih
o
o
d
o
f
lo
s
in
g
co
n
tex
tu
al
m
ater
i
al
an
d
r
u
n
n
in
g
in
to
tr
an
s
latio
n
is
s
u
es wh
en
u
s
in
g
th
e
T
r
an
s
latio
n
ap
p
r
o
ac
h
to
cr
e
ate
s
u
b
jecti
v
e
lex
ico
n
s
[
10
].
Sen
tim
en
t
m
eth
o
d
s
f
o
r
an
aly
s
is
ar
e
d
iv
id
ed
in
to
th
r
ee
c
a
teg
o
r
ies
[
11
]
:
le
x
ico
n
-
b
ased
,
m
ac
h
in
e
-
lear
n
in
g
,
an
d
h
y
b
r
id
ap
p
r
o
ac
h
es.
Po
lar
ity
is
r
etr
iev
ed
u
s
in
g
a
p
r
eset v
o
ca
b
u
lar
y
o
r
d
ictio
n
ar
y
in
lex
ico
n
-
b
ased
tech
n
iq
u
es.
An
em
o
tio
n
d
ictio
n
a
r
y
is
a
co
llectio
n
o
f
wo
r
d
s
th
at
co
r
r
esp
o
n
d
to
p
o
lar
ity
v
alu
es.
T
h
is
wo
r
k
'
s
p
r
im
ar
y
o
b
jectiv
e
is
to
p
r
o
v
id
e
an
ex
ten
s
iv
e
r
ev
iew
o
f
SA
s
tu
d
ies
th
at
co
n
ce
n
tr
ate
o
n
th
e
Hin
d
i
lan
g
u
ag
e,
em
p
h
asizin
g
th
e
a
p
p
r
o
ac
h
es
tak
en
,
th
e
d
if
f
icu
lties
en
c
o
u
n
ter
ed
,
an
d
p
o
s
s
ib
le
d
ir
e
ctio
n
s
f
o
r
f
u
r
th
er
in
v
esti
g
atio
n
to
im
p
r
o
v
e
SA
s
k
ills
f
o
r
Hin
d
i
co
n
ten
t.
T
h
is
wo
r
k
a
n
aly
s
es
s
tu
d
ies
d
o
n
e
to
e
x
am
in
e
an
d
ex
tr
ac
t
s
en
tim
en
t
d
em
o
n
s
tr
atin
g
th
e
p
o
lar
ity
o
f
w
o
r
d
s
,
p
h
r
a
s
es,
o
r
d
o
cu
m
en
ts
wr
itten
in
Hin
d
i.
SA
ex
am
in
es
th
e
in
p
u
t,
p
r
o
ce
s
s
es
it,
an
d
d
eter
m
in
es
its
p
o
lar
ity
.
SA
ca
n
b
e
u
tili
ze
d
f
o
r
lo
o
k
in
g
in
to
d
if
f
er
e
n
t
ty
p
es
o
f
m
o
v
ie
o
r
p
r
o
d
u
ct
r
ev
iews
th
at
ca
n
b
e
o
b
tain
ed
o
n
d
if
f
er
e
n
t
web
s
ites
an
d
in
d
if
f
er
e
n
t
lan
g
u
ag
es,
etc.
SA
m
ay
b
e
u
s
ed
o
n
th
e
s
u
b
s
eq
u
en
t ty
p
es o
f
tex
t e
n
titi
es [
1
2
]:
a.
Do
cu
m
en
t
lev
el:
ev
er
y
s
in
g
l
e
d
o
cu
m
en
t
is
class
if
ied
as
eith
er
p
o
s
itiv
e,
n
eg
ativ
e,
o
r
o
b
jectiv
e
in
th
e
an
aly
s
is
at
th
e
d
o
cu
m
e
n
t le
v
el
.
b.
Sen
ten
ce
lev
el:
th
e
s
en
ten
ce
-
l
ev
el
an
aly
s
is
lo
o
k
s
at
th
e
s
en
t
en
ce
s
in
th
e
r
eso
u
r
ce
s
.
c.
E
ac
h
s
tatem
en
t
is
ex
am
in
e
d
a
n
d
c
ateg
o
r
ized
as
n
eg
ativ
e
,
o
r
p
o
s
itiv
e.
T
h
e
to
tal
tex
t,
th
er
e
f
o
r
e,
c
o
n
s
is
ts
o
f
a
s
er
ies o
f
s
tatem
en
ts
,
ev
er
y
o
n
e
o
f
wh
ic
h
is
lab
elled
with
its
ass
o
ciate
d
p
o
lar
ity
.
d.
Ph
r
ase
lev
el:
th
e
r
esear
ch
g
o
es
m
u
ch
f
u
r
th
er
,
id
en
tify
in
g
wo
r
d
s
in
a
p
h
r
ase
f
o
r
a
ce
r
t
ain
d
o
cu
m
e
n
t,
ex
am
in
in
g
th
e
w
o
r
d
s
,
an
d
ca
teg
o
r
izin
g
th
e
p
h
r
ases
as
n
eg
ativ
e,
o
r
p
o
s
itiv
e.
T
h
e
f
in
e
-
g
r
ain
ed
an
aly
s
is
,
s
o
m
etim
es
r
ef
er
r
ed
to
as
p
h
r
a
s
e
-
lev
el
an
aly
s
is
,
p
r
o
b
es
d
ee
p
l
y
in
to
t
h
e
tex
t
to
f
in
d
an
d
cla
s
s
if
y
s
u
b
jectiv
e
item
s
/en
titi
es.
F
ig
u
r
e
1
s
h
o
ws
th
e
o
u
tli
n
e
o
f
ev
er
y
s
ec
tio
n
in
th
is
r
ev
iew
f
o
r
SA in
Hin
d
i la
n
g
u
ag
e
.
S
EC
T
I
O
N
2
S
tr
u
c
tu
r
e
o
f
th
i
s
s
y
s
te
m
a
ti
c
l
i
te
r
a
tu
r
e
r
e
v
i
e
w
s
p
r
o
c
e
d
u
r
e
•
S
e
a
rc
h
s
t
ra
t
e
g
y
•
Cri
t
e
ri
a
f
o
r
a
rt
i
c
l
e
s
e
l
e
c
t
i
o
n
p
ro
c
e
s
s
•
Q
u
a
l
i
t
y
A
s
s
e
s
s
m
e
n
t
•
D
a
t
a
E
x
t
ra
c
t
i
o
n
S
EC
T
I
O
N
3
Li
te
r
a
tu
r
e
S
u
r
v
e
y
f
o
r
H
i
n
d
i
S
e
n
ti
m
e
n
ta
l
A
n
a
l
y
s
i
s
3
.
1
.
Co
m
p
a
ra
t
i
v
e
v
i
e
w
o
f
Co
rp
o
ra
3
.
2
.
S
e
n
t
i
m
e
n
t
a
l
A
n
a
l
y
s
i
s
o
f
d
a
t
a
s
e
t
s
S
EC
T
I
O
N
4
C
h
a
l
l
e
n
g
e
s
i
n
a
n
a
l
y
z
i
n
g
th
e
H
i
n
d
i
l
a
n
g
u
a
g
e
•
V
a
ri
a
t
i
o
n
s
i
n
s
p
e
l
l
i
n
g
•
A
r
ra
n
g
e
m
e
n
t
o
f
w
o
rd
s
•
L
i
m
i
t
e
d
l
i
n
g
u
i
s
t
i
c
re
s
o
u
rc
e
s
S
EC
T
I
O
N
1
I
n
tr
o
d
u
c
ti
o
n
•
D
o
c
u
m
e
n
t
l
e
v
e
l
•
S
e
n
t
e
n
c
e
l
e
v
e
l
•
P
h
ra
s
e
l
e
v
e
l
•
H
i
n
d
i
S
e
n
t
i
W
o
rd
N
e
t
S
EC
T
I
O
N
5
S
u
m
m
a
r
y
a
n
d
fu
tu
r
e
d
i
s
c
u
s
s
i
o
n
•
L
i
n
g
u
i
s
t
i
c
F
e
a
t
u
re
Ba
s
e
d
W
o
r
d
E
m
b
e
d
d
i
n
g
•
m
u
l
t
i
-
l
i
n
g
u
a
l
m
o
d
e
l
s
l
i
k
e
BE
RT
o
r
X
L
M
-
R
•
c
r
o
s
s
-
l
i
n
g
u
a
l
t
ra
n
s
fe
r
l
e
a
rn
i
n
g
a
p
p
ro
a
c
h
e
s
S
EC
T
I
O
N
6
C
o
n
c
l
u
s
i
o
n
•
S
A
i
n
t
e
rp
re
t
s
a
t
t
i
t
u
d
e
s
i
n
t
e
x
t
,
d
e
t
e
rm
i
n
i
n
g
t
h
e
a
u
t
h
o
r
s
v
i
e
w
•
S
A
i
n
H
i
n
d
i
i
s
d
y
n
a
m
i
c
.
Fig
u
r
e
1
.
Ov
e
r
v
iew
o
f
s
y
s
tem
atic
liter
atu
r
e
r
ev
iew
p
r
o
ce
s
s
f
o
r
Hin
d
i
SA
Du
e
to
th
e
cu
ltu
r
al
an
d
s
o
cio
lo
g
ical
co
m
p
lex
ity
o
f
Hin
d
i,
SA
in
Hin
d
i
h
as
p
ar
ticu
lar
is
s
u
es
with
th
e
in
cr
ea
s
in
g
n
u
m
b
er
o
f
Hin
d
i
c
o
n
ten
t
in
I
n
d
ia.
T
o
ad
d
r
ess
th
is
g
ap
an
d
im
p
r
o
v
e
th
e
ac
c
u
r
a
cy
an
d
ef
f
icien
c
y
o
f
SA
in
Hin
d
i,
th
is
s
tu
d
y
will
e
x
am
in
e
v
ar
io
u
s
te
ch
n
iq
u
es,
r
e
s
o
u
r
ce
s
,
an
d
ch
allen
g
es.
T
h
is
ar
ticle
ev
alu
ates
th
e
wo
r
k
d
o
n
e
to
d
eter
m
in
e
th
e
wo
r
k
an
d
th
e
wo
r
k
th
at
m
ay
d
o
n
e
f
u
r
th
er
to
m
ak
e
SA in
th
e
Hin
d
i la
n
g
u
ag
e
m
o
r
e
ef
f
icien
t a
n
d
ac
c
u
r
ate.
T
h
e
ar
r
an
g
em
en
t o
f
th
is
r
ev
iew
p
a
p
er
is
as f
o
llo
ws:
Sectio
n
1
p
r
o
v
i
d
es a
n
in
tr
o
d
u
ctio
n
to
th
e
to
p
ic
an
d
estab
lis
h
es
t
h
e
co
n
tex
t
f
o
r
th
e
r
ev
iew.
Sectio
n
2
d
ef
in
es
th
e
f
o
r
m
at
o
f
th
is
r
ev
iew
s
tu
d
y
,
o
u
tlin
in
g
th
e
m
et
h
o
d
o
lo
g
y
an
d
s
tr
u
ctu
r
e
ad
o
p
ted
.
Sectio
n
3
p
r
esen
ts
a
s
u
m
m
ar
y
o
f
r
ec
en
t
r
esear
ch
s
u
r
v
ey
s
,
h
ig
h
lig
h
tin
g
th
e
k
ey
c
o
n
tr
ib
u
t
io
n
s
an
d
f
in
d
in
g
s
in
th
e
f
ield
.
Sectio
n
4
e
x
p
lo
r
es
t
h
e
ch
alle
n
g
es
o
f
ev
alu
atin
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
r
ev
iew
b
a
s
ed
o
n
s
en
timen
ta
l
a
n
a
lysi
s
fo
r
Hin
d
i la
n
g
u
a
g
e
(
V
a
ib
h
a
v
P
r
a
ka
s
h
V
a
s
a
n
i
)
1817
Hin
d
i
lan
g
u
ag
es,
d
is
cu
s
s
in
g
th
e
lim
itatio
n
s
an
d
co
m
p
lex
iti
es
in
v
o
lv
ed
.
Sectio
n
5
s
u
m
m
ar
izes
th
e
ar
ticle,
p
r
o
v
id
i
n
g
an
a
n
aly
s
is
o
f
its
f
in
d
in
g
s
an
d
f
u
tu
r
e
s
co
p
e.
Se
cti
o
n
6
c
o
n
clu
d
es
th
e
p
ap
er
,
o
f
f
er
in
g
f
in
al
th
o
u
g
h
ts
an
d
im
p
licatio
n
s
o
f
th
e
s
tu
d
y
.
2.
T
H
E
S
T
RUC
T
UR
E
O
F
T
H
I
S SYS
T
E
M
A
T
I
C
L
I
T
E
RAT
URE R
E
VI
E
W'
S P
RO
CE
D
URE
2
.
1
.
Sea
rc
h
s
t
ra
t
eg
y
I
E
E
E
x
p
l
o
r
e
r
,
S
p
r
i
n
g
e
r
,
S
ci
e
n
c
e
D
i
r
e
ct
,
E
ls
e
v
i
e
r
,
S
c
o
p
u
s
,
a
n
d
G
o
o
g
l
e
S
c
h
o
l
a
r
w
e
r
e
u
s
e
d
t
o
f
i
n
d
r
e
l
e
v
a
n
t
s
t
u
d
i
e
s
.
T
h
e
s
e
d
i
g
i
t
al
l
i
b
r
a
r
i
es
a
r
e
t
h
e
m
o
s
t
w
e
ll
-
k
n
o
w
n
a
n
d
c
o
m
m
o
n
l
y
u
s
e
d
f
o
r
t
h
e
s
e
n
t
i
m
e
n
t
a
l
a
n
a
l
y
s
is
o
f
H
i
n
d
i
(
S
AH
)
.
T
h
e
s
e
a
r
c
h
q
u
e
r
y
s
t
r
i
n
g
i
s
u
s
e
d
t
o
f
i
n
d
r
e
l
e
v
a
n
t
s
t
u
d
ie
s
b
a
s
e
d
o
n
s
t
u
d
y
t
o
p
i
c
s
s
u
c
h
as
"
s
e
n
t
i
m
e
n
t
a
l
a
n
al
y
s
is
,
"
"H
i
n
d
i
t
e
x
t
,
"
"
N
at
u
r
a
l
L
a
n
g
u
a
g
e
P
r
o
ce
s
s
i
n
g
,
"
a
n
d
"
c
l
as
s
i
f
i
c
at
i
o
n
s
"
in
a
c
e
r
t
a
i
n
d
a
t
a
b
as
e
.
T
h
e
r
e
a
r
e
al
s
o
d
i
f
f
e
r
e
n
t
m
e
t
h
o
d
s
f
o
r
r
e
a
c
h
i
n
g
t
h
e
g
o
a
ls
.
"H
i
n
d
i
S
e
n
tiW
o
r
d
N
e
t
,
"
"
S
u
b
je
ct
i
v
e
L
e
x
ic
o
n
,
"
a
n
d
"
M
a
c
h
i
n
e
L
e
a
r
n
i
n
g
"
a
r
e
a
l
l
c
o
n
s
i
d
e
r
e
d
.
F
i
g
u
r
e
2
s
h
o
w
s
t
h
e
s
tr
u
c
t
u
r
e
o
f
t
h
e
l
i
t
e
r
at
u
r
e
r
e
v
i
e
w
i
n
-
d
e
p
t
h
.
Fig
u
r
e
2
.
Stru
ctu
r
e
o
f
liter
atu
r
e
r
ev
iew
2
.
2
.
Resea
rc
h
qu
estio
ns
I
n
th
e
co
n
tex
t
o
f
a
s
y
s
tem
atic
liter
atu
r
e
r
ev
iew,
th
e
r
esear
ch
q
u
esti
o
n
s
ar
e
cr
itical
in
d
eter
m
in
in
g
th
e
s
ea
r
ch
m
eth
o
d
,
d
ata
ex
tr
ac
tio
n
,
an
d
an
aly
s
is
.
T
h
e
f
o
llo
win
g
r
esear
ch
q
u
esti
o
n
s
h
av
e
b
e
en
ch
o
s
en
f
o
r
th
is
in
v
esti
g
atio
n
:
R
Q1
.
W
h
ich
ty
p
e
o
f
m
eth
o
d
is
u
s
ed
in
th
e
e
x
is
tin
g
liter
atu
r
e
p
ap
er
an
d
wh
at
ar
e
th
e
d
is
ad
v
an
tag
es?
R
Q2
.
Ho
w
well
an
d
s
u
cc
ess
f
u
lly
ca
n
th
ese
ap
p
r
o
ac
h
es b
e
u
s
ed
?
R
Q3
.
W
h
at
ar
e
th
e
v
ar
io
u
s
f
u
t
u
r
e
s
u
g
g
esti
v
e
tech
n
iq
u
es f
o
r
s
en
tim
en
tal
an
aly
s
is
b
ae
s
o
n
Hin
d
i te
x
t?
2
.
3
.
Crit
er
ia
f
o
r
a
rt
icle
s
elec
t
io
n pro
ce
s
s
T
h
e
in
clu
s
io
n
an
d
e
x
clu
s
io
n
cr
iter
ia
wer
e
d
ev
elo
p
ed
a
f
ter
ca
r
ef
u
l
ex
am
i
n
atio
n
o
f
th
e
q
u
ality
o
f
th
e
ex
is
tin
g
liter
atu
r
e.
T
h
e
wr
iter
s
an
aly
s
ed
th
e
p
u
b
licatio
n
s
a
n
d
d
eter
m
i
n
ed
th
e
o
n
es
to
in
clu
d
e
an
d
th
o
s
e
to
ex
clu
d
e
with
th
e
ass
is
tan
ce
o
f
an
ex
p
er
ie
n
ce
d
ad
v
is
o
r
.
T
h
e
p
ap
er
s
wer
e
g
r
ad
ed
u
s
in
g
th
e
f
o
llo
win
g
cr
iter
ia:
−
I
n
clu
s
io
n
c
r
iter
ia:
th
e
p
ap
e
r
s
n
ee
d
to
b
e
b
ased
o
n
SAH
tex
t
,
p
u
b
lis
h
ed
b
etwe
en
2
0
1
0
a
n
d
2
0
2
2
,
an
d
f
u
lly
s
ea
r
ch
ab
le.
−
E
x
clu
s
io
n
r
eq
u
ir
em
en
ts
: p
ap
e
r
s
th
at
f
ail
to
s
atis
f
y
th
e
af
o
r
em
en
tio
n
ed
in
cl
u
s
io
n
r
eq
u
ir
em
en
ts
.
2
.
4
.
Q
ua
lit
y
a
s
s
ess
m
ent
T
h
e
p
ap
er
s
ex
am
in
e
d
f
o
r
t
h
is
s
tu
d
y
wer
e
f
r
o
m
r
e
p
u
tab
le
co
n
f
er
en
ce
s
an
d
jo
u
r
n
als,
g
u
a
r
a
n
teein
g
th
at
th
e
au
th
o
r
s
wer
e
tr
ain
ed
p
r
o
f
e
s
s
io
n
als
in
th
eir
f
ield
s
.
Alth
o
u
g
h
th
er
e
wer
e
n
o
s
et
r
eq
u
ir
em
en
ts
f
o
r
q
u
ality
,
th
e
au
th
o
r
ity
o
f
th
ese
s
o
u
r
ce
s
u
s
u
ally
s
u
g
g
ests
a
s
tan
d
ar
d
o
f
d
e
p
en
d
ab
ilit
y
an
d
ac
ad
em
ic
in
te
g
r
ity
.
T
o
p
r
eser
v
e
th
e
q
u
ality
an
d
ap
p
licab
ilit
y
o
f
th
e
ex
am
in
ed
SA
r
esear
ch
,
th
e
s
elec
tio
n
m
eth
o
d
g
av
e
p
r
ef
er
en
ce
to
well
-
k
n
o
wn
p
u
b
licatio
n
s
.
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
815
-
1
8
2
9
1818
2
.
5
.
Da
t
a
e
x
t
ra
ct
i
o
n
T
h
e
q
u
alities
h
av
e
b
ee
n
id
e
n
tifie
d
f
r
o
m
ex
is
tin
g
liter
atu
r
e
an
d
p
u
b
lis
h
ed
liter
atu
r
e
r
ev
i
ews.
T
h
e
ch
ar
ac
ter
is
tics
tak
en
in
to
ac
co
u
n
t in
th
e
e
x
tr
ac
tio
n
a
r
e
as
: a
)
p
u
b
licatio
n
y
ea
r
an
d
b
)
m
eth
o
d
s
d
is
cu
s
s
ed
.
3.
L
I
T
E
R
AT
U
RE
SU
RVE
Y
T
h
e
p
r
im
a
r
y
c
o
n
tr
ib
u
to
r
s
o
f
t
h
is
wo
r
k
co
n
ce
n
tr
ated
o
n
ass
ess
in
g
an
d
ev
alu
atin
g
p
r
ev
io
u
s
s
tu
d
ies
o
n
SA
,
p
ar
ticu
lar
ly
as
it
r
elate
s
to
th
e
Hin
d
i
lan
g
u
ag
e.
I
n
c
o
m
p
ar
is
o
n
to
th
e
s
ig
n
if
ican
t
r
ese
ar
ch
co
n
d
u
cted
i
n
E
n
g
lis
h
SA,
th
ey
an
aly
s
ed
th
e
d
if
f
er
en
t
tech
n
iq
u
es
an
d
m
eth
o
d
o
lo
g
ies
u
s
ed
in
Hin
d
i
SA
an
d
o
u
tlin
ed
t
h
e
s
h
o
r
tco
m
in
g
s
.
T
h
e
p
r
im
ar
y
r
esear
ch
er
s
o
f
th
is
s
tu
d
y
d
is
co
v
er
ed
th
at
alth
o
u
g
h
SA
h
as
b
ee
n
ex
ten
s
iv
ely
r
esear
ch
ed
f
o
r
th
e
E
n
g
lis
h
l
an
g
u
ag
e
,
it
r
em
ain
ed
an
em
er
g
in
g
f
ield
f
o
r
Hin
d
i.
T
h
e
y
n
o
ted
n
u
m
e
r
o
u
s
ch
allen
g
es,
in
clu
d
in
g
th
e
i
n
tr
icac
y
o
f
Hin
d
i
s
y
n
tax
,
th
e
d
ea
r
th
o
f
d
atasets
with
an
n
o
tatio
n
s
,
an
d
th
e
r
eq
u
ir
em
e
n
t
f
o
r
m
o
r
e
ad
v
an
ce
d
alg
o
r
ith
m
s
s
p
ec
if
ically
d
esig
n
ed
f
o
r
Hin
d
i
to
p
er
f
o
r
m
SA
.
T
h
e
E
n
g
lis
h
lan
g
u
ag
e
was
th
e
f
o
cu
s
o
f
m
o
s
t
o
f
th
e
p
r
io
r
ef
f
o
r
ts
.
T
h
e
r
e
h
asn
'
t
b
ee
n
m
u
ch
f
o
cu
s
o
n
o
th
er
lan
g
u
ag
es.
T
h
e
am
o
u
n
t
o
f
web
i
n
f
o
r
m
atio
n
i
s
s
ig
n
if
ican
tly
in
cr
ea
s
in
g
as
m
o
r
e
in
d
iv
id
u
als
ac
ce
s
s
th
e
i
n
ter
n
et
g
lo
b
ally
a
n
d
f
ee
l
m
o
r
e
co
m
f
o
r
tab
le
u
s
in
g
th
eir
lan
g
u
ag
e.
T
h
e
f
o
u
r
th
m
o
s
t
ex
ten
s
iv
ely
s
p
o
k
en
lan
g
u
ag
e
wo
r
ld
wid
e
i
s
Hin
d
i.
T
h
e
r
is
e
in
u
s
er
-
g
e
n
er
a
ted
co
n
ten
t
o
n
t
h
e
in
ter
n
et
is
th
e
im
p
etu
s
f
o
r
SA
s
tu
d
y
.
As
a
co
n
s
eq
u
en
ce
,
it
is
im
p
o
r
tan
t
to
p
ay
atten
tio
n
to
SA
wr
itten
in
Hin
d
i.
W
ith
an
em
p
h
asis
o
n
Hin
d
i
lan
g
u
a
g
e
d
ata,
we
i
n
v
esti
g
ate
m
u
ltip
le
class
if
ier
s
an
d
ap
p
r
o
a
ch
es
u
tili
ze
d
f
o
r
SA in
th
is
wo
r
k
.
T
h
e
ap
p
r
o
ac
h
Mittal
et
a
l
.
[
1
3
]
s
u
g
g
ested
f
o
r
SA
o
f
Hin
d
i
r
ev
iew
d
o
cu
m
en
ts
was
g
iv
en
.
T
h
e
r
ec
o
m
m
en
d
ed
ap
p
r
o
ac
h
is
f
ir
s
t
test
ed
o
n
an
an
n
o
tated
d
ataset.
T
h
e
cr
ea
tio
n
o
f
ce
r
tain
f
u
n
d
am
en
tal
r
u
les
f
o
r
m
an
ag
in
g
n
eg
atio
n
an
d
s
p
ee
ch
h
as
a
s
ig
n
if
ican
t
in
f
lu
en
c
e
o
n
th
e
o
p
in
io
n
s
ex
p
r
ess
ed
in
th
is
s
tu
d
y
.
T
h
e
p
o
lar
ity
o
f
wo
r
d
s
is
also
d
eter
m
in
ed
u
s
in
g
HSW
N
.
T
h
e
p
o
lar
ity
v
alu
es o
f
all
th
e
p
h
r
ases
in
th
is
tex
t a
r
e
ad
d
ed
to
estab
lis
h
th
e
p
ap
er
'
s
to
tal
s
e
m
an
tic
o
r
ien
tatio
n
.
R
an
i
an
d
Ku
m
a
r
[
1
4
]
u
tili
s
ed
s
ev
er
al
co
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
(
C
NN)
co
n
f
ig
u
r
atio
n
o
p
tio
n
s
to
d
ev
elo
p
a
SA
m
o
d
el
f
o
r
Hi
n
d
i
m
o
v
ie
r
e
v
iews.
Ov
er
5
0
%
o
f
th
e
m
an
u
ally
an
n
o
tated
d
at
aset,
th
e
s
u
g
g
ested
m
o
d
el
was
tr
ain
ed
.
T
h
r
ee
n
ati
v
e
Hin
d
i
s
p
ea
k
er
s
an
n
o
tated
t
h
e
d
ataset.
T
h
e
r
ec
o
m
m
en
d
ed
C
NN
m
o
d
el
b
ea
t
o
th
er
cu
ttin
g
-
e
d
g
e
m
o
d
els
b
y
9
5
%
wh
en
th
e
tr
ial
o
u
tco
m
e
s
wer
e
co
n
tr
asted
with
th
o
s
e
o
f
o
th
er
m
ac
h
in
e
-
lear
n
in
g
ap
p
r
o
ac
h
es.
Sh
r
est
h
a
et
a
l
.
[
1
5
]
cr
ea
t
e
d
a
d
y
n
am
ic
a
p
p
r
o
ac
h
f
o
r
f
i
g
u
r
i
n
g
o
u
t
th
e
p
o
la
r
it
y
o
f
ev
e
r
y
s
e
n
te
n
ce
a
n
d
lo
o
k
i
n
g
at
t
h
e
v
i
ew
p
o
i
n
t
o
f
th
e
s
p
ec
if
ic
s
en
te
n
c
e.
Uti
liz
i
n
g
tw
o
s
e
p
a
r
at
e
class
i
f
i
e
r
s
,
t
h
e
SAH
s
cr
ip
t
w
as
r
e
co
m
m
e
n
d
e
d
.
T
e
x
t
e
x
t
r
ac
t
io
n
m
ak
es
u
s
e
o
f
th
e
n
a
v
e
B
ay
es
cl
ass
i
f
ie
r
a
n
d
d
ec
is
i
o
n
tr
e
e
cl
ass
i
f
ie
r
.
T
h
e
v
al
id
ati
o
n
o
f
g
o
o
d
,
n
e
u
tr
al
,
an
d
n
eg
ati
v
e
r
es
u
lts
d
e
m
o
n
s
t
r
a
te
s
t
h
e
c
o
m
p
ar
is
o
n
r
esu
lt
o
f
th
e
em
o
ti
o
n
al
a
n
al
y
s
is
.
SA,
ac
co
r
d
in
g
to
J
h
a
et
a
l
.
[
1
6
]
,
is
a
s
u
b
f
ield
o
f
NL
P
th
a
t
in
v
o
lv
es
e
x
tr
ac
tin
g
u
s
er
o
p
in
io
n
s
an
d
ca
teg
o
r
is
in
g
th
em
b
ased
o
n
t
h
e
p
o
lar
ity
o
f
th
e
tex
t
p
r
esen
ted
.
T
h
e
jo
b
is
d
o
n
e
in
Hin
d
i,
wh
ich
is
I
n
d
ia'
s
n
atio
n
al
lan
g
u
a
g
e.
T
h
ey
d
e
v
e
lo
p
ed
a
s
en
tim
en
t
-
awa
r
e
d
ictio
n
ar
y
co
m
b
in
in
g
lab
elled
d
at
a
f
r
o
m
th
e
o
r
ig
i
n
al
d
o
m
ain
a
n
d
u
n
lab
eled
d
ata
f
r
o
m
b
o
th
th
e
s
o
u
r
ce
an
d
d
es
ir
ed
d
o
m
ain
s
to
a
n
n
o
tate
t
h
e
co
r
p
o
r
a
f
o
r
e
v
er
y
co
n
ce
iv
ab
le
ar
e
a
o
f
i
n
ter
est.
T
wo
en
s
em
b
le
ap
p
r
o
ac
h
es
th
at
m
ak
e
u
s
e
o
f
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
er
s
(
m
u
lt
ilin
g
u
al
-
B
E
R
T
)
ar
e
p
r
esen
ted
in
th
is
p
u
b
licatio
n
b
y
Path
ak
et
a
l
.
[
1
7
]
as
m
B
E
R
T
-
E
-
MV
an
d
m
B
E
R
T
-
E
-
AS.
B
y
cr
ea
tin
g
an
a
u
x
iliar
y
p
h
r
ase
f
r
o
m
t
h
is
asp
ec
t,
th
e
au
th
o
r
em
p
lo
y
s
a
v
ar
iety
o
f
s
tr
ateg
ies
to
co
n
v
e
r
t
th
e
asp
ec
t
-
b
ased
SA
(
AB
SA)
is
s
u
e
in
to
a
s
en
ten
ce
-
pa
ir
class
if
icatio
n
ass
ig
n
m
en
t.
T
h
e
f
in
al
f
o
r
ec
ast
u
s
es
th
e
p
r
o
p
o
s
ed
m
o
d
el
af
ter
m
an
y
p
r
e
-
tr
ain
ed
B
E
R
T
m
o
d
els
h
a
v
e
b
ee
n
p
o
lis
h
ed
,
y
ield
in
g
o
r
ig
in
al,
an
d
g
r
o
u
n
d
-
b
r
ea
k
in
g
f
in
d
i
n
g
s
f
o
r
d
atasets
f
r
o
m
v
ar
io
u
s
d
o
m
ain
s
in
Hin
d
i.
Gu
p
ta
et
a
l
.
[
1
8
]
d
e
v
elo
p
e
d
C
NN
-
ba
s
ed
SA
o
f
T
wee
ts
f
o
r
s
ca
r
ce
-
r
eso
u
r
ce
la
n
g
u
a
g
e
—
Hin
d
i.
I
n
itially
,
th
is
s
tu
d
y
o
f
f
er
s
an
i
n
teg
r
ated
C
NN
tech
n
iq
u
e
,
u
s
e
s
lex
ico
n
-
b
ased
an
d
m
ac
h
i
n
e
-
l
ea
r
n
in
g
s
u
p
e
r
v
is
ed
ap
p
r
o
ac
h
es
to
ex
p
l
o
r
ativ
ely
a
n
aly
s
e
th
e
s
en
tim
en
t
f
o
u
n
d
in
Hin
d
i
lan
g
u
a
g
e
twee
ts
.
T
h
e
s
u
g
g
este
d
d
o
m
ai
n
-
s
p
ec
if
ic
s
en
tim
en
t
d
ictio
n
ar
y
,
Hin
d
i
Sen
ti
-
W
o
r
d
Net
,
an
d
NR
C
em
o
tio
n
lex
ico
n
ar
e
s
o
m
e
o
f
th
e
lex
ical
r
eso
u
r
ce
s
th
at
ar
e
u
s
ed
in
th
e
lex
ico
n
-
b
ased
a
p
p
r
o
ac
h
to
id
en
tify
t
h
e
o
r
ien
tatio
n
o
f
o
p
in
io
n
f
ea
tu
r
es
f
o
r
ca
teg
o
r
is
in
g
th
e
tex
t
in
t
o
p
o
s
itiv
e
o
r
n
eg
ativ
e
class
es.
T
h
e
s
u
g
g
ested
m
eth
o
d
u
s
es
th
e
m
o
v
ie
an
d
p
r
o
d
u
ct
r
ev
iew
twee
ts
th
at
it
h
as
r
etr
i
ev
ed
f
r
o
m
T
witter
.
I
n
a
d
d
itio
n
to
f
ea
t
u
r
e
lear
n
i
n
g
,
c
o
m
p
u
tatio
n
is
co
s
tly
an
d
tim
e
-
co
n
s
u
m
in
g
s
in
ce
lar
g
e
a
m
o
u
n
ts
o
f
d
ata
ar
e
in
v
o
lv
e
d
.
I
n
th
eir
ap
p
r
o
ac
h
,
B
ak
l
i
wal
et
a
l
.
[
1
9
]
co
m
b
i
n
ed
elem
en
t
s
o
f
th
e
s
tan
d
ar
d
N
-
g
r
am
an
d
th
e
PO
S
tag
g
ed
n
g
r
am
m
eth
o
d
o
l
o
g
ies.
Hin
d
i
lan
g
u
ag
e
r
ev
iews
ar
e
ad
d
itio
n
ally
ca
teg
o
r
ize
d
as
f
av
o
u
r
ab
le
o
r
b
ad
u
s
in
g
lin
g
u
is
tic
ap
p
r
o
ac
h
es.
Fo
r
th
e
Hin
d
i
lan
g
u
ag
e,
Ar
o
r
a
et
a
l
.
[
2
0
]
cr
ea
te
d
a
s
u
b
jectiv
e
l
ex
ico
n
.
Utilizin
g
a
g
r
ap
h
-
b
ased
ap
p
r
o
ac
h
,
th
e
le
x
ico
n
was
b
u
ilt
o
n
to
p
o
f
W
o
r
d
Net.
T
h
ey
s
co
r
ed
7
4
%
ac
cu
r
ac
y
in
r
e
v
iew
ca
teg
o
r
izatio
n
,
a
n
d
6
9
% a
cc
u
r
ac
y
in
Hin
d
i w
h
e
n
co
m
p
ar
ed
t
o
h
u
m
a
n
an
n
o
tato
r
s
.
T
o
d
eter
m
in
e
h
o
w
p
e
o
p
le
g
e
n
er
ally
f
ee
l
ab
o
u
t
th
e
s
to
r
y
,
Y
ad
av
et
a
l
.
[
2
1
]
m
eth
o
d
f
o
r
SA
o
f
Hin
d
i
h
ea
lth
h
ea
d
lin
es
u
s
es
d
ata
f
r
o
m
a
s
ep
a
r
ate
co
r
p
u
s
.
T
h
e
in
v
esti
g
atio
n
'
s
p
o
lar
ity
is
r
etr
iev
ed
f
r
o
m
th
e
c
o
r
p
u
s
,
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
r
ev
iew
b
a
s
ed
o
n
s
en
timen
ta
l
a
n
a
lysi
s
fo
r
Hin
d
i la
n
g
u
a
g
e
(
V
a
ib
h
a
v
P
r
a
ka
s
h
V
a
s
a
n
i
)
1819
an
d
th
e
f
i
n
al
ag
g
r
e
g
ated
p
o
lar
ity
—
wh
ich
m
ay
b
e
p
o
s
itiv
e,
n
eg
ativ
e,
o
r
n
eu
tr
al
—
is
estab
lis
h
ed
.
A
d
ictio
n
ar
y
is
em
p
lo
y
ed
t
o
d
e
f
in
e
th
e
p
o
s
itiv
e
an
d
n
eg
ativ
e
p
o
lar
ity
.
T
h
e
p
r
o
p
o
s
ed
s
y
s
tem
h
as
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
an
d
r
ec
all
o
f
8
5
.
4
2
%,
8
5
.
1
5
%,
an
d
9
7
.
2
3
%,
r
esp
ec
tiv
ely
.
Gh
o
s
h
an
d
Du
tta
[
2
2
]
wo
r
k
s
u
g
g
ests
a
s
tr
ateg
y
t
h
at
em
p
lo
y
s
a
two
-
s
tep
m
eth
o
d
t
o
d
ev
elo
p
a
n
an
ato
m
y
f
o
r
an
aly
s
in
g
f
ee
lin
g
s
in
Hin
d
i
ev
alu
atio
n
s
.
T
h
e
f
o
llo
win
g
s
tep
s
ar
e
s
u
g
g
es
ted
:
i
)
en
h
an
cin
g
Hin
d
iSen
tiW
o
r
d
Net
an
d
ii
)
s
e
n
tim
en
t
ex
tr
ac
tio
n
.
Mo
r
e
ad
je
ctiv
es,
v
er
b
s
,
an
d
ad
v
er
b
s
a
r
e
ad
d
ed
to
HSW
N
to
en
h
an
ce
it.
I
t a
ls
o
s
u
g
g
ests
u
s
in
g
th
e
s
y
s
n
et
r
e
p
lace
m
en
t te
ch
n
iq
u
e
to
i
n
cr
ea
s
e
th
e
ac
cu
r
ac
y
o
f
SA.
B
ak
liwal
et
a
l
.
[
2
3
]
.
u
s
ed
a
lex
i
co
n
-
b
ased
tech
n
i
q
u
e
to
ac
h
iev
e
Hin
d
i
s
en
tim
en
t
ca
teg
o
r
izatio
n
.
Utilizin
g
ad
jectiv
es
an
d
ad
v
er
b
s
u
s
in
g
th
e
an
aly
s
is
o
f
d
ata,
th
ey
cr
ea
te
a
s
ee
d
lis
t,
wh
ich
th
ey
th
e
n
ex
p
a
n
d
u
tili
zin
g
co
r
r
elatio
n
s
b
etwe
en
s
y
n
o
n
y
m
s
a
n
d
a
n
to
n
y
m
s
.
B
e
ca
u
s
e
HSW
N
is
n
o
t
p
u
b
licly
a
v
ailab
le
f
o
r
u
s
ag
e,
th
ey
cr
ea
te
th
eir
o
wn
v
ia
wo
r
d
lin
k
in
g
.
T
h
ey
c
r
ea
ted
th
eir
l
ex
ico
n
u
s
in
g
g
r
ap
h
-
b
ased
tr
a
v
er
s
al.
T
h
ey
awa
r
d
p
o
lar
ity
r
atin
g
s
to
ter
m
s
in
t
h
e
lex
ico
n
.
T
h
eir
s
co
r
in
g
a
p
p
r
o
ac
h
g
en
er
ates
a
p
o
s
itiv
e,
n
eg
ativ
e,
o
r
n
e
u
tr
al
o
u
tco
m
e
b
y
ad
d
in
g
t
h
e
s
co
r
es
o
f
all
o
p
in
i
o
n
wo
r
d
s
in
a
s
en
te
n
ce
.
T
h
e
s
tu
d
y
b
y
Sh
a
r
m
a
et
a
l
.
[
2
4
]
ass
ess
es
in
d
iv
id
u
al
twee
ts
o
n
th
e
T
witter
a
p
p
licatio
n
i
n
Hin
d
i
f
o
r
f
ee
lin
g
s
,
as
well
as
s
en
tim
en
ts
lin
k
ed
with
ce
r
tain
h
ash
ta
g
s
,
b
y
ac
cu
m
u
latin
g
s
p
ec
if
ic
h
ash
tag
twee
ts
an
d
d
eter
m
i
n
in
g
t
h
eir
av
e
r
ag
e
s
en
tim
en
ts
.
T
h
ey
ar
e
tr
ai
n
in
g
th
e
ir
m
o
d
el
with
th
e
n
aiv
e
B
ay
e
s
class
if
ier
,
wh
ich
allo
ws
r
ea
l
-
tim
e
an
aly
s
is
o
f
1
0
s
ets
o
f
twee
ts
g
ath
er
ed
o
v
er
a
h
alf
-
h
o
u
r
in
ter
v
al.
Ho
wev
er
,
th
e
class
if
ier
b
ec
o
m
es
s
tu
ck
wh
ile
ca
lcu
latin
g
th
e
s
en
tim
en
t
v
al
u
e
o
f
a
ce
r
tain
h
ash
tag
s
in
ce
th
e
class
if
ier
d
o
es
n
o
t
in
co
r
p
o
r
ate
th
e
p
o
lar
ity
o
f
o
p
i
n
io
n
s
,
wh
ich
m
i
g
h
t r
esu
lt in
th
e
s
en
tim
en
t v
alu
e
b
ei
n
g
n
e
u
tr
a
lized
o
r
n
e
u
tr
al.
T
h
e
r
esear
ch
u
n
d
er
tak
e
n
b
y
No
r
m
an
et
a
l
.
[
2
5
]
c
o
m
p
ar
es
f
iv
e
I
n
d
ia
n
r
eg
io
n
al
lan
g
u
a
g
e
s
:
B
en
g
a
li,
Hin
d
i,
Ma
lay
alam
,
T
am
il,
a
n
d
T
elu
g
u
.
T
h
e
s
tu
d
y
also
d
is
cu
s
s
es
d
if
f
icu
lties
th
at
ev
er
y
lan
g
u
a
g
e
f
ac
es
in
co
m
p
ar
is
o
n
to
E
n
g
lis
h
.
T
h
is
ar
ticle
in
d
icate
s
th
at
f
ee
lin
g
s
ex
p
r
ess
ed
in
twee
ts
in
E
n
g
lis
h
ar
e
s
im
p
ler
to
an
aly
s
e
th
an
s
en
tim
en
ts
ex
p
r
ess
ed
in
I
n
d
ia
n
r
eg
i
o
n
al
lan
g
u
ag
es.
T
h
e
f
u
tu
r
e
s
co
p
e
m
i
g
h
t
in
clu
d
e
u
s
in
g
a
n
alg
o
r
ith
m
to
c
o
m
p
a
r
e
SA a
m
o
n
g
s
u
ch
.
Hasan
et
a
l
.
[
2
6
]
cr
ea
te
d
a
s
y
s
tem
th
at
ca
n
an
al
y
s
e
n
o
t
o
n
ly
Hin
d
i
b
u
t
also
Ma
r
ath
i
wo
r
d
s
.
T
o
ca
teg
o
r
is
e
th
e
Hin
d
i
an
d
Ma
r
ath
i
wo
r
d
s
,
t
h
e
s
u
g
g
ested
s
y
s
tem
em
p
lo
y
s
a
f
ew
ap
p
r
o
ac
h
e
s
s
u
ch
as
K
n
ea
r
est
n
eig
h
b
o
u
r
(
KNN)
,
n
av
e
B
ay
e
s
,
an
d
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
(
SVM)
.
T
h
e
r
ep
o
r
t
also
an
a
ly
s
es
th
e
r
esu
lts
o
f
th
e
th
r
ee
an
d
d
eter
m
in
es wh
ic
h
is
th
e
b
est.
T
h
e
r
esu
lt o
f
th
e
Hin
d
i a
n
aly
s
is
is
les
s
th
an
8
0
% a
cc
u
r
ate,
wh
er
ea
s
th
e
r
esu
lt
o
f
th
e
Ma
r
ath
i
an
al
y
s
is
is
9
0
%
ac
cu
r
ate.
T
h
ey
p
r
esen
t
th
e
an
aly
s
is
to
b
e
d
o
n
e
n
o
t
o
n
r
u
le
-
b
ased
lan
g
u
ag
e
i
d
en
tific
atio
n
f
o
r
f
u
t
u
r
e
d
ev
el
o
p
m
en
t.
T
o
d
eter
m
i
n
e
th
e
r
i
g
h
t
s
en
s
e
o
f
a
wo
r
d
in
a
co
n
te
x
t,
Mu
l
atk
ar
[
2
7
]
em
p
lo
y
ed
th
e
W
SD
m
eth
o
d
(
Sen
s
e
d
is
am
b
ig
u
atio
n
)
.
SV
M
s
ep
ar
ates
th
e
two
ca
teg
o
r
ies
an
d
b
u
ild
s
a
wid
e
g
ap
b
etwe
en
th
em
.
T
h
e
tech
n
iq
u
e'
s
p
r
o
ce
s
s
es
in
v
o
lv
e
s
tem
m
in
g
,
s
o
r
tin
g
s
in
g
le
-
c
o
lu
m
n
wo
r
d
lin
es,
o
r
r
em
o
v
in
g
s
to
p
wo
r
d
s
.
A
wo
r
d
is
b
r
o
k
en
d
o
wn
as
th
e
s
u
b
s
tr
in
g
p
lu
s
th
e
o
th
er
letter
s
o
f
th
e
wo
r
d
if
it
a
p
p
ea
r
s
as
a
s
u
b
s
tr
i
n
g
in
th
e
f
o
llo
win
g
wo
r
d
.
T
h
e
ad
d
itio
n
al
letter
s
ar
e
r
eg
ar
d
e
d
as a
f
f
ix
es wh
er
ea
s
th
e
s
u
b
s
tr
in
g
is
r
eg
ar
d
ed
as th
e
r
o
o
t o
r
b
ase
f
o
r
m
.
P
a
n
d
e
y
e
t
a
l
.
[
2
8
]
e
m
p
l
o
y
e
d
H
S
W
N
t
o
d
e
t
e
r
m
i
n
e
t
h
e
w
h
o
l
e
e
m
o
t
i
o
n
c
o
n
n
e
c
t
e
d
w
i
t
h
a
H
i
n
d
i
m
o
v
i
e
r
e
v
i
e
w
.
T
o
c
a
l
c
u
l
a
t
e
t
h
e
u
l
t
i
m
a
t
e
a
g
g
r
e
g
a
t
e
d
p
o
l
a
r
i
t
y
,
w
h
i
c
h
m
a
y
b
e
p
o
s
i
t
i
v
e
,
n
e
g
a
t
i
v
e
,
o
r
n
e
u
t
r
a
l
,
t
h
e
p
o
l
a
r
i
t
y
o
f
t
h
e
w
o
r
d
s
i
n
t
h
e
e
v
a
l
u
a
t
i
o
n
i
s
r
e
t
r
i
e
v
e
d
f
r
o
m
H
S
W
N
.
T
o
d
e
t
e
c
t
t
h
e
p
o
l
a
r
i
t
y
o
f
w
o
r
d
s
i
n
H
S
W
N
t
h
a
t
d
o
n
'
t
a
l
r
e
a
d
y
h
a
v
e
p
o
l
a
r
i
t
y
a
t
t
a
c
h
e
d
t
o
t
h
e
m
,
t
h
e
s
y
n
s
e
t
r
e
p
l
a
c
e
m
e
n
t
m
e
t
h
o
d
i
s
e
m
p
l
o
y
e
d
.
T
o
i
m
p
r
o
v
e
s
y
s
t
e
m
p
e
r
f
o
r
m
a
n
c
e
,
n
e
g
a
t
i
v
i
t
y
,
a
n
d
d
i
s
c
o
u
r
s
e
i
n
t
e
r
a
c
t
i
o
n
s
,
w
h
i
c
h
a
r
e
f
r
e
q
u
e
n
t
i
n
H
i
n
d
i
m
o
v
i
e
r
e
v
i
e
w
s
,
a
r
e
a
l
s
o
h
a
n
d
l
e
d
.
J
o
s
h
i
e
t
a
l
.
[
2
9
]
t
o
o
k
t
h
r
e
e
a
p
p
r
o
a
c
h
e
s
:
t
h
e
y
t
r
a
i
n
e
d
a
c
l
a
s
s
i
f
ie
r
,
t
r
a
n
s
l
a
t
e
d
a
g
i
v
e
n
c
o
n
t
e
n
t
i
n
t
o
E
n
g
l
i
s
h
,
a
n
d
c
r
e
a
t
e
d
a
l
e
x
i
c
al
r
es
o
u
r
c
e
c
a
l
l
e
d
H
SW
N
.
S
e
n
tiW
o
r
d
N
e
t
1
.
1
i
s
e
m
p
l
o
y
e
d
i
n
t
h
is
n
ai
v
e
te
c
h
n
i
q
u
e
.
W
o
r
d
N
et
l
i
n
k
i
n
g
i
s
a
t
e
c
h
n
i
q
u
e
f
o
r
m
a
p
p
i
n
g
s
y
n
s
e
ts
o
f
W
o
r
d
N
e
ts
f
r
o
m
v
a
r
i
o
u
s
l
a
n
g
u
a
g
e
s
.
W
ith
2
5
0
H
i
n
d
i
m
o
v
i
e
r
e
v
i
e
w
s
,
PO
S
t
a
g
g
i
n
g
i
s
c
o
m
p
l
e
t
e
d
.
R
a
p
i
d
Mi
n
e
r
5
.
0
i
s
u
s
e
d
t
o
c
l
a
s
s
i
f
y
d
o
c
u
m
e
n
t
s
.
F
o
r
c
la
s
s
i
f
i
c
at
i
o
n
,
u
s
e
t
h
e
L
i
b
S
V
M
t
y
p
e
-
C
l
e
a
r
n
e
r
.
T
h
e
g
r
e
a
t
e
s
t
a
c
c
u
r
a
c
y
f
o
r
r
e
s
o
u
r
c
e
-
b
a
s
e
d
SA
w
as
7
8
.
1
9
a
n
d
6
0
.
3
1
f
o
r
T
F
-
I
D
F
.
Ak
h
tar
et
a
l
.
[
3
0
]
ev
alu
ate
th
e
d
if
f
icu
lties
o
f
SA
in
Hin
d
i
b
y
o
f
f
e
r
in
g
a
b
en
ch
m
a
r
k
s
etu
p
in
wh
ich
th
ey
d
ev
elo
p
a
h
i
g
h
-
q
u
ality
a
n
n
o
tated
d
ata
s
et,
b
u
ild
ML
al
g
o
r
ith
m
s
f
o
r
SA
to
d
em
o
n
s
tr
ate
s
u
cc
ess
f
u
l
d
ataset
co
n
s
u
m
p
tio
n
,
an
d
th
e
n
co
n
s
eq
u
en
tly
cr
ea
te
th
e
r
eso
u
r
c
e
ac
ce
s
s
ib
le
to
th
e
co
m
m
u
n
ity
f
o
r
ad
d
itio
n
al
in
v
esti
g
atio
n
ad
v
a
n
ce
m
en
t.
T
h
e
Hin
d
i
p
r
o
d
u
ct
r
e
v
iews
in
th
e
d
ataset
wer
e
ac
q
u
ir
ed
f
r
o
m
v
ar
io
u
s
web
s
ites
.
An
asp
ec
t
p
h
r
ase
with
t
h
e
em
o
tio
n
ass
o
ciate
d
with
it
is
n
o
te
d
n
ex
t
t
o
ea
ch
s
en
ten
ce
in
t
h
e
r
ev
iew.
Fo
r
asp
ec
t
p
h
r
ase
ex
tr
ac
tio
n
a
n
d
SA,
we
em
p
lo
y
co
n
d
itio
n
al
r
an
d
o
m
f
ield
(
C
R
F)
an
d
SVM
as
th
e
class
if
icatio
n
m
eth
o
d
s
,
ac
co
r
d
in
g
ly
.
T
h
e
a
v
er
ag
e
F
-
m
ea
s
u
r
e
f
o
r
asp
ec
t
p
h
r
ase
ex
tr
ac
tio
n
is
4
1
.
0
7
%,
with
a
s
en
tim
en
t
class
if
icatio
n
ac
cu
r
ac
y
o
f
5
4
.
0
5
%.
T
h
r
ee
I
n
d
ian
lan
g
u
ag
es
—
Hin
d
i,
B
en
g
ali,
a
n
d
T
am
il
—
ar
e
u
s
ed
in
Se
et
a
l
.
[
3
1
]
an
aly
s
is
o
f
s
en
tim
en
t
in
T
witter
d
ata
u
s
in
g
a
n
aiv
e
B
ay
es
class
if
ier
.
T
h
e
twee
ts
ar
e
to
k
en
ized
f
ir
s
t,
an
d
th
en
th
e
n
o
r
m
aliza
tio
n
m
eth
o
d
is
d
o
n
e
to
th
ese
to
k
en
s
.
Fo
r
f
ea
tu
r
e
ex
tr
ac
tio
n
,
Sen
tiW
o
r
d
Net
o
f
I
n
d
ian
la
n
g
u
a
g
es
is
em
p
lo
y
ed
.
B
ay
esian
in
f
er
en
ce
class
if
ier
s
f
u
n
ctio
n
well
f
o
r
s
m
all
d
atasets
an
d
ar
e
in
ef
f
ec
tiv
e
wh
en
id
en
tify
in
g
ch
ar
ac
ter
is
tics
m
an
u
ally
,
as
in
th
e
ca
s
e
o
f
h
u
g
e
d
atasets
.
Fin
ally
,
th
ey
s
co
r
e
d
5
5
.
6
7
%
ac
cu
r
ac
y
in
Hin
d
i
em
o
tio
n
al
a
n
aly
s
is
.
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
815
-
1
8
2
9
1820
B
ased
o
n
twee
ts
in
a
v
ar
iety
o
f
I
n
d
ian
lan
g
u
ag
es,
ele
ctio
n
r
esu
lts
ca
n
n
o
t
b
e
p
r
ed
icted
.
T
o
g
ath
er
twee
ts
in
Hin
d
i,
Sh
ar
m
a
an
d
Mo
h
[
3
2
]
u
s
ed
th
e
T
witter
Ar
ch
iv
er
ap
p
licatio
n
.
B
o
t
h
s
u
p
er
v
is
ed
an
d
u
n
s
u
p
er
v
is
ed
tech
n
i
q
u
es
wer
e
em
p
lo
y
ed
b
y
th
e
r
esear
ch
er
s
.
T
h
e
class
if
ier
ca
teg
o
r
is
ed
th
e
test
d
ata
as
eith
er
p
o
s
iti
v
e,
n
eg
ativ
e,
o
r
n
e
u
tr
al
em
p
lo
y
in
g
d
ictio
n
a
r
y
b
ased
,
n
aiv
e
B
ay
es,
an
d
SVM
tech
n
iq
u
es.
T
h
e
y
also
lo
o
k
ed
in
to
h
o
w
T
witter
p
e
o
p
l
e
f
elt
ab
o
u
t
in
d
iv
id
u
als
o
f
th
e
m
ajo
r
I
n
d
ian
p
o
liti
ca
l
p
ar
ties
.
T
h
ey
u
s
ed
SVM
to
m
ak
e
th
e
f
in
al
f
o
r
ec
ast
s
in
ce
th
e
m
eth
o
d
is
m
o
r
e
a
cc
u
r
ate.
T
h
ey
also
a
n
ticip
ated
th
at
th
e
B
J
P
wo
u
ld
h
av
e
a
h
ig
h
er
c
h
an
ce
o
f
win
n
in
g
th
e
g
en
er
al
elec
tio
n
i
n
2
0
1
6
.
A
ty
p
ical
s
o
cial
n
etwo
r
k
in
g
p
latf
o
r
m
is
T
witter
,
wh
er
e
u
s
er
s
ca
n
s
h
ar
e
th
eir
o
p
in
io
n
s
o
n
a
r
an
g
e
o
f
to
p
ics,
in
clu
d
in
g
r
ev
iews o
f
e
-
co
m
m
er
ce
p
r
o
d
u
cts.
T
h
er
e
ar
e
r
ev
iews o
f
p
r
o
d
u
cts p
r
o
v
id
e
d
in
Hin
d
i a
s
well
as
o
th
er
lan
g
u
a
g
es.
A
SA
s
tr
ate
g
y
d
ev
elo
p
ed
in
Hin
d
i
is
d
escr
ib
ed
b
y
Yad
a
v
et
a
l
.
[
3
3
]
.
T
h
e
lo
n
g
s
h
o
r
t
-
ter
m
m
em
o
r
y
(
L
STM
)
m
o
d
el
is
u
s
ed
in
co
m
b
in
atio
n
with
th
e
c
o
n
tin
u
o
u
s
b
a
g
o
f
w
o
r
d
s
m
o
d
el
s
in
ce
Hin
d
i
SA
is
a
ch
allen
g
in
g
to
p
ic.
T
h
e
ap
p
r
o
p
r
iate
wo
r
d
s
ize
o
f
th
e
i
n
p
u
t
v
ec
to
r
is
d
ete
r
m
in
ed
to
in
c
r
e
ase
ac
cu
r
ac
y
.
Ov
er
8
7
%
ac
cu
r
ac
y
o
n
a
v
er
ag
e
is
ac
co
m
p
lis
h
ed
.
Ad
d
itio
n
ally
,
th
e
r
ec
o
m
m
en
d
ed
r
esu
lts
a
r
e
co
n
t
r
asted
with
cu
ttin
g
-
ed
g
e
m
et
h
o
d
s
.
I
t
was
f
o
u
n
d
th
at
th
e
r
ec
o
m
m
en
d
e
d
s
tr
ateg
y
wo
r
k
s
b
etter
th
an
th
e
o
th
er
s
tr
ateg
ies
ex
am
in
ed
.
T
o
ad
d
r
ess
th
is
p
r
o
b
lem
,
Sh
ar
m
a
an
d
Ar
y
a
[
3
4
]
p
r
o
d
u
ce
d
a
b
r
a
n
d
-
n
ew
Hin
d
i
f
ak
e
n
ews
(
Hin
Fak
eNe
ws)
d
ataset
th
at
co
n
tain
s
an
esti
m
ated
3
3
,
3
0
0
item
s
.
T
h
is
d
ataset
m
ay
b
e
u
s
ed
to
c
r
ea
te
au
to
n
o
m
o
u
s
f
ak
e
n
ews
id
en
tific
atio
n
alg
o
r
ith
m
s
.
L
in
g
u
is
tic
f
ea
tu
r
e
b
ased
wo
r
d
em
b
ed
d
in
g
(
L
FW
E
)
is
a
s
u
g
g
ested
ap
p
r
o
ac
h
th
at
c
r
ea
tes
wo
r
d
em
b
ed
d
in
g
b
ased
o
n
lin
g
u
is
tic
f
ac
to
r
s
.
L
ex
ical,
s
em
an
tic,
s
y
n
tactic,
p
s
y
ch
o
-
lin
g
u
is
tic,
r
ea
d
ab
ilit
y
,
an
d
q
u
an
tit
y
ar
e
th
e
s
ev
er
al
c
ateg
o
r
ies
f
o
r
th
ese
tr
aits
.
T
h
e
co
n
tr
ib
u
tio
n
is
o
f
a
d
u
al
k
i
n
d
.
I
n
th
e
in
itial
s
tag
e,
th
e
d
ataset
is
p
r
e
-
p
r
o
c
ess
ed
,
an
d
lan
g
u
ag
e
tr
aits
ar
e
r
etr
ie
v
ed
.
T
h
e
2
nd
s
tag
e
in
v
o
lv
es
cr
ea
tin
g
f
ea
tu
r
e
s
ets
as
wo
r
d
e
m
b
ed
d
in
g
s
an
d
cl
ass
if
y
in
g
th
em
v
ia
en
s
em
b
le
v
o
tin
g
.
T
h
e
L
FW
E
alg
o
r
ith
m
s
u
cc
ess
f
u
lly
r
ec
o
g
n
i
ze
s
an
d
ca
teg
o
r
izes f
a
k
e
n
ews
in
Hin
d
i w
ith
a
9
8
.
4
9
% a
cc
u
r
a
cy
,
b
y
test
d
ata.
DL
an
d
ML
ar
e
s
u
g
g
ested
f
o
r
th
e
em
o
tio
n
al
in
ter
p
r
etatio
n
o
f
p
o
liti
ca
l
twee
ts
in
Hin
d
i
b
y
J
ain
et
a
l
.
[
3
5
]
.
T
h
ese
tex
t
p
iece
s
ar
e
i
n
Hin
d
i,
a
n
d
th
e
d
ata
h
as
b
e
en
p
r
e
-
p
r
o
ce
s
s
ed
an
d
tr
ain
e
d
u
s
in
g
s
ev
er
al
ML
class
if
ier
s
s
u
ch
as
KN
N,
SV
M,
L
STM
,
an
d
s
o
o
n
.
Utilizi
n
g
th
e
m
eth
o
d
u
s
ed
in
th
is
s
t
u
d
y
,
p
o
liti
ca
l
id
ea
s
ex
p
r
ess
ed
in
Hin
d
i
v
ia
twee
ts
o
n
T
witter
m
ay
b
e
id
en
tifie
d
a
s
h
av
in
g
eith
er
a
p
o
s
itiv
e
o
r
n
e
g
ativ
e
p
o
lar
ity
.
T
h
is
p
r
o
ject
aim
s
t
o
d
e
v
elo
p
a
co
ar
s
e
-
g
r
ai
n
ed
(
b
in
ar
y
)
an
d
f
i
n
e
-
g
r
ain
e
d
(
m
u
lti
-
class
an
d
m
u
lti
-
lab
el)
class
if
icatio
n
-
b
ased
h
o
s
tile
co
n
ten
t
id
en
tific
atio
n
d
ev
ice
i
n
Hin
d
i.
T
o
class
if
y
s
o
cial
n
etwo
r
k
in
g
p
o
s
ts
in
Hin
d
i
Dev
an
ag
ar
i
s
cr
ip
t
as
h
o
s
tile
o
r
n
o
n
-
h
o
s
tile,
C
h
ak
r
ab
o
r
ty
et
a
l
.
[
3
6
]
p
r
o
p
o
s
e
a
B
E
R
T
b
ased
o
n
a
co
n
tex
tu
al
em
b
ed
d
i
n
g
a
p
p
r
o
a
ch
with
a
c
o
n
ca
ten
atio
n
o
f
e
m
o
ji2
v
ec
em
b
ed
in
g
s
b
y
em
p
l
o
y
in
g
th
e
co
n
s
tr
ain
t
2
0
2
1
Hin
d
i
d
ataset
.
I
n
a
d
d
itio
n
,
we
d
ev
elo
p
an
e
n
s
em
b
le
c
lass
if
ier
b
y
m
o
d
if
y
in
g
s
ev
er
al
lear
n
in
g
s
tr
ateg
ies
an
d
em
b
ed
d
in
g
s
tr
u
ctu
r
es
f
o
r
f
in
e
-
g
r
ain
e
d
task
s
in
wh
ich
h
o
s
tile
m
es
s
ag
es
ar
e
cla
s
s
ed
as
d
ef
am
ato
r
y
,
u
n
tr
u
e,
x
en
o
p
h
o
b
ic,
o
r
o
f
f
en
s
iv
e.
I
n
t
h
is
w
o
r
k
,
Vel
an
k
ar
e
t
a
l
.
[
3
7
]
e
x
a
m
i
n
e
t
h
e
id
e
n
ti
f
ic
ati
o
n
o
f
h
a
t
e
a
n
d
a
b
u
s
iv
e
s
p
ee
c
h
i
n
Hin
d
i
a
n
d
Ma
r
at
h
i
lit
er
at
u
r
e
.
T
h
e
c
h
all
e
n
g
e
is
p
o
s
e
d
as
a
te
x
t
c
lass
i
f
i
ca
ti
o
n
jo
b
a
n
d
is
s
o
lv
ed
b
y
u
t
ilis
i
n
g
c
u
t
ti
n
g
-
ed
g
e
d
e
ep
le
ar
n
i
n
g
tec
h
n
i
q
u
es
.
T
h
e
y
l
o
o
k
i
n
t
o
a
v
a
r
i
et
y
o
f
al
g
o
r
i
th
m
s
f
o
r
d
e
e
p
l
ea
r
n
i
n
g
,
i
n
cl
u
d
i
n
g
C
NN
,
L
STM
,
an
d
B
E
R
T
v
a
r
i
a
n
ts
li
k
e
m
u
l
tili
n
g
u
al
B
E
R
T
,
I
n
d
i
cBER
T
,
a
n
d
m
o
n
o
li
n
g
u
al
R
o
B
E
R
T
a
.
Qu
ic
k
te
x
t
wo
r
d
em
b
e
d
d
i
n
g
s
a
r
e
a
d
d
e
d
t
o
t
h
e
u
n
d
e
r
l
y
i
n
g
d
esi
g
n
s
,
t
h
a
t
d
e
p
e
n
d
o
n
C
NN
a
n
d
L
S
T
M
.
Util
izi
n
g
t
h
e
HASOC
2
0
2
1
Hin
d
i
a
n
d
M
ar
at
h
i
h
a
te
s
p
ee
c
h
d
a
tase
ts
,
t
h
e
y
co
n
t
r
as
t
v
a
r
i
o
u
s
ap
p
r
o
ac
h
es
.
F
u
r
t
h
e
r
m
o
r
e
,
o
n
t
h
e
f
in
e
-
g
r
ai
n
ed
Hin
d
i
d
at
aset
,
t
h
e
b
asic
m
o
d
e
l
s
o
u
t
p
e
r
f
o
r
m
B
E
R
T
-
b
ase
d
m
o
d
els
wi
t
h
s
t
a
n
d
ar
d
h
y
p
e
r
p
a
r
a
m
ete
r
twe
ak
in
g
.
T
o
r
ec
o
g
n
ize
m
o
o
d
s
an
d
id
e
n
tify
h
ate
s
p
ee
ch
in
Hin
d
i
a
n
d
B
en
g
ali
d
ata,
Kh
a
n
a
n
d
S
h
ah
id
[
3
8
]
ex
am
in
e
h
o
w
d
ee
p
n
eu
r
al
n
etwo
r
k
s
m
ay
b
e
em
p
lo
y
ed
ef
f
e
ctiv
ely
in
tr
an
s
f
er
lear
n
in
g
a
n
d
jo
in
t
d
u
al
in
p
u
t
lear
n
in
g
s
ettin
g
s
.
Fo
r
t
h
e
Hin
d
i
HASOC
d
atasets
an
d
B
en
g
ali
h
ate
s
p
ee
ch
,
th
e
y
f
ir
s
t
t
r
ain
W
o
r
d
2
Vec
wo
r
d
em
b
ed
d
in
g
s
.
Nex
t,
th
ey
tr
ain
L
STM
,
an
d
f
in
ally
,
th
ey
co
n
s
tr
u
ct
B
en
g
ali
s
en
tim
e
n
t
class
if
ier
s
u
s
in
g
p
ar
am
eter
s
h
ar
in
g
-
b
ased
tr
a
n
s
f
er
lear
n
in
g
b
y
r
e
u
s
in
g
an
d
f
in
e
-
tu
n
in
g
th
e
tr
ain
ed
Hin
d
i
class
if
ier
s
'
weig
h
ts
.
T
h
ese
two
class
if
ier
s
s
er
v
e
as th
e
b
aselin
e
in
th
eir
r
esear
ch
.
Path
ak
et
a
l
.
[
3
9
]
Su
g
g
est
ed
two
en
s
em
b
le
m
o
d
els
b
ased
o
n
m
u
ltil
in
g
u
al
-
B
E
R
T
,
n
am
ely
,
m
B
E
R
T
-
E
-
MV
an
d
m
B
E
R
T
-
E
-
AS
f
o
r
SA
in
th
e
Hin
d
i
lan
g
u
ag
e.
Gen
e
r
ate
an
au
x
iliar
y
s
en
ten
ce
f
r
o
m
th
is
f
ea
tu
r
e
u
s
in
g
v
a
r
io
u
s
tec
h
n
iq
u
es,
th
en
c
o
n
v
e
r
t
th
e
asp
ec
t
-
s
en
tim
en
t
an
aly
s
is
(
AB
SA)
is
s
u
e,
wh
ich
d
eter
m
in
es
ev
er
y
asp
ec
t
c
o
n
tain
ed
in
th
e
s
tatem
en
t
an
d
th
e
em
o
tio
n
co
n
v
e
y
ed
f
o
r
ea
ch
asp
ec
t
t
o
a
s
en
ten
ce
-
p
air
ca
teg
o
r
is
atio
n
task
.
Af
ter
th
at
,
co
m
b
i
n
e
s
ev
er
al
p
r
e
v
io
u
s
ly
t
r
ai
n
ed
B
E
R
T
m
o
d
els
f
o
r
a
f
in
a
l
f
o
r
ec
ast
b
y
f
in
e
-
tu
n
in
g
th
em
.
H
o
wev
er
,
th
is
m
o
d
el
wo
u
l
d
r
eq
u
ir
e
e
x
ten
s
iv
e
co
m
p
u
tatio
n
al
r
eso
u
r
ce
s
f
o
r
f
in
e
-
t
u
n
in
g
an
d
in
f
er
en
ce
,
wh
ich
co
u
ld
p
r
o
v
id
e
d
if
f
icu
lties
with
lim
ited
h
ar
d
war
e.
Als
o
,
th
e
m
o
d
els
h
ad
lim
ited
r
eso
u
r
ce
s
f
o
r
ap
p
ly
in
g
to
o
th
e
r
lan
g
u
ag
es.
Sh
r
iv
astav
a
an
d
Ku
m
ar
[
4
0
]
s
u
g
g
ested
th
e
g
en
etic
alg
o
r
ith
m
-
g
ated
r
ec
u
r
r
en
t
u
n
it
(
GA
-
G
R
U)
m
o
d
el,
wh
ich
h
elp
s
th
e
n
etwo
r
k
to
ap
p
r
o
p
r
iately
ca
teg
o
r
is
e
Hin
d
i
wo
r
d
s
i
n
to
s
en
tim
en
t
clas
s
es
b
y
ef
f
ec
tiv
el
y
ca
p
tu
r
in
g
th
e
s
y
n
tactic
a
n
d
s
em
an
tic
r
elatio
n
s
h
ip
s
b
etwe
e
n
th
em
.
A
g
e
n
etic
alg
o
r
ith
m
is
also
u
s
ed
in
th
is
r
esear
ch
ef
f
o
r
t
to
au
t
o
m
atica
lly
d
esig
n
a
g
ated
r
ec
u
r
r
en
t
n
et
wo
r
k
ar
ch
itectu
r
e,
e
n
ab
lin
g
it
to
ch
o
o
s
e
th
e
b
est
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
r
ev
iew
b
a
s
ed
o
n
s
en
timen
ta
l
a
n
a
lysi
s
fo
r
Hin
d
i la
n
g
u
a
g
e
(
V
a
ib
h
a
v
P
r
a
ka
s
h
V
a
s
a
n
i
)
1821
p
o
s
s
ib
le
h
y
p
er
-
p
ar
am
eter
s
.
T
h
e
p
er
f
o
r
m
an
ce
o
f
a
g
ated
r
ec
u
r
r
en
t
u
n
i
t
n
etwo
r
k
is
h
ig
h
ly
d
ep
en
d
en
t
u
p
o
n
t
h
e
s
elec
tio
n
o
f
its
h
y
p
er
-
p
ar
am
eter
s
.
T
h
e
s
u
g
g
ested
m
o
d
el
was
m
o
r
e
s
u
cc
ess
f
u
l
th
an
o
th
er
c
o
n
v
e
n
tio
n
al
r
eso
u
r
ce
-
b
ased
an
d
m
ac
h
in
e
-
l
ea
r
n
in
g
tec
h
n
iq
u
es,
ac
h
iev
in
g
g
r
o
u
n
d
-
b
r
ea
k
i
n
g
p
e
r
f
o
r
m
an
c
e
o
u
tco
m
es
o
n
th
e
Hin
d
i
m
o
v
ie
r
ev
i
ew
d
ataset.
T
h
is
m
o
d
el
is
lim
ited
to
clas
s
if
y
in
g
th
e
r
ev
iews
in
to
p
o
s
itiv
e,
n
eg
ativ
e,
an
d
n
eu
tr
al
class
es;
s
till
,
th
e
r
ev
iews
ca
n
b
e
class
if
ied
as
a
d
if
f
er
en
t
k
in
d
o
f
f
in
e
-
g
r
ain
e
d
em
o
tio
n
.
T
ab
le
1
(
in
Ap
p
en
d
ix
)
.
s
h
o
ws th
e
s
u
r
v
ey
o
f
s
ev
er
al
s
en
tim
en
tal
an
a
ly
s
es o
f
th
e
Hin
d
i la
n
g
u
ag
e
.
3
.
1
.
Su
mm
a
ry
B
y
r
em
o
v
in
g
th
e
p
o
la
r
ity
f
r
o
m
th
e
c
o
n
ten
t,
SA
h
as
m
ad
e
it
p
o
s
s
ib
le
to
d
eter
m
in
e
o
n
e'
s
attitu
d
e
wh
en
eith
er
wr
itin
g
o
r
s
p
ea
k
in
g
.
D
o
cu
m
en
ts
,
twee
ts
,
n
ews
ar
ticl
es,
co
m
m
en
ts
,
b
lo
g
s
,
s
o
cial
n
etwo
r
k
s
,
as
well
a
s
o
th
er
o
f
f
lin
e
as
well
as
o
n
lin
e
s
o
u
r
ce
s
f
o
r
d
ata
ca
n
all
b
e
u
s
ed
to
g
ath
er
s
en
tim
en
ts
.
SA
h
as
g
r
o
w
n
i
n
p
o
p
u
lar
ity
,
r
esu
ltin
g
in
b
etter
p
r
o
d
u
cts,
a
g
r
ea
ter
u
n
d
er
s
ta
n
d
in
g
o
f
u
s
er
o
p
in
io
n
s
,
an
d
b
etter
ex
ec
u
tio
n
an
d
m
an
ag
em
en
t
o
f
b
u
s
in
ess
ch
o
ices.
Peo
p
le
r
ely
o
n
ev
alu
atio
n
s
an
d
o
p
in
io
n
s
to
m
ak
e
d
ec
i
s
io
n
s
.
Acc
o
r
d
in
g
to
th
e
wr
iter
s
o
f
th
e
af
o
r
em
en
ti
o
n
ed
r
ev
iew
s
tu
d
ies,
ass
ess
in
g
th
e
s
en
tim
en
t
f
o
r
th
e
Hin
d
i
lan
g
u
ag
e
is
s
till
d
if
f
icu
lt.
Ho
wev
er
,
n
u
m
er
o
u
s
r
esear
ch
p
r
esen
ted
d
if
f
er
e
n
t
way
s
f
o
r
alter
in
g
HSW
N
an
d
o
th
er
v
a
r
ied
s
tr
ateg
ies in
th
is
s
tu
d
y
.
T
h
e
f
o
llo
win
g
s
ec
tio
n
ex
p
lain
s
th
e
c
h
allen
g
es in
an
aly
s
in
g
th
e
Hin
d
i la
n
g
u
a
g
e.
3.
2.
Co
m
pa
ra
t
iv
e
v
iew
o
f
co
rpo
ra
T
h
is
r
ev
iew
in
clu
d
es
d
etails
ab
o
u
t
th
e
m
an
y
c
o
r
p
o
r
a
u
tili
ze
d
in
t
h
e
r
esear
c
h
in
v
esti
g
atio
n
s
in
t
h
is
tab
le,
in
clu
d
in
g
th
eir
n
am
es,
s
o
u
r
ce
s
,
an
d
d
escr
ip
tio
n
s
.
T
h
ese
co
r
p
o
r
a
f
o
r
m
th
e
b
asis
o
f
Hin
d
i
SA
,
an
d
r
esear
ch
er
s
f
r
e
q
u
en
tly
u
s
e
th
em
to
test
an
d
im
p
r
o
v
e
SA
m
o
d
els.
T
ab
le
2
ex
p
lain
s
th
e
co
r
p
o
r
a
u
s
ed
in
th
e
r
esear
ch
s
tu
d
ies
in
th
is
r
ev
iew
p
ap
er
.
T
ab
le
2
.
C
o
m
p
a
r
ativ
e
v
iew
o
f
d
atasets
C
o
r
p
u
s
n
a
me
D
e
scri
p
t
i
o
n
S
o
u
r
c
e
U
ses
H
i
n
d
i
so
c
i
a
l
med
i
a
c
o
r
p
us
Th
i
s
c
o
r
p
u
s
i
s
a
c
o
m
b
i
n
a
t
i
o
n
o
f
H
i
n
d
i
t
e
x
t
f
r
o
m
se
v
e
r
a
l
s
o
c
i
a
l
me
d
i
a
si
t
e
s
.
I
t
c
o
n
t
a
i
n
s a
v
a
r
i
e
t
y
o
f
i
t
e
ms
,
i
n
c
l
u
d
i
n
g
c
o
mm
e
n
t
s
,
F
a
c
e
b
o
o
k
p
o
s
t
s,
a
n
d
t
w
e
e
t
s
.
S
o
u
r
c
e
s
i
n
c
l
u
d
e
Tw
i
t
t
e
r
,
F
a
c
e
b
o
o
k
,
o
n
l
i
n
e
f
o
r
u
ms
Th
i
s
d
a
t
a
se
t
i
s fr
e
q
u
e
n
t
l
y
u
se
d
b
y
r
e
s
e
a
r
c
h
e
r
s t
o
e
x
a
mi
n
e
l
a
n
g
u
a
g
e
u
se
i
n
g
e
n
e
r
a
l
o
n
l
i
n
e
c
o
mm
u
n
i
c
a
t
i
o
n
a
n
d
t
o
a
ssess
p
u
b
l
i
c
s
e
n
t
i
m
e
n
t
o
n
so
c
i
a
l
me
d
i
a
.
H
i
n
d
i
n
e
w
s
c
o
r
p
u
s
A
sel
e
c
t
i
o
n
o
f
a
r
t
i
c
l
e
s fr
o
m
d
i
f
f
e
r
e
n
t
n
e
w
s we
b
s
i
t
e
s a
n
d
n
e
w
s
p
a
p
e
r
s
t
h
a
t
a
r
e
a
v
a
i
l
a
b
l
e
i
n
H
i
n
d
i
.
S
o
u
r
c
e
s
i
n
c
l
u
d
e
n
e
w
s w
e
b
s
i
t
e
s a
n
d
v
a
r
i
o
u
s
p
u
b
l
i
c
a
t
i
o
n
s.
A
s i
t
i
s
c
o
mp
i
l
e
d
b
y
v
a
r
i
o
u
s
n
e
w
s
w
e
b
si
t
e
s,
i
t
i
s a
n
e
sse
n
t
i
a
l
r
e
s
o
u
r
c
e
f
o
r
r
e
s
e
a
r
c
h
i
n
g
n
e
w
s se
n
t
i
me
n
t
,
l
a
n
g
u
a
g
e
p
a
t
t
e
r
n
s,
a
n
d
d
e
v
e
l
o
p
m
e
n
t
s
i
n
t
h
e
H
i
n
d
i
-
sp
e
a
k
i
n
g
m
e
d
i
a
e
n
v
i
r
o
n
m
e
n
t
.
M
o
v
i
e
r
e
v
i
e
w
s
A
n
e
x
c
l
u
si
v
e
c
o
l
l
e
c
t
i
o
n
o
f
H
i
n
d
i
-
l
a
n
g
u
a
g
e
mo
v
i
e
r
e
v
i
e
w
s
a
n
d
r
a
t
i
n
g
s.
S
o
u
r
c
e
s
i
n
c
l
u
d
e
mo
v
i
e
r
e
v
i
e
w
w
e
b
si
t
e
s
a
n
d
u
ser
-
g
e
n
e
r
a
t
e
d
c
o
n
t
e
n
t
R
e
v
i
e
w
s
f
r
o
m w
e
b
s
i
t
e
s
t
h
a
t
p
r
o
v
i
d
e
mo
v
i
e
r
e
v
i
e
w
s
a
r
e
i
n
c
o
r
p
o
r
a
t
e
d
w
i
t
h
u
ser
-
g
e
n
e
r
a
t
e
d
c
o
n
t
e
n
t
.
Th
i
s
d
a
t
a
se
t
i
s
u
se
d
b
y
r
e
s
e
a
r
c
h
e
r
s
t
o
i
n
v
e
s
t
i
g
a
t
e
sen
t
i
m
e
n
t
i
n
t
h
e
c
o
n
t
e
x
t
o
f
e
n
t
e
r
t
a
i
n
m
e
n
t
a
n
d
t
o
d
e
v
e
l
o
p
S
A
a
l
g
o
r
i
t
h
ms
f
o
r
m
o
v
i
e
r
e
v
i
e
w
s.
H
i
n
d
i
p
r
o
d
u
c
t
r
e
v
i
e
w
s
Th
e
H
i
n
d
i
p
r
o
d
u
c
t
r
e
v
i
e
w
s
d
a
t
a
se
t
i
n
c
l
u
d
e
s ra
t
i
n
g
s a
n
d
o
p
i
n
i
o
n
s
w
r
i
t
t
e
n
i
n
H
i
n
d
i
o
n
p
r
o
d
u
c
t
s
a
n
d
serv
i
c
e
s
.
S
o
u
r
c
e
s
i
n
c
l
u
d
e
e
-
c
o
mm
e
r
c
e
w
e
b
si
t
e
s
a
n
d
r
e
v
i
e
w
p
l
a
t
f
o
r
ms
.
Th
i
s
i
s
a
u
s
e
f
u
l
t
o
o
l
f
o
r
c
o
m
p
a
n
i
e
s
a
n
d
r
e
s
e
a
r
c
h
e
r
s
w
h
o
w
i
s
h
t
o
k
n
o
w
a
b
o
u
t
c
o
n
su
m
e
r
v
i
e
w
s,
p
r
o
d
u
c
t
sen
t
i
m
e
n
t
,
a
n
d
p
r
o
d
u
c
t
q
u
a
l
i
t
y
a
n
d
s
e
r
v
i
c
e
s
i
n
t
h
e
H
i
n
d
i
-
s
p
e
a
k
i
n
g
m
a
r
k
e
t
.
H
i
n
d
i
sen
t
i
m
e
n
t
l
e
x
i
c
o
n
Th
i
s
l
e
x
i
c
o
n
i
s
a
c
a
r
e
f
u
l
l
y
sel
e
c
t
e
d
l
i
st
o
f
H
i
n
d
i
w
o
r
d
s a
n
d
e
x
p
r
e
ss
i
o
n
s w
i
t
h
s
e
n
t
i
m
e
n
t
l
a
b
e
l
s.
Th
e
l
e
x
i
c
o
n
i
s
d
e
v
e
l
o
p
e
d
t
h
r
o
u
g
h
man
u
a
l
c
u
r
a
t
i
o
n
b
y
l
i
n
g
u
i
s
t
s a
n
d
l
a
n
g
u
a
g
e
e
x
p
e
r
t
s.
W
i
t
h
t
h
e
h
e
l
p
o
f
l
i
n
g
u
i
st
s
a
n
d
l
a
n
g
u
a
g
e
sp
e
c
i
a
l
i
s
t
s,
i
t
i
s ma
n
u
a
l
l
y
c
o
n
st
r
u
c
t
e
d
.
S
A
t
a
s
k
s
i
m
p
r
o
v
e
g
r
e
a
t
l
y
f
r
o
m t
h
e
l
e
x
i
c
o
n
,
w
h
i
c
h
ma
k
e
s
i
t
p
o
ss
i
b
l
e
t
o
c
r
e
a
t
e
mo
d
e
l
s
a
n
d
a
l
g
o
r
i
t
h
ms t
h
a
t
a
r
e
d
e
p
e
n
d
e
n
t
u
p
o
n
sen
t
i
m
e
n
t.
3.
3
.
Sentim
ent
a
l
a
na
ly
s
is
o
f
da
t
a
s
et
s
T
h
is
tab
le
c
o
m
p
ar
es
th
e
ef
f
e
ctiv
en
ess
o
f
s
ev
er
al
SA
tech
n
iq
u
es
o
n
v
ar
io
u
s
d
atasets
.
Asp
ec
ts
o
f
ass
es
s
m
en
t
m
ea
s
u
r
es
in
clu
d
e
F1
-
s
co
r
e,
r
ec
all,
ac
cu
r
ac
y
,
a
n
d
p
r
ec
is
io
n
.
B
y
em
p
lo
y
i
n
g
u
n
if
o
r
m
ass
ess
m
en
t
cr
iter
ia,
th
is
co
m
p
ar
is
o
n
en
ab
les
r
ea
d
er
s
to
ev
alu
ate
th
e
ef
f
icac
y
o
f
v
ar
io
u
s
ap
p
r
o
ac
h
es
ac
r
o
s
s
d
is
tin
ct
d
atasets
.
T
ab
le
3
ex
p
lain
s
th
e
s
en
tim
en
tal
an
aly
s
is
o
f
d
atasets
s
u
g
g
ested
in
th
is
wo
r
k
.
Fig
u
r
e
3
s
h
o
ws
th
e
co
m
p
ar
is
o
n
o
f
ac
cu
r
ac
y
m
etr
ics
with
an
n
o
tate
d
d
atasets
,
m
o
v
ie
r
ev
iew
d
atasets
an
d
s
p
ee
ch
d
etec
tio
n
d
atasets
(
s
ee
in
Ap
p
en
d
ix
)
.
T
h
e
ac
cu
r
ac
y
an
d
p
e
r
f
o
r
m
an
ce
o
f
s
ev
er
al
m
o
d
els
an
d
d
atasets
h
av
e
b
ee
n
ass
ess
ed
i
n
SA
f
o
r
th
e
Hin
d
i
lan
g
u
ag
e.
T
h
e
HSW
N
m
o
d
el,
wh
ic
h
is
b
u
ilt
o
n
lex
ico
n
s
,
d
em
o
n
s
tr
ated
a
m
o
d
es
t
lev
el
o
f
p
er
f
o
r
m
a
n
ce
with
ac
cu
r
a
cies
o
f
7
6
%.
A
C
NN
m
o
d
el
o
u
tp
er
f
o
r
m
ed
o
th
e
r
m
o
d
els
s
ig
n
if
ican
tly
with
an
ac
cu
r
ac
y
o
f
9
5
%,
wh
ile
m
o
r
e
s
o
p
h
is
ticated
m
o
d
els
s
u
ch
a
s
I
n
teg
r
ated
C
NN
-
R
NN
an
d
L
STM
d
is
p
lay
ed
in
cr
ea
s
ed
ac
cu
r
ac
y
at
8
5
%.
T
h
e
L
FW
E
m
o
d
el
p
er
f
o
r
m
ed
ex
ce
p
tio
n
ally
well,
with
a
n
ea
r
-
p
er
f
ec
t
ac
cu
r
ac
y
o
f
9
8
.
4
9
%.
W
h
ile
class
ic
m
o
d
els
lik
e
n
a
i
v
e
B
ay
es
o
b
tain
ed
7
6
.
7
%
ac
cu
r
ac
y
o
n
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
815
-
1
8
2
9
1822
v
ar
io
u
s
d
atasets
,
tech
n
iq
u
es
s
u
ch
as
GA
-
GR
U
ac
h
iev
ed
8
8
.
0
2
%.
T
h
e
Hin
d
i
s
en
tim
en
t
o
r
ien
tatio
n
s
y
s
tem
,
o
n
th
e
o
th
er
h
a
n
d
,
p
er
f
o
r
m
ed
p
o
o
r
ly
,
s
co
r
in
g
b
etwe
en
6
5
an
d
6
6
%
f
o
r
ac
cu
r
ac
y
an
d
p
r
ec
is
io
n
.
A
d
if
f
er
en
t
C
NN
-
L
STM
m
o
d
el
p
er
f
o
r
m
ed
wo
r
s
e,
with
an
ac
cu
r
ac
y
o
f
7
2
.
4
%.
Ov
er
all,
p
er
f
o
r
m
a
n
ce
v
a
r
ies
g
r
ea
tly
b
ased
o
n
t
h
e
n
atu
r
e
an
d
co
m
p
lex
ity
o
f
th
e
d
atasets
an
d
m
o
d
els
th
at
ar
e
u
tili
s
ed
.
T
h
e
s
tu
d
y
o
f
SA
in
Hi
n
d
i
em
p
h
asis
es
th
e
im
p
o
r
tan
ce
o
f
a
d
d
itio
n
al
r
esear
ch
an
d
ad
v
a
n
ce
m
en
t.
W
h
i
le
Hin
d
i
lack
s
a
d
ev
elo
p
ed
co
r
p
u
s
o
f
r
esear
ch
,
E
n
g
lis
h
SA
h
as
b
ee
n
e
x
am
in
ed
ex
ten
s
iv
ely
.
W
h
ile
p
r
o
m
is
in
g
,
ad
v
an
ce
d
m
o
d
els
lik
e
C
NN
an
d
L
FW
E
s
till
n
ee
d
to
b
e
im
p
r
o
v
ed
.
Fo
r
c
o
n
s
is
ten
t
SA
o
u
tp
u
ts
,
h
ig
h
-
q
u
ality
,
s
tan
d
ar
d
is
ed
Hin
d
i
d
a
tasets
ar
e
r
eq
u
ir
ed
.
I
m
p
r
o
v
in
g
Hin
d
i SA
m
ig
h
t r
es
u
lt in
m
o
r
e
p
r
ec
is
e
ap
p
licatio
n
s
.
T
ab
le
3
.
Sen
tim
en
tal
a
n
aly
s
is
o
f
d
atasets
D
a
t
a
s
e
t
s
M
e
t
h
o
d
s
M
e
t
r
i
c
s
p
e
r
f
o
r
m
e
d
Ti
t
l
e
o
f
p
a
p
e
r
Li
mi
t
a
t
i
o
n
s/
F
u
t
u
r
e
w
o
r
k
H
i
n
d
i
p
r
o
d
u
c
t
r
e
v
i
e
w
s
A
B
S
A
[
4
1
]
P
r
e
c
i
s
i
o
n
5
4
.
0
5
%
E
-
c
o
mm
e
r
c
e
p
r
o
d
u
c
t
r
e
v
i
e
w
s
u
si
n
g
a
s
p
e
c
t
-
b
a
s
e
d
H
i
n
d
i
sen
t
i
m
e
n
t
a
n
a
l
y
si
s
A
d
v
a
n
c
e
d
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
ms f
o
r
e
x
t
r
a
c
t
i
o
n
i
n
N
LP
C
R
F
a
n
d
S
V
M
[
42]
A
c
c
u
r
a
c
y
5
4
.
0
5
%
;
F
-
sc
o
r
e
4
1
.
0
7
%
A
B
S
A
i
n
H
i
n
d
i
:
r
e
s
o
u
r
c
e
c
r
e
a
t
i
o
n
a
n
d
e
v
a
l
u
a
t
i
o
n
Ex
p
l
o
r
e
D
L
me
t
h
o
d
s f
o
r
sen
t
i
m
e
n
t
a
l
a
n
a
l
y
si
s
A
n
n
o
t
a
t
e
d
d
a
t
a
se
t
s
D
i
f
f
e
r
e
n
t
C
N
N
[
1
4
]
A
c
c
u
r
a
c
y
9
5
%
D
e
e
p
l
e
a
r
n
i
n
g
-
b
a
s
e
d
S
A
u
s
i
n
g
c
o
n
v
o
l
u
t
i
o
n
n
e
u
r
a
l
n
e
t
w
o
r
k
Ex
p
e
r
i
m
e
n
t
a
l
w
i
t
h
o
t
h
e
r
D
L
a
l
o
n
g
w
i
t
h
o
t
h
e
r
d
a
t
a
s
e
t
s
n
e
e
d
t
o
b
e
i
m
p
r
o
v
e
d
H
S
W
N
[
43]
A
c
c
u
r
a
c
y
7
6
%
A
n
o
v
e
l
s
e
n
t
i
me
n
t
-
a
w
a
r
e
d
i
c
t
i
o
n
a
r
y
f
o
r
mu
l
t
i
-
d
o
m
a
i
n
sen
t
i
m
e
n
t
c
l
a
ss
i
f
i
c
a
t
i
o
n
M
u
s
t
b
e
u
s
e
d
t
o
s
o
l
v
e
o
t
h
e
r
N
LP
i
ssu
e
s
I
n
t
e
g
r
a
t
e
d
C
N
N
-
R
N
N
a
n
d
LST
M
[
1
8
]
A
c
c
u
r
a
c
y
8
5
%
To
w
a
r
d
I
n
t
e
g
r
a
t
e
d
C
N
N
-
b
a
s
e
d
S
A
o
f
t
w
e
e
t
s
f
o
r
s
c
a
r
c
e
-
r
e
s
o
u
r
c
e
l
a
n
g
u
a
ge
—
H
i
n
d
i
D
a
t
a
s
e
t
s
n
e
e
d
t
o
b
e
e
x
p
a
n
d
e
d
f
o
r
b
e
t
t
e
r
r
e
su
l
t
s
C
N
N
[
4
4
]
A
c
c
u
r
a
c
y
9
5
%
D
e
e
p
l
e
a
r
n
i
n
g
-
b
a
s
e
d
S
A
u
s
i
n
g
c
o
n
v
o
l
u
t
i
o
n
n
e
u
r
a
l
n
e
t
w
o
r
k
Th
e
me
t
h
o
d
n
e
e
d
s
t
o
b
e
i
mp
r
o
v
e
d
LFW
E
[
3
4
]
A
c
c
u
r
a
c
y
9
8
.
4
9
%
LFW
E:
l
i
n
g
u
i
st
i
c
f
e
a
t
u
r
e
-
b
a
s
e
d
w
o
r
d
e
m
b
e
d
d
i
n
g
f
o
r
H
i
n
F
a
k
e
N
e
w
s
d
e
t
e
c
t
i
o
n
W
a
n
t
t
o
e
x
p
a
n
d
t
h
e
s
c
o
p
e
o
f
t
h
i
s
e
f
f
o
r
t
t
o
c
o
v
e
r
a
d
d
i
t
i
o
n
a
l
c
h
a
r
a
c
t
e
r
i
s
t
i
c
s
Bi
-
d
i
r
e
c
t
i
o
n
a
l
e
n
c
o
d
e
r
[
3
7
]
F1
-
sc
o
r
e
9
7
.
2
1
%
En
se
mb
l
e
c
l
a
ssi
f
i
e
r
f
o
r
H
i
n
d
i
h
o
s
t
i
l
e
c
o
n
t
e
n
t
d
e
t
e
c
t
i
o
n
C
o
n
t
a
i
n
s
a
d
d
i
t
i
o
n
a
l
l
o
w
-
r
e
s
o
u
r
c
e
l
a
n
g
u
a
g
e
s
l
i
k
e
B
e
n
g
a
l
i
a
n
d
M
a
r
a
t
h
i
i
n
t
h
i
s
p
r
o
j
e
c
t
H
i
n
d
i
r
e
v
i
e
w
a
n
d
p
o
l
i
t
i
c
a
l
t
w
e
e
t
s
mBE
R
T
-
E
-
M
V
a
n
d
mBE
R
T
-
E
-
AS
[
3
9
]
A
c
c
u
r
a
c
y
7
9
.
7
7
%
A
B
S
A
i
n
H
i
n
d
i
l
a
n
g
u
a
g
e
b
y
e
n
s
e
mb
l
i
n
g
p
r
e
-
t
r
a
i
n
e
d
m
B
ER
T
mo
d
e
l
s
M
u
s
t
b
e
u
s
e
d
t
o
s
o
l
v
e
o
t
h
e
r
N
LP
i
ssu
e
s
N
a
ï
v
e
B
a
y
e
s
a
n
d
d
e
c
i
si
o
n
t
r
e
e
c
l
a
ss
i
f
i
e
r
[
1
5
]
NB
-
p
o
si
t
i
v
e
0
.
1
8
7
5
a
n
d
D
T
-
p
o
s
i
t
i
v
e
0
.
5
6
2
5
N
LP
-
b
a
se
d
S
A
H
s
c
r
i
p
t
a
n
o
p
t
i
m
i
z
a
t
i
o
n
a
p
p
r
o
a
c
h
Th
e
ma
j
o
r
i
t
y
o
f
r
e
s
e
a
r
c
h
c
o
n
c
e
n
t
r
a
t
e
s
o
n
e
x
t
r
a
c
t
i
n
g
i
n
f
o
r
mat
i
o
n
f
r
o
m
t
e
x
t
s
a
v
a
i
l
a
b
l
e
o
n
l
i
n
e
M
L
a
n
d
D
L
[
3
5
]
K
N
N
7
4
%
a
n
d
LSTM
7
7
%
A
n
a
l
y
z
i
n
g
p
o
l
i
t
i
c
a
l
t
w
e
e
t
s
i
n
H
i
n
d
i
l
a
n
g
u
a
g
e
u
s
i
n
g
mac
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
I
mp
l
e
me
n
t
d
e
e
p
l
e
a
r
n
i
n
g
met
h
o
d
s
i
n
t
h
e
f
u
t
u
r
e
M
o
v
i
e
r
e
v
i
e
w
d
a
t
a
se
t
s
a
n
d
sp
e
e
c
h
d
e
t
e
c
t
i
o
n
d
a
t
a
se
t
GA
-
G
R
U
[
40]
A
c
c
u
r
a
c
y
8
8
.
0
2
%
A
S
A
sy
s
t
e
m f
o
r
t
h
e
H
i
n
d
i
l
a
n
g
u
a
g
e
b
y
i
n
t
e
g
r
a
t
i
n
g
g
a
t
e
d
r
e
c
u
r
r
e
n
t
u
n
i
t
w
i
t
h
g
e
n
e
t
i
c
a
l
g
o
r
i
t
h
m
Th
i
s
st
u
d
y
ma
y
b
e
e
x
p
a
n
d
e
d
t
o
c
o
v
e
r
o
t
h
e
r
t
y
p
e
s
o
f
e
v
a
l
u
a
t
i
o
n
s
Na
i
v
e
B
a
y
e
s
c
l
a
ss
i
f
i
e
r
[
4
5
]
A
c
c
u
r
a
c
y
7
6
.
7
%
a
n
d
F
-
sc
o
r
e
7
8
.
1
7
%
Le
a
r
n
i
n
g
-
b
a
s
e
d
a
p
p
r
o
a
c
h
f
o
r
H
i
n
d
i
t
e
x
t
S
A
u
s
i
n
g
n
a
i
v
e
b
a
y
e
s
c
l
a
ss
i
f
i
e
r
U
si
n
g
e
n
s
e
mb
l
e
t
e
c
h
n
i
q
u
e
s
c
a
n
i
n
c
r
e
a
s
e
t
h
e
c
l
a
ss
i
f
i
e
r
's
a
c
c
u
r
a
c
y
H
S
W
N
[
2
9
]
A
c
c
u
r
a
c
y
7
8
.
1
4
%
A
f
a
l
l
-
b
a
c
k
st
r
a
t
e
g
y
f
o
r
S
A
i
n
H
i
n
d
i
:
a
c
a
se
s
t
u
d
y
Th
e
h
-
S
W
N
v
e
r
s
i
o
n
n
e
e
d
s t
o
i
mp
r
o
v
e
f
o
r
b
e
t
t
e
r
c
o
v
e
r
a
g
e
D
L
mo
d
e
l
s (
C
N
N
,
LSTM
)
[
3
7
]
A
c
c
u
r
a
c
y
7
2
.
4
%
H
a
t
e
a
n
d
o
f
f
e
n
si
v
e
s
p
e
e
c
h
d
e
t
e
c
t
i
o
n
i
n
H
i
n
d
i
a
n
d
M
a
r
a
t
h
i
Ex
t
e
n
d
t
h
i
s w
o
r
k
f
o
r
b
e
t
t
e
r
a
c
c
u
r
a
c
y
T
h
e
F
ig
u
r
e
3
(
a
)
illu
s
tr
ates
th
e
ac
cu
r
ac
y
o
f
d
if
f
er
en
t
m
eth
o
d
s
u
s
in
g
an
n
o
tated
d
atasets
.
T
h
e
m
eth
o
d
s
co
m
p
ar
ed
in
clu
d
e
HSW
N,
I
n
teg
r
ated
C
NN
-
R
NN
an
d
L
STM
,
C
NN,
an
d
L
FW
E
.
Am
o
n
g
t
h
ese,
HSW
N
ac
h
iev
es
th
e
lo
west
ac
cu
r
ac
y
at
ap
p
r
o
x
im
ately
7
5
%,
in
d
icatin
g
its
lim
ited
ab
ilit
y
to
c
a
p
tu
r
e
th
e
d
ataset'
s
co
m
p
lex
ities
.
T
h
e
I
n
teg
r
ated
C
NN
-
R
N
N
an
d
L
STM
m
eth
o
d
s
h
o
ws
im
p
r
o
v
ed
p
er
f
o
r
m
an
ce
with
an
ac
cu
r
ac
y
o
f
ar
o
u
n
d
8
5
%,
d
em
o
n
s
tr
atin
g
th
e
ad
v
an
ta
g
e
o
f
c
o
m
b
in
i
n
g
C
NNs f
o
r
s
p
atial
f
ea
tu
r
e
ex
tr
a
ctio
n
with
R
NN
an
d
L
STM
f
o
r
tem
p
o
r
al
s
eq
u
e
n
ce
lear
n
i
n
g
.
T
h
e
C
NN
m
et
h
o
d
p
er
f
o
r
m
s
s
ig
n
if
ican
tly
b
etter
,
ac
h
iev
in
g
an
ac
cu
r
ac
y
clo
s
e
to
9
5
%,
h
ig
h
lig
h
tin
g
its
s
tr
en
g
th
in
ex
tr
ac
tin
g
s
p
atial
f
ea
tu
r
es
ef
f
ec
tiv
ely
.
Ho
wev
er
,
L
FW
E
o
u
tp
er
f
o
r
m
s
all
o
th
er
m
eth
o
d
s
,
with
an
ac
cu
r
ac
y
a
p
p
r
o
ac
h
in
g
1
0
0
%,
s
u
g
g
esti
n
g
th
at
it
i
n
c
o
r
p
o
r
ates
a
d
v
an
ce
d
tech
n
iq
u
es
to
o
p
tim
ize
f
ea
t
u
r
e
lear
n
in
g
an
d
class
if
icatio
n
.
Ov
er
all,
L
FW
E
em
er
g
es
as
th
e
m
o
s
t
ef
f
ec
tiv
e
m
eth
o
d
,
f
o
llo
wed
clo
s
ely
b
y
C
NN,
wh
ile
HS
W
N
lag
s
b
eh
in
d
.
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
r
ev
iew
b
a
s
ed
o
n
s
en
timen
ta
l
a
n
a
lysi
s
fo
r
Hin
d
i la
n
g
u
a
g
e
(
V
a
ib
h
a
v
P
r
a
ka
s
h
V
a
s
a
n
i
)
1823
T
h
e
F
ig
u
r
e
3
(
b
)
co
m
p
ar
es th
e
ac
cu
r
ac
y
o
f
d
if
f
er
e
n
t m
eth
o
d
s
ap
p
lied
to
Mo
v
i
e
R
ev
iew
d
atasets
an
d
a
Sp
ee
ch
d
etec
tio
n
d
ataset.
T
h
e
m
eth
o
d
s
in
clu
d
e
GA
-
GR
U,
Naiv
e
B
ay
es
C
la
s
s
if
ier
,
Hin
d
i
Sen
tim
en
t
Or
ien
tatio
n
Sy
s
tem
,
HSW
N,
an
d
C
NN
-
L
STM
.
Am
o
n
g
th
ese,
GA
-
GR
U
ac
h
iev
es
th
e
h
ig
h
est
ac
cu
r
ac
y
a
t
8
8
.
0
2
%,
d
em
o
n
s
tr
atin
g
its
r
o
b
u
s
tn
ess
b
y
c
o
m
b
i
n
in
g
Gen
et
ic
Alg
o
r
ith
m
s
(
GA)
with
Gat
ed
R
ec
u
r
r
en
t
Un
its
(
GR
U)
to
h
an
d
le
co
m
p
lex
d
ata
ef
f
ec
tiv
ely
.
T
h
e
Naiv
e
B
ay
es
C
las
s
if
ier
r
ec
o
r
d
s
an
ac
cu
r
ac
y
o
f
7
6
.
7
%,
s
h
o
win
g
d
ec
en
t
p
er
f
o
r
m
an
c
e
b
u
t
f
allin
g
b
eh
in
d
d
ee
p
lear
n
in
g
-
b
ased
m
eth
o
d
s
.
T
h
e
Hin
d
i
Sen
tim
en
t
Or
ien
t
atio
n
Sy
s
tem
ac
h
iev
es
t
h
e
lo
west
ac
cu
r
ac
y
at
6
5
.
0
%,
r
ef
lectin
g
its
lim
itatio
n
s
in
d
e
alin
g
with
in
tr
icate
r
elatio
n
s
h
ip
s
in
th
e
d
atasets
.
Me
an
wh
ile,
HSW
N
im
p
r
o
v
es
its
p
er
f
o
r
m
an
ce
co
m
p
a
r
ed
to
th
e
f
ir
s
t
ch
a
r
t,
ac
h
iev
in
g
7
8
.
1
4
%
ac
cu
r
ac
y
,
in
d
icatin
g
m
o
d
e
r
ate
ef
f
ec
ti
v
en
ess
.
L
astl
y
,
C
NN
-
L
STM
ac
h
iev
es
7
2
.
4
%,
co
m
b
in
in
g
s
p
atial
an
d
s
eq
u
e
n
tial
lear
n
in
g
b
u
t
s
till
u
n
d
er
p
er
f
o
r
m
in
g
co
m
p
ar
e
d
to
GA
-
GR
U.
Ov
er
all,
GA
-
GR
U
em
er
g
es
as
th
e
b
est
-
p
er
f
o
r
m
in
g
m
eth
o
d
,
wh
ile
tr
a
d
itio
n
al
ap
p
r
o
ac
h
es
lik
e
th
e
Hin
d
i
Sen
tim
en
t
Or
ien
tatio
n
Sy
s
tem
s
h
o
w
lo
w
er
ef
f
ec
tiv
en
ess
co
m
p
a
r
ed
to
a
d
v
an
ce
d
d
ee
p
lear
n
in
g
tec
h
n
iq
u
es.
(
a)
(
b
)
Fig
u
r
e
3
.
C
o
m
p
a
r
is
o
n
o
f
ac
cu
r
ac
y
(
a)
u
s
in
g
an
n
o
tated
d
atase
ts
an
d
(
b
)
u
s
in
g
m
o
v
ie
r
ev
iew
an
d
s
p
ee
ch
d
etec
tio
n
d
ataset
4.
CH
AL
L
E
NG
E
S IN A
NAL
Y
Z
I
NG
T
H
E
H
I
ND
I
L
A
NG
U
AG
E
E
ac
h
d
ialec
t
h
as
a
u
n
iq
u
e
wr
itin
g
s
ty
le
as
well
as
a
u
n
iq
u
e
s
et
o
f
d
if
f
icu
lties
.
T
h
e
r
e
ar
e
two
ap
p
r
o
ac
h
es
f
o
r
in
co
r
p
o
r
atin
g
s
u
ch
ch
ar
ac
ter
is
tics
in
to
th
e
p
ar
s
in
g
p
r
o
ce
s
s
.
T
h
ese
tech
n
iq
u
es
m
ig
h
t
th
o
u
g
h
t
o
f
as
d
is
tin
ct
m
o
d
u
lar
ity
p
ar
ad
ig
m
s
[
4
6
]
.
I
n
s
tead
o
f
s
p
litt
in
g
th
e
p
r
o
ce
d
u
r
e
in
to
two
p
ar
ts
,
it
is
b
etter
to
em
p
lo
y
th
is
d
ata
as
a
f
ea
tu
r
es
f
o
r
p
r
ac
tical
r
ea
s
o
n
s
(
in
clu
d
in
g
POS
tag
g
er
/ch
u
n
k
er
ac
cu
r
ac
y
)
.
W
o
r
k
in
g
with
Hin
d
i
lan
g
u
ag
e
wr
itin
g
p
r
esen
ts
v
ar
i
o
u
s
d
if
f
icu
lties
[
46
]
.
Fo
llo
win
g
is
a
s
u
m
m
ar
y
o
f
th
em
:
Var
iatio
n
s
in
s
p
ellin
g
s
:
n
u
m
er
o
u
s
Hin
d
i
wo
r
d
s
h
a
v
e
th
e
s
a
m
e
m
ea
n
in
g
s
b
u
t
d
if
f
e
r
en
t
s
p
ellin
g
s
.
T
h
e
ex
ten
s
iv
e
v
ar
iety
o
f
v
o
wels
in
th
e
lan
g
u
ag
e
m
i
g
h
t
b
e
to
b
lam
e.
T
h
is
r
en
d
er
s
it
ch
allen
g
in
g
to
co
v
er
all
lex
ica
l
v
ar
ian
ts
o
f
a
s
in
g
le
p
h
r
ase.
Ad
d
itio
n
ally
,
it
is
ch
allen
g
in
g
t
o
tr
ain
a
m
o
d
el
to
h
an
d
le
e
v
er
y
p
o
ten
tial
s
p
ellin
g
v
ar
ian
t.
Fo
r
i
n
s
tan
ce
,
th
er
e
ar
e
m
u
ltip
le
way
s
to
s
p
ell
cr
ick
et
in
Hin
d
i,
in
clu
d
es “
ि
क
े
ट
”,
“
ि
क
े
त
”,
“
िकट
”.
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
815
-
1
8
2
9
1824
Ar
r
an
g
em
e
n
t
o
f
wo
r
d
s
:
th
e
H
in
d
i
lan
g
u
a
g
e'
s
f
r
ee
-
o
r
d
er
ch
a
r
ac
ter
is
tic
en
ab
les
s
en
ten
ce
s
to
tr
an
s
m
it
th
e
s
am
e
id
ea
u
tili
zin
g
a
v
ar
ie
ty
o
f
wo
r
d
s
.
I
n
co
n
tr
ast,
th
e
r
e'
s
a
s
u
b
ject
in
E
n
g
lis
h
,
th
e
n
an
ac
tio
n
v
er
b
,
th
e
n
an
o
b
ject.
T
h
e
s
tatem
en
t
m
a
y
o
cc
asio
n
ally
b
e
ex
p
r
ess
ed
in
r
ev
er
s
e
o
r
d
er
,
o
r
in
wh
at
is
r
ef
er
r
ed
t
o
as
th
e
p
ass
iv
e
f
o
r
m
.
T
h
e
o
r
d
e
r
o
f
w
o
r
d
s
in
a
p
h
r
ase
d
eter
m
in
es
t
h
e
p
o
lar
ity
o
f
th
e
tex
t.
I
n
E
n
g
lis
h
,
f
o
r
ex
a
m
p
le,
"Vir
at
Ko
h
li
p
lay
ed
a
th
r
illi
n
g
in
n
in
g
.
"
T
h
e
f
o
llo
win
g
f
o
r
m
o
f
s
en
ten
ce
s
ar
e
p
o
s
s
ib
le
in
H
in
d
i
f
o
r
th
e
p
h
r
ase
"I
t
was
a
f
ascin
atin
g
m
atch
an
d
k
ep
t
th
e
au
d
ien
ce
i
n
tr
ig
u
ed
:
िर
ा
ट
क
ो
हली
ने
र
ो
मा
चक
पा
र
ी
ख
े
ल
ी
।
यह
एक
आ
क
ष
क
म
ै
च
था
और
श
सक
ो
को
र
ो
मा
चत
क
र
त
ा
र
हा
।
or
िर
ा
ट
क
ो
हली
ने
ख
े
ल
ी
र
ो
मा
चक
पा
र
ी
।
यह
एक
आ
क
ष
क
म
ै
च
था
और
श
सक
ो
को
ब
ा
ध
े
र
खता
था
सा
िश
ई
।
.
L
im
ited
lin
g
u
is
tic
r
eso
u
r
ce
s
:
wh
en
an
aly
zin
g
s
en
tim
en
t
i
n
Hin
d
i,
in
ad
eq
u
ate
f
ac
ilit
ies,
p
o
o
r
ly
p
o
p
u
lated
lin
g
u
is
tic
d
ev
ices,
with
th
e
ab
s
en
ce
o
f
a
n
an
n
o
tated
co
r
p
u
s
ar
e
all
is
s
u
es.
W
i
th
ab
o
u
t
3
,
0
0
0
item
s
,
Hin
d
i
-
Sen
tiW
o
r
d
Net
h
as
a
f
ar
s
m
aller
v
o
ca
b
u
lar
y
th
a
n
E
n
g
lis
h
lin
g
u
is
tic
r
eso
u
r
ce
s
lik
e
Sen
tiW
o
r
d
Net.
Fin
d
in
g
all
ar
ticu
latio
n
s
th
at
d
escr
ib
e
th
e
s
am
e
co
n
ce
p
t
is
a
ch
allen
g
e
k
n
o
wn
as
co
-
r
ef
er
e
n
ce
r
eso
lu
tio
n
.
T
a
k
e
th
e
f
o
llo
win
g
Hin
d
i e
x
am
p
le,
f
o
r
ex
a
m
p
le:
“
मी
र
ा
ने
सी
ता
के
सा
थ
खाना
खाय
ा
।
िह
सो
न
े
चल
ी
गई
”
“
िह
”
in
th
e
latter
s
en
ten
ce
talk
s
ab
o
u
t
“
मी
र
ा
”
t
h
at
is
an
en
tity
.
T
o
ef
f
ec
tiv
ely
a
n
aly
ze
th
e
em
o
tio
n
f
o
u
n
d
at
th
e
asp
ec
t
lay
er
,
it
is
ess
en
tial
to
u
n
d
er
s
tan
d
th
ese
c
o
-
r
ef
er
e
n
ce
r
elatio
n
s
h
ip
s
.
T
h
e
n
ex
t
s
ec
tio
n
g
iv
es
a
s
u
m
m
ar
y
a
n
d
f
u
tu
r
e
d
is
cu
s
s
i
o
n
o
f
th
is
r
ev
iew
p
ap
e
r
.
5.
SUM
M
ARY
A
ND
F
UT
URE
DIS
CU
SS
I
O
N
SA
is
a
p
o
p
u
lar
d
is
cip
lin
e
th
ese
d
ay
s
s
in
ce
it
a
id
s
in
u
n
d
er
s
tan
d
in
g
p
eo
p
le'
s
att
itu
d
es
o
r
th
o
u
g
h
ts
to
war
d
s
s
p
ec
if
ic
to
p
ics.
I
t
is
u
s
ef
u
l
in
a
v
ar
iety
o
f
in
d
u
s
tr
i
es,
in
clu
d
in
g
s
o
cial
m
ed
ia
m
o
n
ito
r
in
g
,
b
u
s
in
ess
,
m
ar
k
et
r
esear
ch
a
n
d
a
n
aly
s
is
,
cu
s
to
m
er
s
er
v
ice,
an
d
s
o
f
o
r
th
.
B
ec
au
s
e
th
e
m
ajo
r
ity
o
f
th
e
s
tu
d
y
is
d
o
n
e
u
tili
s
in
g
HS
W
N
a
s
it
s
r
eso
u
r
ce
,
it m
ay
in
co
r
p
o
r
ate
ad
d
itio
n
a
l r
eso
u
r
ce
s
s
u
ch
as d
ictio
n
ar
ie
s
.
Acc
o
r
d
in
g
to
SA
r
esear
ch
,
a
r
an
g
e
o
f
task
s
h
as
b
ee
n
co
m
p
leted
u
p
to
t
h
is
p
o
in
t.
E
v
en
wh
ile
s
o
m
e
wo
r
k
h
as
b
ee
n
b
e
g
u
n
in
a
v
ar
iety
o
f
lan
g
u
ag
es,
u
tili
zin
g
v
ar
io
u
s
m
eth
o
d
o
lo
g
ies,
tactic
s
,
an
d
a
p
p
r
o
ac
h
es,
th
e
r
e
is
s
till
a
lo
n
g
way
to
g
o
in
th
e
s
am
e
d
ir
ec
tio
n
.
T
h
is
r
ev
iew
ar
ticle
will
class
if
y
th
e
r
ev
iews
i
n
Hin
d
i
an
d
it
w
ill
also
an
aly
s
e
th
e
p
er
f
o
r
m
an
ce
b
y
wo
r
k
in
g
o
n
t
h
e
alg
o
r
ith
m
u
s
in
g
v
ar
io
u
s
m
etr
ics.
T
h
u
s
,
we
co
n
s
id
er
s
o
m
e
f
u
tu
r
e
d
ir
ec
tio
n
s
f
o
r
s
en
tim
en
tal
an
aly
s
is
in
th
e
Hin
d
i la
n
g
u
a
g
e.
−
Mo
r
e
ac
cu
r
ate
tech
n
iq
u
es
m
a
y
b
e
d
ev
elo
p
ed
.
Du
e
to
its
ex
i
s
tin
g
lim
i
tatio
n
s
,
Hin
d
i'
s
lex
ical
co
v
er
a
g
e
ca
n
b
e
in
cr
ea
s
ed
.
−
Ma
n
y
r
esear
ch
er
s
n
e
ed
to
r
esear
ch
s
en
tim
en
tal
an
aly
s
is
f
o
r
t
h
e
Hin
d
i la
n
g
u
ag
e
u
s
in
g
v
ar
i
o
u
s
alg
o
r
ith
m
s
.
−
I
n
th
e
f
u
t
u
r
e,
th
is
m
u
lti
-
class
SA
in
f
o
r
m
atio
n
m
ig
h
t
b
e
u
s
ed
to
s
u
m
m
ar
is
e
a
Hin
d
i
d
o
c
u
m
en
t.
Her
e’
s
a
class
if
icatio
n
o
f
th
e
m
eth
o
d
an
aly
ze
d
in
th
is
r
ev
iew
p
ap
er
.
T
ab
le
s
4
-
7
ex
p
lain
s
th
e
m
eth
o
d
s
u
s
ed
in
th
is
r
ev
iew
p
ap
er
.
T
ab
le
4
.
Asp
ec
t
-
b
ased
s
en
tim
e
n
t a
n
aly
s
is
f
o
r
Hin
d
i
la
n
g
u
a
g
e
A
u
t
h
o
r
s
M
e
t
h
o
d
s
K
e
y
f
e
a
t
u
r
e
s
Y
a
d
a
v
e
t
a
l
.
[
4
1
]
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
,
u
n
su
p
e
r
v
i
s
e
d
st
r
a
t
e
g
y
,
a
n
d
N
LP
U
sag
e
o
f
u
n
s
u
p
e
r
v
i
se
d
s
t
r
a
t
e
g
y
,
e
x
p
l
o
r
e
mu
l
t
i
p
l
e
a
r
e
a
s f
o
r
d
a
t
a
se
t
c
o
l
l
e
c
t
i
o
n
Jh
a
e
t
a
l
.
[
4
3
]
S
e
n
t
i
me
n
t
-
a
w
a
r
e
d
i
c
t
i
o
n
a
r
y
,
l
a
b
e
l
l
e
d
,
a
n
d
u
n
l
a
b
e
l
e
d
d
a
t
a
C
r
e
a
t
i
n
g
a
se
n
t
i
me
n
t
-
a
w
a
r
e
d
i
c
t
i
o
n
a
r
y
,
a
c
o
mb
i
n
a
t
i
o
n
o
f
l
a
b
e
l
l
e
d
a
n
d
u
n
l
a
b
e
l
e
d
d
a
t
a
P
a
t
h
a
k
e
t
a
l
.
[
4
4
]
En
se
mb
l
e
a
p
p
r
o
a
c
h
e
s
a
n
d
p
r
e
-
t
r
a
i
n
e
d
B
E
R
T
m
o
d
e
l
s
Tr
a
n
sf
o
r
mat
i
o
n
o
f
A
B
S
A
i
n
t
o
se
n
t
e
n
c
e
-
p
a
i
r
c
l
a
ss
i
f
i
c
a
t
i
o
n
,
u
sa
g
e
o
f
p
r
e
-
t
r
a
i
n
e
d
B
ER
T
a
p
p
r
o
a
c
h
G
u
p
t
a
e
t
a
l
.
[1
8
]
Le
x
i
c
o
n
-
b
a
s
e
d
me
t
h
o
d
o
l
o
g
i
e
s
,
C
N
N
-
R
N
N
,
a
n
d
LST
M
U
sag
e
o
f
l
e
x
i
c
o
n
s
su
c
h
a
s
H
i
n
d
i
S
e
n
t
i
-
W
o
r
d
N
e
t
a
n
d
N
R
C
Emo
t
i
o
n
Le
x
i
c
o
n
i
n
t
e
g
r
a
t
e
d
C
N
N
-
R
N
N
a
n
d
L
S
TM
t
e
c
h
n
i
q
u
e
T
ab
le
5
.
Mo
v
ie
r
ev
iew
SA
A
u
t
h
o
r
s
M
e
t
h
o
d
s
K
e
y
f
e
a
t
u
r
e
s
R
a
n
i
a
n
d
K
u
m
a
r
[
1
4
]
C
N
N
c
o
n
f
i
g
u
r
a
t
i
o
n
s
,
n
a
t
i
v
e
sp
e
a
k
e
r
s'
a
n
n
o
t
a
t
i
o
n
s
U
sag
e
o
f
C
N
N
s
t
r
u
c
t
u
r
e
,
n
a
t
i
v
e
s
p
e
a
k
e
r
a
n
n
o
t
a
t
i
o
n
s
T
ab
le
6
.
SA
o
n
s
o
cial
m
ed
ia
d
ata
R
e
se
a
r
c
h
p
a
p
e
r
C
l
a
s
si
f
i
c
a
t
i
o
n
M
e
t
h
o
d
s
K
e
y
f
e
a
t
u
r
e
s
N
o
r
ma
n
e
t
a
l
.
[
2
5
]
S
A
i
n
I
n
d
i
a
n
r
e
g
i
o
n
a
l
l
a
n
g
u
a
g
e
s
C
o
m
p
a
r
a
t
i
v
e
a
n
a
l
y
si
s
S
A
w
a
s c
o
m
p
a
r
e
d
w
i
t
h
v
a
r
i
o
u
s I
n
d
i
a
n
r
e
g
i
o
n
a
l
l
a
n
g
u
a
g
e
s
w
i
t
h
En
g
l
i
s
h
H
a
san
e
t
a
l
.
[
4
7
]
M
u
l
t
i
l
i
n
g
u
a
l
SA
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
(
K
N
N
,
N
a
i
v
e
B
a
y
e
s,
a
n
d
S
V
M
)
A
n
a
l
y
s
i
s
o
f
H
i
n
d
i
a
n
d
M
a
r
a
t
h
i
,
K
N
N
,
n
a
i
v
e
b
a
y
e
s
,
a
n
d
S
V
M
me
t
h
o
d
s
Jai
n
e
t
a
l
.
[
3
5
]
P
o
l
i
t
i
c
a
l
t
w
e
e
t
SA
D
e
e
p
l
e
a
r
n
i
n
g
a
n
d
M
L
c
l
a
ss
i
f
i
e
r
s
(
K
N
N
,
S
V
M
,
a
n
d
LSTM
)
U
sag
e
o
f
d
e
e
p
l
e
a
r
n
i
n
g
a
n
d
M
L
c
l
a
ssi
f
i
e
r
s fo
r
p
o
l
i
t
i
c
a
l
t
w
e
e
t
SA
Evaluation Warning : The document was created with Spire.PDF for Python.