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.
23
,
No
.
2
,
A
u
g
u
s
t
2
0
2
1
,
p
p
.
1
0
4
9
~
1
0
5
8
I
SS
N:
2
5
0
2
-
4
7
5
2
,
DOI
: 1
0
.
1
1
5
9
1
/ijeecs.v
23
.i
2
.
pp
1
0
4
9
-
1
0
5
8
1049
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ee
cs.ia
esco
r
e.
co
m
Dete
rmi
ning
subj
ect
hea
ding
s o
f
do
cuments
using
inf
o
rma
tion
retr
iev
a
l mo
dels
E
v
i Y
uli
a
nti
,
L
a
k
s
m
it
a
Ra
ha
dia
nti
F
a
c
u
lt
y
o
f
C
o
m
p
u
ter
S
c
ien
c
e
,
Un
iv
e
rsitas
In
d
o
n
e
sia
,
I
n
d
o
n
e
sia
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Ap
r
3
0
,
2
0
2
1
R
ev
is
ed
J
u
l 4
,
2
0
2
1
Acc
ep
ted
J
u
l 7
,
2
0
2
1
S
u
b
jec
t
h
e
a
d
i
n
g
is
a
c
o
n
tr
o
ll
e
d
v
o
c
a
b
u
lar
y
th
a
t
d
e
sc
rib
e
s
th
e
to
p
ic
o
f
a
d
o
c
u
m
e
n
t,
wh
ich
is
imp
o
rtan
t
to
fin
d
a
n
d
o
r
g
a
n
ize
l
ib
ra
ry
re
so
u
rc
e
s.
As
sig
n
in
g
a
p
p
r
o
p
riate
su
b
jec
t
h
e
a
d
in
g
s
t
o
a
d
o
c
u
m
e
n
t,
h
o
we
v
e
r,
is
a
ti
m
e
-
c
o
n
su
m
in
g
p
ro
c
e
ss
.
We
t
h
e
re
fo
re
c
o
n
d
u
c
t
a
n
o
v
e
l
st
u
d
y
o
n
th
e
e
ff
e
c
ti
v
e
n
e
ss
o
f
i
n
fo
rm
a
ti
o
n
re
tri
e
v
a
l
m
o
d
e
ls,
i
.
e
.
,
lan
g
u
a
g
e
m
o
d
e
l
(L
M
)
a
n
d
v
e
c
to
r
sp
a
c
e
m
o
d
e
l
(VSM
)
,
to
a
u
t
o
m
a
ti
c
a
ll
y
g
e
n
e
ra
te
a
ra
n
k
e
d
li
st
o
f
re
lev
a
n
t
su
b
jec
t
h
e
a
d
in
g
s,
with
t
h
e
a
im
to
g
iv
e
a
re
c
o
m
m
e
n
d
a
ti
o
n
fo
r
li
b
ra
rian
s
to
d
e
term
i
n
e
th
e
su
b
jec
t
h
e
a
d
i
n
g
s
e
ffe
c
ti
v
e
ly
a
n
d
e
fficie
n
tl
y
.
Ou
r
re
su
lt
s
sh
o
w
th
a
t
th
e
re
a
re
a
h
ig
h
n
u
m
b
e
r
o
f
o
u
r
q
u
e
ri
e
s
(u
p
to
6
1
%
)
th
a
t
h
a
v
e
re
lev
a
n
t
su
b
jec
t
h
e
a
d
in
g
s
in
th
e
te
n
t
o
p
-
ra
n
k
e
d
r
e
c
o
m
m
e
n
d
a
ti
o
n
s;
a
n
d
o
n
a
v
e
ra
g
e
,
th
e
first
re
lev
a
n
t
su
b
jec
t
h
e
a
d
i
n
g
is
f
o
u
n
d
a
t
th
e
e
a
rly
p
o
sit
io
n
(3
rd
r
a
n
k
).
Th
is
in
d
ica
tes
th
a
t
d
o
c
u
m
e
n
t
re
tri
e
v
a
l
m
e
th
o
d
s
c
a
n
h
e
lp
th
e
su
b
jec
t
h
e
a
d
in
g
a
ss
ig
n
m
e
n
t
p
r
o
c
e
ss
.
LM
a
n
d
VSM
a
re
sh
o
wn
to
h
a
v
e
c
o
m
p
a
ra
b
le
p
e
rfo
rm
a
n
c
e
,
e
x
c
e
p
t
wh
e
n
t
h
e
se
a
rc
h
u
n
it
is
ti
tl
e
,
VSM
is
su
p
e
rio
r
to
LM
b
y
8
-
2
2
%
.
O
u
r
fu
r
th
e
r
a
n
a
l
y
sis
e
x
h
i
b
it
s
th
re
e
fa
c
u
lt
y
p
a
irs
t
h
a
t
a
re
p
o
ten
ti
a
l
t
o
h
a
v
e
re
se
a
rc
h
c
o
ll
a
b
o
ra
ti
o
n
a
s
th
e
ir
st
u
d
e
n
ts’
th
e
sis
o
ften
h
a
v
e
o
v
e
rlap
su
b
jec
t
h
e
a
d
i
n
g
s:
i)
e
c
o
n
o
m
y
a
n
d
b
u
sin
e
ss
-
so
c
ial
a
n
d
p
o
li
ti
c
a
l
sc
ien
c
e
s,
ii
)
n
u
rsin
g
-
p
u
b
li
c
h
e
a
lt
h
,
a
n
d
ii
i)
m
e
d
icin
e
-
p
u
b
l
ic h
e
a
lt
h
.
K
ey
w
o
r
d
s
:
Do
cu
m
en
t r
etr
ie
v
al
I
n
f
o
r
m
atio
n
r
etr
ie
v
al
L
an
g
u
ag
e
m
o
d
el
Su
b
ject
h
ea
d
in
g
Vec
to
r
s
p
ac
e
m
o
d
el
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 BY
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
E
v
i Y
u
lian
ti
Facu
lty
o
f
C
o
m
p
u
ter
Scien
ce
Un
iv
er
s
itas
I
n
d
o
n
esia
Kam
p
u
s
UI
Dep
o
k
,
J
awa
B
ar
at,
I
n
d
o
n
esia
E
m
ail: e
v
i.y
@
cs.u
i.a
c.
id
1.
I
NT
RO
D
UCT
I
O
N
Un
iv
er
s
itas
I
n
d
o
n
esia
(
UI
)
L
i
b
r
ar
y
ar
ch
i
v
es
s
o
m
e
d
o
c
u
m
en
t
co
llectio
n
s
,
s
u
ch
as:
th
esis
,
a
n
d
b
o
o
k
.
I
t
is
cr
u
cial
to
o
r
g
an
ize
th
ese
co
l
lectio
n
s
in
o
r
d
er
to
h
elp
u
s
er
s
ea
s
ily
f
in
d
th
e
r
eso
u
r
ce
s
t
h
at
t
h
ey
n
ee
d
.
On
e
way
to
o
r
g
an
ize
lib
r
ar
y
co
llectio
n
is
ca
lled
ca
talo
g
u
in
g
,
co
n
d
u
ct
ed
b
y
cr
ea
tin
g
b
ib
lio
g
r
ap
h
ic
r
ec
o
r
d
s
to
r
ep
r
esen
t
d
o
cu
m
e
n
ts
[
1
]
.
A
p
ar
t
o
f
c
atalo
g
u
in
g
p
r
o
ce
s
s
th
at
is
u
s
u
ally
p
er
f
o
r
m
ed
b
y
lib
r
ar
ian
s
is
s
u
b
ject
an
aly
s
is
(
o
r
s
u
b
ject
ca
talo
g
u
in
g
)
,
wh
ich
r
e
q
u
ir
e
th
em
t
o
f
in
d
th
e
s
u
b
ject
h
ea
d
in
g
s
o
f
a
p
ar
ticu
la
r
lib
r
ar
y
r
eso
u
r
ce
[
1
]
-
[
4
]
.
Su
b
ject
h
ea
d
in
g
is
a
c
o
n
tr
o
lled
to
p
ical
v
o
ca
b
u
lar
y
th
at
is
a
s
s
ig
n
ed
to
b
i
b
lio
g
r
a
p
h
ic
r
ec
o
r
d
s
to
f
in
d
an
d
o
r
g
a
n
ize
lib
r
ar
y
m
ater
ial
s
b
ased
o
n
to
p
ics
[
5
]
-
[
7
]
.
I
t
is
i
m
p
o
r
tan
t
esp
ec
ially
wh
en
s
ea
r
ch
er
s
h
av
e
lim
ited
in
f
o
r
m
atio
n
ab
o
u
t
th
e
titl
e
o
r
au
th
o
r
s
o
f
d
o
c
u
m
en
ts
,
s
o
th
at
th
ey
ca
n
s
till
f
in
d
d
o
cu
m
en
ts
b
y
th
eir
to
p
ics.
A
f
ew
s
tu
d
ies
h
av
e
in
v
esti
g
ated
th
e
b
en
ef
it
o
f
s
u
b
ject
h
ea
d
in
g
s
f
o
r
lib
r
ar
y
u
s
er
s
[
8
]
,
[
9
]
.
L
ee
-
Sm
eltze
r
an
d
Hac
k
lem
an
[
8
]
f
o
u
n
d
th
at
li
b
r
ar
y
u
s
er
s
p
r
ef
er
r
e
d
k
ey
wo
r
d
s
ea
r
ch
in
g
in
s
u
b
ject
f
ield
s
o
v
er
o
t
h
er
ca
talo
g
s
ea
r
ch
in
g
m
et
h
o
d
s
.
later
,
s
ap
o
n
-
wh
ite
an
d
h
an
s
b
r
o
u
g
h
[
9
]
s
h
o
wed
th
at
th
e
d
is
s
er
tatio
n
s
with
s
u
b
ject
h
ea
d
in
g
s
ar
e
f
o
u
n
d
to
b
e
m
o
r
e
lik
ely
t
o
cir
cu
late
th
an
t
h
o
s
e
with
o
u
t
s
u
b
ject
h
ea
d
in
g
s
.
So
m
e
r
esear
ch
er
s
h
av
e
also
s
tu
d
ied
s
u
b
ject
h
ea
d
in
g
f
o
r
v
ar
io
u
s
p
u
r
p
o
s
es,
s
u
ch
as:
an
al
y
s
in
g
s
u
b
jec
ts
lis
ted
in
d
if
f
er
en
t
s
u
b
ject
h
ea
d
in
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci,
Vo
l.
23
,
No
.
2
,
Au
g
u
s
t 2
0
2
1
:
1
0
4
9
-
1
0
5
8
1050
lis
t
[
1
0
]
-
[
1
2
]
,
g
en
er
atin
g
m
a
p
o
f
s
cien
ce
[
1
3
]
,
[
1
4
]
,
d
ev
elo
p
in
g
s
im
p
lifie
d
s
u
b
ject
h
ea
d
i
n
g
lis
t
[
1
5
]
,
[
1
6
]
,
an
d
ass
ig
n
in
g
s
u
b
ject
h
ea
d
in
g
au
to
m
atica
lly
[
1
7
]
-
[
1
9
]
.
Dete
r
m
in
in
g
a
n
ac
cu
r
ate
s
u
b
ject
h
ea
d
in
g
f
o
r
a
d
o
cu
m
e
n
t
is
a
co
m
p
le
x
p
r
o
ce
s
s
.
I
t
is
b
ec
au
s
e
a
lib
r
ar
ian
n
ee
d
s
to
d
eter
m
i
n
e
t
h
e
to
p
ic/s
u
b
ject
o
f
th
e
d
o
cu
m
en
t
b
y
c
o
n
d
u
ctin
g
a
th
o
r
o
u
g
h
s
u
b
ject
an
al
y
s
is
;
th
en
f
in
d
co
n
tr
o
lled
v
o
ca
b
u
la
r
ies
f
r
o
m
t
h
e
s
u
b
ject
h
ea
d
in
g
lis
t
to
ass
ig
n
ap
p
r
o
p
r
iate
s
u
b
ject
h
ea
d
in
g
s
[
1
]
.
T
h
er
ef
o
r
e,
n
o
t
all
lib
r
ar
ian
s
ar
e
ca
p
ab
le
to
d
o
th
is
task
ac
cu
r
ately
[
2
0
]
.
I
n
ad
d
itio
n
,
d
if
f
e
r
e
n
t
p
er
s
o
n
s
m
ay
u
s
e
d
if
f
er
en
t
in
ter
p
r
etatio
n
s
o
f
t
h
e
s
u
b
ject
as
th
ey
u
s
e
d
if
f
er
en
t
v
o
ca
b
u
lar
ies
to
r
ep
r
ese
n
t
it.
T
h
is
is
s
u
e
i
s
co
n
f
ir
m
e
d
in
th
e
r
esu
lts
o
f
p
r
ev
io
u
s
s
tu
d
y
th
at
s
h
o
wn
t
h
at
h
u
m
an
s
o
f
ten
ass
ig
n
in
g
class
es
o
r
k
ey
wo
r
d
s
to
b
ib
lio
g
r
ap
h
ic
r
ec
o
r
d
s
in
a
n
in
co
n
s
is
ten
t
m
an
n
er
,
b
o
t
h
co
m
p
ar
ed
to
t
h
em
s
elv
es
an
d
co
m
p
a
r
ed
to
o
th
e
r
h
u
m
an
s
[
2
1
]
.
B
ased
o
n
th
e
af
o
r
em
en
tio
n
ed
p
r
o
b
lem
,
t
h
is
r
esea
r
ch
aim
s
to
h
elp
lib
r
ar
ian
s
to
co
n
d
u
ct
t
h
e
s
u
b
ject
h
ea
d
in
g
ass
ig
n
m
en
t
p
r
o
ce
s
s
b
y
p
r
o
v
id
i
n
g
a
r
ec
o
m
m
e
n
d
atio
n
lis
t
o
f
s
u
b
ject
h
ea
d
in
g
s
f
o
r
a
g
iv
en
d
o
cu
m
e
n
t
.
T
h
e
lib
r
ar
ia
n
s
th
en
ca
n
lo
o
k
a
t
th
e
r
ec
o
m
m
en
d
atio
n
lis
t
an
d
d
ec
id
e
wh
ich
s
u
b
ject
h
ea
d
in
g
s
ar
e
r
elev
a
n
t
to
b
e
ass
ig
n
ed
to
a
d
o
cu
m
en
t
.
I
t
is
ex
p
ec
ted
th
at
lib
r
ar
ian
s
ca
n
f
in
d
r
elev
an
t
s
u
b
ject
h
ea
d
in
g
s
f
r
o
m
th
e
r
ec
o
m
m
en
d
atio
n
lis
t,
o
r
at
th
e
v
er
y
least
th
ey
ca
n
o
b
tain
s
o
m
e
in
s
ig
h
t
ab
o
u
t
th
e
to
p
ics
o
f
s
im
ilar
d
o
cu
m
en
ts
in
th
e
c
o
llectio
n
.
Fin
ally
,
th
e
g
o
al
o
f
th
is
s
tu
d
y
is
to
en
a
b
l
e
th
e
s
u
b
ject
h
ea
d
in
g
ass
ig
n
m
en
t
p
r
o
ce
s
s
to
r
u
n
m
o
r
e
ef
f
ec
tiv
e
ly
an
d
ef
f
icien
t
ly
b
y
lib
r
ar
ian
s
.
Ou
r
wo
r
k
is
d
if
f
er
en
t
to
tex
t
ca
teg
o
r
izatio
n
[
2
2
]
,
[
2
3
]
s
in
c
e
we
d
o
n
o
t
ca
teg
o
r
ize
d
o
cu
m
en
ts
in
to
ce
r
tain
to
p
ics.
So
m
e
p
r
e
v
io
u
s
wo
r
k
th
at
ar
e
m
o
s
tly
r
elate
d
to
o
u
r
w
o
r
k
ar
e
s
u
m
m
ar
ized
in
T
ab
le
1
.
T
h
ey
in
v
esti
g
ated
th
e
way
s
to
ass
ig
n
s
u
b
ject
h
ea
d
in
g
s
au
t
o
m
ati
ca
lly
[
1
7
]
-
[
1
9
]
.
Ou
r
wo
r
k
is
d
if
f
er
en
t
t
o
th
em
i
n
wh
ich
we
f
o
r
m
u
late
o
u
r
tas
k
as
a
s
u
b
ject
h
ea
d
in
g
r
ec
o
m
m
en
d
atio
n
/r
et
r
iev
al
in
s
tead
o
f
s
u
b
ject
h
ea
d
i
n
g
ass
ig
n
m
en
t
o
r
class
if
icatio
n
.
T
h
er
ef
o
r
e,
we
d
o
n
o
t
ass
ig
n
a
s
u
b
ject
h
ea
d
in
g
to
a
d
o
cu
m
e
n
t,
b
u
t
we
r
an
k
th
e
s
u
b
ject
h
ea
d
in
g
s
o
f
s
im
ilar
d
o
cu
m
en
ts
in
th
e
co
lle
ctio
n
i
n
o
r
d
er
to
g
e
n
er
ate
s
u
b
ject
h
ea
d
in
g
s
r
ec
o
m
m
en
d
atio
n
lis
t.
I
n
a
ty
p
ical
d
o
cu
m
en
t
r
etr
iev
al
s
y
s
tem
,
th
e
s
ea
r
ch
r
esu
lts
ar
e
th
e
r
an
k
ed
lis
t
o
f
d
o
cu
m
en
ts
,
h
o
wev
er
in
th
is
wo
r
k
,
th
e
r
esu
lts
ar
e
th
e
r
an
k
ed
lis
t
o
f
s
u
b
ject
h
ea
d
i
n
g
s
f
r
o
m
s
im
ilar
d
o
cu
m
en
ts
.
T
h
e
ev
alu
atio
n
is
th
er
ef
o
r
e
p
er
f
o
r
m
ed
o
n
th
e
u
n
it
o
f
s
u
b
ject
h
ea
d
in
g
s
in
s
tead
o
f
d
o
cu
m
en
t
s
,
wh
ich
f
u
r
th
e
r
d
if
f
er
e
n
tiates
th
is
wo
r
k
with
a
s
tan
d
ar
d
d
o
cu
m
e
n
t r
etr
iev
al
task
.
T
ab
le
1
.
C
o
m
p
a
r
is
o
n
with
s
o
m
e
m
o
s
tly
r
elate
d
wo
r
k
Ta
sk
M
e
t
h
o
d
Ty
p
e
K
a
v
u
l
u
r
u
a
n
d
H
e
[
1
7
]
S
u
b
j
e
c
t
h
e
a
d
i
n
g
a
ss
i
g
n
me
n
t
U
si
n
g
n
a
me
d
e
n
t
i
t
y
r
e
c
o
g
n
i
t
i
o
n
,
r
e
l
a
t
i
o
n
s
h
i
p
e
x
t
r
a
c
t
i
o
n
,
a
n
d
o
u
t
p
u
t
l
a
b
e
l
c
o
-
o
c
c
u
r
r
e
n
c
e
f
r
e
q
u
e
n
c
i
e
s
o
f
med
i
c
a
l
s
u
b
j
e
c
t
h
e
a
d
i
n
g
t
e
r
m
p
a
i
r
s.
U
n
su
p
e
r
v
i
s
e
d
K
a
v
u
l
u
r
u
a
n
d
Lu
[
1
8
]
S
u
b
j
e
c
t
h
e
a
d
i
n
g
a
ss
i
g
n
me
n
t
& c
l
a
ssi
f
i
c
a
t
i
o
n
U
si
n
g
t
e
r
m
c
o
-
o
c
c
u
r
r
e
n
c
e
f
r
e
q
u
e
n
c
i
e
s
a
n
d
l
a
t
e
n
t
t
e
r
m a
ss
o
c
i
a
t
i
o
n
U
n
su
p
e
r
v
i
s
e
d
S
u
p
e
r
v
i
se
d
G
o
l
u
b
e
t
a
l
.
[
1
9
]
S
u
b
j
e
c
t
h
e
a
d
i
n
g
c
l
a
ss
i
f
i
c
a
t
i
o
n
U
si
n
g
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
m
(
e
.
g
.
,
S
V
M
)
S
u
p
e
r
v
i
se
d
O
u
r
w
o
r
k
S
u
b
j
e
c
t
h
e
a
d
i
n
g
r
e
c
o
m
me
n
d
a
t
i
o
n
/
r
e
t
r
i
e
v
a
l
U
si
n
g
i
n
f
o
r
mat
i
o
n
r
e
t
r
i
e
v
a
l
me
t
h
o
d
s,
i
.
e
.
,
L
M
a
n
d
V
S
M
U
n
su
p
e
r
v
i
s
e
d
Ou
r
id
ea
is
s
u
p
p
o
r
ted
b
y
Go
lu
b
et
a
l.
[
1
9
]
wh
o
co
n
clu
d
e
d
th
at
au
to
m
atic
s
u
b
ject
h
ea
d
in
g
ass
ig
n
m
en
t
s
h
o
u
ld
n
ev
er
b
e
im
p
lem
e
n
te
d
o
n
its
o
wn
;
in
s
tead
,
a
s
y
s
t
em
s
h
o
u
ld
co
m
b
in
e
th
e
ef
f
ic
ien
cy
o
f
au
to
m
atic
s
u
g
g
esti
o
n
s
with
q
u
ality
o
f
h
u
m
an
d
ec
is
io
n
s
at
th
e
f
in
al
s
tag
e.
T
h
ey
f
o
u
n
d
th
at
a
p
p
ly
in
g
p
u
r
ely
au
to
m
atic
s
u
b
ject
h
ea
d
in
g
class
if
icatio
n
d
o
es
n
o
t
w
o
r
k
,
b
ec
au
s
e
th
er
e
ar
e
a
lar
g
e
n
u
m
b
er
o
f
s
u
b
je
ct
h
ea
d
in
g
s
class
es.
Fo
r
th
is
r
ea
s
o
n
,
th
ey
lim
it
to
u
s
e
o
n
ly
to
p
th
r
ee
lev
els
in
d
ewe
y
d
ec
im
al
class
if
icatio
n
(
DD
C
)
s
u
b
ject
h
ea
d
in
g
lis
t
(
by
u
s
in
g
8
0
3
in
s
tead
o
f
1
4
,
4
1
3
class
es),
wh
ich
th
er
ef
o
r
e
d
o
es
n
o
t
r
ef
lect
th
e
o
p
er
atio
n
al
s
y
s
tem
.
B
a
s
ed
o
n
th
is
r
ea
s
o
n
,
we
p
r
o
p
o
s
e
to
p
r
o
d
u
ce
a
s
y
s
tem
th
at
ca
n
r
ec
o
m
m
en
d
a
lis
t
o
f
r
elev
an
t
s
u
b
ject
h
ea
d
in
g
s
,
an
d
th
en
let
th
e
lib
r
ar
ian
s
to
d
ec
id
e
wh
ich
o
f
th
o
s
e
ar
e
ap
p
r
o
p
r
i
ate
to
b
e
ass
ig
n
ed
to
a
g
iv
en
d
o
cu
m
e
n
t.
Her
e,
we
u
s
e
in
f
o
r
m
atio
n
r
etr
iev
al
m
et
h
o
d
s
,
i.e
.
,
q
u
er
y
lik
elih
o
o
d
la
n
g
u
ag
e
m
o
d
el
(
L
M)
[
2
4
]
,
[
2
5
]
an
d
v
ec
to
r
s
p
ac
e
m
o
d
el
(
VSM)
[
2
6
]
,
[
2
7
]
,
to
g
e
n
er
ate
th
e
r
ec
o
m
m
en
d
atio
n
lis
t.
I
n
s
u
m
m
a
r
y
,
o
u
r
c
o
n
tr
ib
u
ti
o
n
s
ar
e
as
f
o
llo
ws:
i
)
.
W
e
c
o
n
d
u
ct
a
n
o
v
el
s
tu
d
y
o
n
th
e
ef
f
ec
tiv
en
ess
o
f
in
f
o
r
m
atio
n
r
etr
iev
al
m
o
d
els,
i.e
.
,
L
M
an
d
VSM,
to
au
to
m
a
tically
g
en
er
ate
a
r
an
k
ed
lis
t
o
f
r
elev
an
t
s
u
b
jec
t
h
ea
d
in
g
s
,
with
th
e
g
o
al
to
g
iv
e
a
r
ec
o
m
m
en
d
atio
n
f
o
r
lib
r
ar
ian
s
in
d
eter
m
in
in
g
th
e
s
u
b
ject
h
ea
d
i
n
g
s
ef
f
ec
tiv
ely
a
n
d
e
f
f
icien
tly
.
T
o
th
e
b
est
o
f
o
u
r
k
n
o
wled
g
e,
n
o
n
e
o
f
th
e
p
r
ev
io
u
s
p
u
b
lis
h
ed
wo
r
k
h
av
e
r
ep
o
r
ted
ab
o
u
t
th
is
s
tu
d
y
.
ii
)
.
W
e
p
er
f
o
r
m
an
em
p
ir
ical
ev
alu
atio
n
to
test
th
e
ef
f
ec
tiv
en
ess
o
f
L
M
an
d
VSM
r
etr
iev
al
m
eth
o
d
s
in
g
en
er
atin
g
s
u
b
je
ct
h
ea
d
i
n
g
s
r
ec
o
m
m
e
n
d
atio
n
u
s
in
g
o
u
r
co
llectio
n
o
f
b
a
ch
elo
r
th
esis
f
r
o
m
Un
iv
er
s
itas
I
n
d
o
n
esia
a
n
d
t
h
r
ee
d
if
f
er
e
n
t
s
ea
r
ch
u
n
its
:
titl
e,
ab
n
s
tr
ac
t,
a
n
d
titl
e+
ab
s
tr
ac
t.
iii
)
.
W
e
an
aly
ze
d
if
f
er
en
t
f
ac
u
lties
th
at
o
f
ten
s
h
ar
e
s
im
ilar
s
u
b
ject
h
ea
d
i
n
g
s
.
T
h
e
r
esu
lt
o
f
t
h
is
an
aly
s
is
ca
n
s
u
g
g
est
th
e
p
o
ten
tial r
esear
ch
co
lla
b
o
r
atio
n
b
etwe
en
th
ese
f
ac
u
lties
.
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:
2502
-
4
7
5
2
Dete
r
min
in
g
s
u
b
ject
h
ea
d
in
g
s
o
f d
o
cu
men
ts
u
s
in
g
in
fo
r
ma
tio
n
r
etri
ev
a
l m
o
d
els
(
E
vi
Yu
lia
n
ti
)
1051
2.
P
RO
B
L
E
M
ST
A
T
E
M
E
NT
T
h
e
task
th
at
is
s
tu
d
ied
in
th
is
wo
r
k
is
s
u
b
ject
h
ea
d
in
g
r
ec
o
m
m
en
d
atio
n
/r
et
r
iev
al
p
r
o
b
lem
,
aim
ed
to
h
elp
lib
r
ar
ian
s
in
ass
ig
n
in
g
s
u
b
ject
h
ea
d
in
g
s
m
o
r
e
ef
f
ec
tiv
ely
an
d
ef
f
icien
tly
.
Ou
r
g
e
n
er
al
f
r
am
ewo
r
k
is
illu
s
tr
ated
in
Fig
u
r
e
1
.
Giv
e
n
a
d
o
cu
m
en
t
m
etad
ata
(
th
at
wil
l
b
e
ass
ig
n
ed
th
e
s
u
b
ject
h
ea
d
i
n
g
s
)
as
a
q
u
er
y
,
we
will
r
etr
iev
e
a
r
an
k
ed
lis
t
o
f
s
u
b
ject
h
ea
d
in
g
s
f
r
o
m
th
e
m
o
s
t
s
im
ilar
d
o
cu
m
en
t
m
etad
at
a
in
th
e
co
llectio
n
u
s
in
g
L
M
a
n
d
VSM
r
etr
iev
al
m
eth
o
d
s
as
th
e
o
u
t
p
u
t.
I
t
is
ex
p
ec
ted
th
at
th
e
o
u
tp
u
t
ca
n
d
is
p
lay
r
elev
an
t
s
u
b
ject
h
ea
d
in
g
s
in
th
e
ea
r
l
y
p
o
s
itio
n
o
f
th
e
r
ec
o
m
m
en
d
atio
n
lis
t,
s
o
th
at
lib
r
ar
ian
s
ca
n
f
in
d
th
em
q
u
ick
ly
.
T
h
e
b
asic
id
ea
o
f
u
tili
zin
g
d
o
cu
m
en
t
r
etr
iev
al
m
eth
o
d
s
to
g
en
e
r
ate
s
u
b
ject
h
ea
d
in
g
r
ec
o
m
m
en
d
atio
n
is
th
at
“two
d
o
cu
m
e
n
ts
th
at
ar
e
ass
ig
n
ed
s
im
ilar
s
u
b
ject
h
ea
d
in
g
s
u
s
u
a
lly
d
escr
ib
e
a
b
o
u
t
th
e
s
am
e
t
o
p
ics
an
d
th
er
ef
o
r
e
th
ey
m
ay
h
av
e
h
ig
h
d
o
c
u
m
e
n
t
s
im
ilar
ity
”.
T
h
e
s
im
ilar
ity
b
etwe
en
d
o
cu
m
en
t
s
is
ass
e
s
s
ed
u
s
in
g
titl
e
an
d
ab
s
tr
ac
t te
x
t r
ep
r
esen
tatio
n
,
s
in
ce
o
u
r
d
ataset
d
o
es n
o
t
h
av
e
t
h
e
f
u
ll tex
t o
f
d
o
cu
m
en
ts
(
s
ee
Sectio
n
4
.
1
)
.
Fig
u
r
e
1
.
Ou
r
g
en
e
r
al
f
r
am
ew
o
r
k
I
n
itially
o
u
r
in
d
ex
in
g
d
ata
is
in
d
ex
ed
u
s
in
g
f
ield
in
d
ex
i
n
g
[
2
8
]
,
[
2
9
]
,
s
o
we
h
av
e
titl
e
a
n
d
a
b
s
tr
ac
t
f
ield
s
in
th
e
in
d
e
x
.
T
h
is
en
ab
les
u
s
to
s
ea
r
ch
o
n
titl
e
an
d
a
b
s
tr
ac
t
p
ar
ts
in
d
iv
id
u
ally
.
Mo
r
e
s
p
ec
if
ically
,
th
e
titl
e
o
f
q
u
e
r
ies
is
co
m
p
ar
e
d
a
g
ain
s
t
th
e
titl
e
f
ield
o
f
th
e
in
d
e
x
,
an
d
th
e
ab
s
tr
ac
t
o
f
q
u
er
ies
is
co
m
p
ar
ed
ag
ain
s
t
th
e
ab
s
tr
ac
t
f
ield
o
f
th
e
in
d
e
x
.
T
h
is
way
,
in
g
e
n
er
al
we
c
alcu
lated
s
im
ilar
ity
o
n
t
h
e
titl
es
an
d
ab
s
tr
ac
ts
o
f
d
o
cu
m
e
n
ts
.
T
h
e
L
M
an
d
VSM
r
etr
iev
al
m
eth
o
d
s
will
r
an
k
k
d
o
cu
m
e
n
t
m
etad
a
ta
in
th
e
co
llectio
n
th
at
ar
e
m
o
s
t
s
im
ilar
to
th
e
q
u
er
y
ac
co
r
d
in
g
t
o
th
e
titl
es
an
d
ab
s
tr
ac
ts
.
T
h
en
,
th
e
s
u
b
ject
h
ea
d
in
g
s
o
f
th
e
k
to
p
-
r
an
k
e
d
d
o
cu
m
e
n
ts
ar
e
ex
tr
ac
ted
as
th
e
s
u
b
ject
h
ea
d
in
g
r
ec
o
m
m
e
n
d
atio
n
lis
t.
No
te
th
at
a
m
etad
at
a
m
ay
h
av
e
s
ev
er
al
s
u
b
ject
h
ea
d
in
g
s
.
As
a
r
esu
lt,
th
er
e
co
u
ld
b
e
m
o
r
e
th
a
n
o
n
e
s
u
b
ject
h
ea
d
in
g
at
ev
er
y
r
an
k
in
th
e
to
p
k
r
ec
o
m
m
en
d
atio
n
lis
t.
T
h
e
n
u
m
b
er
o
f
m
etad
ata
r
etr
ie
v
ed
in
th
e
to
p
r
an
k
(
k
)
ca
n
b
e
s
et
ac
co
r
d
in
g
to
p
eo
p
le
n
ee
d
s
.
I
n
o
u
r
e
x
p
er
im
e
n
ts
,
we
r
etr
iev
e
1
,
5
,
an
d
1
0
d
o
c
u
m
en
ts
in
th
e
to
p
r
an
k
,
an
d
tak
e
th
e
ir
s
u
b
ject
h
ea
d
in
g
s
in
th
e
r
ec
o
m
m
en
d
atio
n
lis
t.
3.
RE
T
RI
E
VA
L
M
E
T
H
O
DS
3
.
1
.
Q
uery
lik
eliho
o
d la
ng
ua
g
e
m
o
del
(
L
M
)
L
an
g
u
ag
e
m
o
d
el
(
L
M)
u
s
es
p
r
o
b
ab
ilis
tic
co
n
ce
p
t
to
esti
m
ate
th
e
p
r
o
b
a
b
ilit
y
o
f
r
el
ev
an
ce
o
f
d
o
cu
m
e
n
ts
to
u
s
er
q
u
er
ies
[
2
4
]
,
[
2
5
]
.
A
la
n
g
u
a
g
e
m
o
d
el
M
d
f
o
r
ea
ch
d
o
c
u
m
en
t
d
is
in
f
er
r
e
d
an
d
th
en
u
s
ed
to
esti
m
ate
th
e
p
r
o
b
ab
ilit
y
o
f
g
en
er
atin
g
t
h
e
q
u
er
y
ac
co
r
d
in
g
to
t
h
e
m
o
d
el
p
(
q
|M
d
)
.
T
h
is
p
r
o
b
ab
ilit
y
is
also
r
ef
er
r
ed
to
as
q
u
er
y
g
e
n
er
atio
n
p
r
o
b
a
b
ilit
y
.
Do
cu
m
en
ts
in
th
e
co
llectio
n
ar
e
th
en
r
an
k
e
d
ac
co
r
d
in
g
to
th
ese
p
r
o
b
a
b
ilit
ies.
T
h
e
h
ig
h
er
th
e
p
r
o
b
a
b
ilit
y
o
f
a
d
o
c
u
m
en
t,
th
e
m
o
r
e
p
r
o
b
a
b
le
th
e
q
u
er
y
ca
n
b
e
g
e
n
er
ated
f
r
o
m
th
e
d
o
cu
m
e
n
t,
wh
ich
im
p
lies
th
at
th
e
m
o
r
e
r
elev
a
n
t
th
e
d
o
c
u
m
en
t
to
th
e
q
u
er
y
.
T
h
e
p
r
o
b
ab
ilit
y
o
f
p
r
o
d
u
cin
g
th
e
q
u
er
y
q
g
i
v
en
th
e
lan
g
u
a
g
e
m
o
d
el
o
f
d
o
cu
m
en
t
d
,
i.e
.
,
M
d
,
u
s
in
g
Ma
x
im
u
m
L
i
k
elih
o
o
d
E
s
tim
ate
an
d
Dir
ich
let
s
m
o
o
th
in
g
ca
n
b
e
c
o
m
p
u
ted
as f
o
llo
ws
(
1
)
:
lo
g
(
|
)
=
∑
l
og
,
+
|
|
|
|
+
∈
(
1
)
w
h
e
r
e
t
is
a
t
e
r
m
i
n
q
u
e
r
y
q
,
,
d
e
n
o
t
e
s
t
h
e
f
r
e
q
u
e
n
c
y
o
f
t
er
m
t
i
n
q
u
e
r
y
q
,
(
|
)
i
s
p
r
o
b
a
b
i
l
it
y
o
f
g
e
n
e
r
a
t
i
n
g
t
e
r
m
t
f
r
o
m
d
o
c
u
m
e
n
t
l
a
n
g
u
a
g
e
m
o
d
e
l
M
d
,
a
n
d
|
d
|
d
e
n
o
t
e
s
t
o
t
a
l
n
u
m
b
e
r
o
f
t
e
r
m
s
i
n
d
o
c
u
m
e
n
t
d
,
|
C
|
d
e
n
o
t
e
s
t
h
e
t
o
t
al
t
e
r
m
s
i
n
t
h
e
co
l
l
e
c
t
i
o
n
,
cf
t
d
e
n
o
t
es
t
h
e
f
r
e
q
u
en
c
y
o
f
t
e
r
m
t
i
n
t
h
e
c
o
l
l
e
ct
i
o
n
.
D
i
r
i
c
h
l
et
s
m
o
o
t
h
i
n
g
i
s
a
d
e
f
a
u
lt
s
m
o
o
t
h
i
n
g
te
c
h
n
i
q
u
e
p
r
o
v
i
d
e
d
i
n
t
h
e
s
e
a
r
c
h
e
n
g
i
n
e
l
i
b
r
a
r
y
I
n
d
r
i
(
d
e
f
a
u
l
t
v
a
l
u
e
i
s
2
5
0
0
)
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci,
Vo
l.
23
,
No
.
2
,
Au
g
u
s
t 2
0
2
1
:
1
0
4
9
-
1
0
5
8
1052
3
.
2
.
Vec
t
o
r
s
pa
ce
m
o
del
(
VSM
)
Vec
to
r
s
p
ac
e
m
o
d
el
(
VSM)
r
ep
r
esen
ts
d
o
cu
m
en
ts
as
v
ec
to
r
s
o
f
ter
m
s
in
th
e
v
ec
to
r
s
p
ac
e
[
2
6
]
,
[
2
7
]
.
C
o
s
in
e
s
im
ilar
ity
is
th
en
u
s
ed
to
ca
lcu
late
th
e
s
im
ilar
ity
b
etwe
en
two
d
o
cu
m
e
n
t
v
ec
to
r
s
i
n
th
e
v
ec
to
r
s
p
ac
e.
W
h
en
two
d
o
cu
m
e
n
ts
s
h
ar
e
m
an
y
s
im
ilar
wo
r
d
s
,
th
en
th
e
ir
v
ec
to
r
s
in
th
e
v
ec
t
o
r
s
p
ac
e
is
clo
s
e
(
th
e
an
g
le
b
etwe
en
th
em
is
s
m
all)
.
As
a
r
esu
lt,
th
e
co
s
in
e
o
f
t
h
e
an
g
le
b
etwe
en
th
em
is
h
ig
h
.
T
wo
id
en
tical
d
o
c
u
m
en
ts
will h
av
e
co
s
in
e
s
im
ilar
ity
s
co
r
e
o
f
1
.
T
h
e
g
r
ea
ter
th
e
s
im
ilar
ity
b
etwe
en
two
d
o
c
u
m
en
ts
,
th
e
co
s
in
e
s
im
ilar
ity
s
co
r
e
will
b
e
g
ettin
g
clo
s
er
t
o
1
.
T
h
e
r
an
g
e
o
f
co
s
in
e
s
im
ilar
ity
s
co
r
e
is
b
etwe
en
0
a
n
d
1
.
T
h
e
f
o
r
m
u
la
to
co
m
p
u
te
c
o
s
in
e
s
im
ilar
ity
is
as f
o
llo
ws
(
2
)
:
(
,
)
=
∑
.
(
,
)
=
1
√
∑
2
=
1
.
∑
2
=
1
(
2
)
wh
er
e
is
th
e
weig
h
t
o
f
te
r
m
k
in
d
o
c
u
m
en
t
;
is
th
e
weig
h
t
o
f
ter
m
k
in
d
o
cu
m
en
t
;
is
th
e
n
u
m
b
er
o
f
ter
m
s
in
d
o
c
u
m
en
t
;
an
d
is
th
e
n
u
m
b
e
r
o
f
ter
m
s
in
d
o
cu
m
e
n
t
.
T
h
e
weig
h
t
o
f
ter
m
s
is
ca
lcu
lated
u
s
in
g
T
F
-
I
DF f
o
r
m
u
la.
4.
E
XP
E
R
I
M
E
N
T
4
.
1
.
Da
t
a
s
et
T
h
is
wo
r
k
u
s
es
a
m
etad
ata
c
o
l
lectio
n
o
f
b
ac
h
elo
r
th
esis
o
f
U
I
s
tu
d
en
ts
as
o
u
r
d
ataset.
T
h
is
m
etad
ata
co
llectio
n
was
cr
awle
d
f
r
o
m
UI
lib
r
ar
y
web
s
ite
(
h
ttp
://l
ib
.
u
i.a
c.
id
/
)
in
th
e
s
ec
tio
n
“UI
-
Sk
r
ip
s
i
(
Op
en
)
”
u
s
in
g
a
p
r
o
g
r
am
wr
itten
in
Per
l
an
d
L
W
P
m
o
d
u
le.
T
h
is
co
llectio
n
co
n
tain
s
d
o
cu
m
en
t
m
etad
at
a
o
f
b
ac
h
elo
r
th
esis
f
r
o
m
1
4
f
ac
u
lties
in
UI
d
u
r
in
g
2
0
0
8
-
2
0
1
1
,
with
a
to
tal
o
f
1
4
,
7
2
2
m
etad
ata.
I
t
is
im
p
o
r
t
an
t
to
n
o
te
th
at
w
e
us
ed
d
o
c
u
m
en
t
m
eta
d
ata,
in
s
tead
o
f
f
u
ll
tex
t,
b
ec
au
s
e
m
o
s
t
o
f
d
o
cu
m
en
ts
in
th
e
web
s
ite
d
o
n
o
t
h
av
e
f
u
ll
tex
t
(
an
d
s
o
m
e
f
u
ll tex
t o
f
d
o
c
u
m
e
n
ts
ca
n
n
o
t b
e
ac
ce
s
s
ed
s
in
ce
t
h
ey
ar
e
s
ec
u
r
el
y
en
cr
y
p
ted
)
.
W
e
o
n
ly
c
r
awl
f
o
u
r
in
f
o
r
m
at
io
n
f
o
r
ea
ch
m
etad
ata
t
h
at
is
n
ee
d
e
d
b
y
o
u
r
m
eth
o
d
:
titl
e,
ab
s
tr
ac
t,
f
ac
u
lty
,
a
n
d
s
u
b
ject
h
ea
d
in
g
.
Ab
s
tr
ac
t
co
n
tain
s
th
e
s
u
m
m
ar
y
o
f
th
e
th
esis
.
T
itle
an
d
ab
s
tr
ac
t
in
f
o
r
m
atio
n
a
r
e
im
p
o
r
tan
t
as
th
e
m
etad
ata
r
ep
r
esen
tatio
n
f
o
r
in
d
ex
in
g
a
n
d
q
u
er
y
in
g
p
u
r
p
o
s
e.
R
elate
d
m
etad
ata
in
t
h
e
co
llectio
n
ar
e
d
eter
m
in
e
d
b
y
o
u
r
m
eth
o
d
s
b
ased
o
n
th
e
s
im
ilar
ity
o
n
th
ese
tex
t.
Facu
lty
in
f
o
r
m
atio
n
is
im
p
o
r
tan
t
f
o
r
o
u
r
an
aly
s
is
o
n
cr
o
s
s
-
f
ac
u
lty
s
im
ilar
d
o
c
u
m
en
ts
.
Su
b
ject
h
ea
d
in
g
in
f
o
r
m
atio
n
is
s
u
r
ely
im
p
o
r
tan
t
as
th
is
will
b
e
ex
tr
ac
ted
f
r
o
m
r
elate
d
m
etad
ata
t
o
p
r
o
d
u
ce
r
ec
o
m
m
en
d
atio
n
.
I
n
ad
d
itio
n
,
it
is
also
u
s
ed
as
a
g
o
ld
s
tan
d
ar
d
in
t
h
e
ev
alu
atio
n
p
r
o
ce
s
s
.
I
t
is
im
p
o
r
tan
t
to
n
o
te
t
h
at
o
n
e
d
o
cu
m
e
n
t
m
ay
b
e
ass
ig
n
ed
s
ev
er
al
s
u
b
ject
h
ea
d
in
g
s
.
Fig
u
r
e
2
d
escr
ib
es a
n
ex
am
p
le
o
f
s
u
b
ject
h
ea
d
in
g
s
f
r
o
m
two
m
etad
ata
in
o
u
r
d
ataset.
T
o
h
e
lp
r
ea
d
er
s
u
n
d
er
s
tan
d
th
e
co
n
te
x
t
o
f
th
e
m
etad
ata
w
h
er
e
t
h
is
s
u
b
ject
h
ea
d
in
g
co
m
e
f
r
o
m
,
we
also
d
is
p
lay
th
e
titl
e
o
f
m
etad
ata.
T
h
ese
two
m
etad
ata
h
av
e
m
o
r
e
t
h
an
o
n
e
s
u
b
ject
h
ea
d
in
g
th
at
ar
e
d
elim
ited
with
“;”
ch
ar
ac
ter
.
F
o
r
th
e
s
ec
o
n
d
m
eta
d
ata,
it
h
as
th
e
s
u
b
ject
h
ea
d
in
g
“M
ater
n
al
-
C
h
ild
Nu
r
s
in
g
”
th
at
is
co
m
p
o
s
ed
o
f
m
ain
s
u
b
ject
h
ea
d
in
g
“M
ater
n
al”
an
d
ad
d
iti
o
n
al
s
u
b
ject
h
ea
d
i
n
g
“Ch
ild
N
u
r
s
in
g
”.
Fig
u
r
e
2
.
E
x
am
p
les o
f
s
u
b
ject
h
ea
d
in
g
s
o
f
d
o
c
u
m
en
ts
As
we
r
eq
u
ir
e
ea
ch
m
etad
at
a
in
o
u
r
d
ataset
to
h
av
e
titl
e
,
ab
s
tr
ac
t,
s
u
b
ject
h
ea
d
in
g
,
a
n
d
f
ac
u
lty
in
f
o
r
m
atio
n
,
s
o
we
ex
clu
d
e
all
m
etad
ata
th
at
d
o
n
o
t
h
a
v
e
an
y
o
f
th
is
in
f
o
r
m
atio
n
.
I
t
r
esu
lt
s
in
9
,
2
6
1
m
etad
ata
in
o
u
r
d
ataset.
T
h
e
s
tatis
tic
s
o
f
to
tal
d
o
cu
m
en
t
s
f
o
r
ea
ch
f
ac
u
lty
is
d
escr
ib
ed
in
T
ab
le
2
.
W
e
ca
n
s
ee
f
r
o
m
th
e
tab
le
th
at
m
o
s
t
m
etad
ata
in
o
u
r
d
ataset
co
m
es
f
r
o
m
Facu
lt
y
o
f
E
n
g
in
ee
r
in
g
.
T
h
is
m
ay
b
e
d
u
e
to
th
e
h
i
g
h
n
u
m
b
er
o
f
s
tu
d
y
p
r
o
g
r
a
m
in
t
h
is
f
ac
u
lty
wh
ich
r
esu
lts
in
th
e
h
ig
h
n
u
m
b
e
r
o
f
s
tu
d
en
ts
.
Acc
o
r
d
in
g
to
th
e
d
ata
o
f
UI
’
s
b
ac
h
elo
r
s
tu
d
en
t
ad
m
is
s
io
n
in
2
0
2
0
,
th
e
Facu
lty
o
f
E
n
g
in
ee
r
in
g
h
as
th
e
h
i
g
h
est
n
u
m
b
er
o
f
s
tu
d
e
n
ts
am
o
n
g
all
f
ac
u
lt
ies
in
UI
.
T
o
en
s
u
r
e
th
at
ea
ch
f
ac
u
lty
h
as
en
o
u
g
h
m
etad
ata,
we
d
ec
id
e
t
o
ex
clu
d
e
th
e
o
n
es
th
at
h
av
e
a
s
m
all
n
u
m
b
er
o
f
m
etad
ata.
W
e
m
ak
e
a
ju
s
tific
atio
n
o
f
1
0
0
d
o
cu
m
e
n
ts
as
o
u
r
lo
wer
b
o
u
n
d
o
f
th
e
Ti
t
le
:
An
a
li
s
i
s
s
e
le
c
ti
v
i
ty
d
a
n
ma
r
k
e
t
ti
mi
n
g
p
a
d
a
r
e
k
s
a
d
a
n
a
s
a
ha
m
d
i
In
d
on
e
s
i
a
:
p
e
r
i
od
e
2
0
0
5
-
2007
(
A
na
l
ys
is
of
t
h
e
s
e
l
e
ct
ivi
t
y
a
n
d
m
a
rk
e
t
t
i
m
ing
on
e
qu
it
y
fu
nd
s
i
n
I
nd
o
ne
s
ia
:
2
0
0
5
-
2
0
0
7
pe
rio
d
)
S
u
b
j
ect
hea
d
in
g
s
:
Mu
tu
al f
u
n
d
s
; Sha
r
es
Ti
t
le:
H
u
bu
n
g
a
n
ti
n
g
k
a
t
p
e
n
g
e
ta
hua
n
s
u
a
mi
d
e
n
g
a
n
p
r
a
k
ti
k
p
e
mb
e
r
i
a
n
AS
I
e
k
s
k
l
u
s
i
f
d
i
R
W
2
5
Ba
k
ti
ja
y
a
Su
k
ma
ja
y
a
D
e
p
ok
(
T
h
e
cor
relat
ion
bet
wee
n
h
u
s
band
's
k
nowl
e
g
d
e
l
e
v
e
l
s
of
e
x
cl
u
s
ive
br
e
a
s
t
fe
e
d
ing
pra
ct
ice
in
R
W
2
5
B
a
k
t
ija
ya
,
S
u
k
m
a
ja
ya
,
D
e
po
k
)
S
ub
j
et
head
i
n
gs
:
Bre
a
s
t
fe
e
d
in
g;
M
a
t
e
rnal
--
C
h
il
d
Nu
rs
ing
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:
2502
-
4
7
5
2
Dete
r
min
in
g
s
u
b
ject
h
ea
d
in
g
s
o
f d
o
cu
men
ts
u
s
in
g
in
fo
r
ma
tio
n
r
etri
ev
a
l m
o
d
els
(
E
vi
Yu
lia
n
ti
)
1053
n
u
m
b
er
o
f
m
etad
ata
f
o
r
ea
ch
f
ac
u
lty
.
T
h
is
ca
u
s
es
th
e
m
etad
a
ta
f
r
o
m
f
o
u
r
f
ac
u
lties
(
i.e
.
,
Fa
cu
lty
o
f
Ph
ar
m
ac
y
,
Facu
lty
o
f
C
o
m
p
u
ter
Scien
ce
,
Facu
lty
o
f
Den
tis
tr
y
,
an
d
Facu
lty
o
f
Ad
m
in
is
tr
ativ
e
Scien
ce
)
to
b
e
f
ilter
ed
o
u
t
f
r
o
m
o
u
r
d
ataset.
T
h
ey
r
esp
ec
t
iv
ely
h
av
e
9
5
,
9
4
,
8
4
,
an
d
3
5
d
o
cu
m
e
n
ts
.
Fin
ally
,
th
er
e
a
r
e
8
,
9
5
3
m
etad
ata
t
h
at
co
m
es f
r
o
m
1
0
f
ac
u
lt
ies
at
UI
in
o
u
r
f
in
al
d
ataset.
T
ab
le
3
d
escr
ib
es
th
e
ter
m
s
t
atis
tics
in
in
o
u
r
d
at
aset.
On
av
er
ag
e,
titl
e
co
n
tain
s
ar
o
u
n
d
2
3
ter
m
s
,
wh
ich
is
q
u
ite
lo
n
g
f
o
r
a
titl
e.
W
e
f
o
u
n
d
th
is
h
ap
p
en
ed
b
ec
a
u
s
e
s
o
m
e
m
etad
ata
h
as a
m
ix
ed
o
f
I
n
d
o
n
esian
an
d
E
n
g
lis
h
titl
es
in
th
e
web
s
ite,
wh
ich
m
ak
es
t
h
e
o
v
er
all
titl
e
lo
n
g
.
T
h
e
a
v
er
ag
e
ter
m
in
a
b
s
tr
ac
t
is
ar
o
u
n
d
2
0
5
wo
r
d
s
.
T
h
e
av
er
ag
e
n
u
m
b
e
r
o
f
s
u
b
ject
h
ea
d
in
g
s
in
a
m
etad
ata
is
1
.
5
.
T
h
is
d
en
o
tes
th
at
m
o
s
t
o
f
m
etad
ata
ar
e
ass
ig
n
ed
m
o
r
e
th
a
n
o
n
e
s
u
b
jec
t h
ea
d
in
g
s
th
at
co
n
s
is
ts
o
f
3
ter
m
s
o
n
av
e
r
ag
e.
T
ab
le
2
.
T
h
e
s
tatis
tics
o
f
to
tal
m
etad
ata
f
o
r
ea
c
h
f
ac
u
lty
in
o
u
r
d
ataset
F
a
c
u
l
t
y
A
b
b
r
e
v
i
a
t
i
o
n
#
st
u
d
y
p
r
o
g
r
a
m
#
d
o
c
u
m
e
n
t
m
e
t
a
d
a
t
a
F
a
c
u
l
t
y
o
f
E
n
g
i
n
e
e
r
i
n
g
En
g
13
2
,
4
0
0
F
a
c
u
l
t
y
o
f
H
u
ma
n
i
t
i
e
s
H
u
m
15
1
,
4
8
7
F
a
c
u
l
t
y
o
f
M
a
t
h
e
ma
t
i
c
s
a
n
d
N
a
t
u
r
a
l
S
c
i
e
n
c
e
M
a
t
h
9
1
,
1
8
8
F
a
c
u
l
t
y
o
f
S
o
c
i
a
l
a
n
d
P
o
l
i
t
i
c
a
l
S
c
i
e
n
c
e
s
S
o
c
7
1
,
0
8
1
F
a
c
u
l
t
y
o
f
P
u
b
l
i
c
H
e
a
l
t
h
P
u
b
H
4
9
7
1
F
a
c
u
l
t
y
o
f
L
a
w
La
w
1
7
1
2
F
a
c
u
l
t
y
o
f
E
c
o
n
o
m
y
a
n
d
B
u
s
i
n
e
ss
Ec
o
5
5
2
1
F
a
c
u
l
t
y
o
f
P
s
y
c
h
o
l
o
g
y
P
sy
1
2
5
3
F
a
c
u
l
t
y
o
f
M
e
d
i
c
i
n
e
M
e
d
1
2
1
9
F
a
c
u
l
t
y
o
f
N
u
r
si
n
g
N
u
r
s
1
1
2
1
To
t
a
l
9
,
2
6
1
T
ab
le
3
.
T
h
e
s
tatis
tics
o
f
ter
m
s
an
d
s
u
b
ject
h
ea
d
i
n
g
s
in
o
u
r
d
ataset
A
v
e
r
a
g
e
t
e
r
ms
f
o
r
e
a
c
h
t
i
t
l
e
2
3
.
1
2
A
v
e
r
a
g
e
t
e
r
ms
f
o
r
e
a
c
h
a
b
st
r
a
c
t
2
0
5
.
1
6
A
v
e
r
a
g
e
t
e
r
ms
f
o
r
e
a
c
h
su
b
j
e
c
t
h
e
a
d
i
n
g
3
.
1
7
A
v
e
r
a
g
e
n
u
m
b
e
r
o
f
s
u
b
j
e
c
t
h
e
a
d
i
n
g
s
f
o
r
e
a
c
h
m
e
t
a
d
a
t
a
1
.
5
1
Ou
r
d
ataset
is
s
p
lit
in
to
2
p
ar
ts
:
in
d
ex
in
g
d
ata
an
d
q
u
e
r
ies
.
T
h
is
is
im
p
o
r
tan
t
b
ec
au
s
e
w
e
h
av
e
to
en
s
u
r
e
th
at
th
e
m
etad
ata
to
b
e
q
u
er
ied
(
e
v
alu
ated
)
d
o
n
o
t
ap
p
ea
r
in
th
e
in
d
ex
.
Oth
er
wis
e,
t
h
e
r
etr
iev
al
s
y
s
tem
will
alwa
y
s
o
b
tain
p
er
f
ec
t
m
a
tch
as
th
e
m
o
s
t
-
r
elev
an
t
m
eta
d
ata
f
o
u
n
d
is
b
asically
th
e
m
etad
ata
in
p
u
t
its
elf
.
Fo
r
th
is
r
ea
s
o
n
,
th
e
d
ataset
h
as
to
b
e
s
p
lit.
W
e
tak
e
a
r
an
d
o
m
s
am
p
le
o
f
3
0
m
etad
a
ta
f
o
r
ea
c
h
f
ac
u
lt
y
,
r
esu
ltin
g
3
0
0
m
etad
ata
as
th
e
q
u
er
ies
.
T
h
e
n
th
e
r
e
m
ain
in
g
d
ata
with
th
e
to
tal
o
f
8
,
6
5
3
m
etad
ata
is
u
s
ed
f
o
r
in
d
ex
in
g
.
4
.
2
.
I
m
plem
ent
a
t
io
n
W
e
u
s
e
a
s
ea
r
ch
en
g
in
e
lib
r
ar
y
I
n
d
r
i
v
e
r
s
io
n
5
.
1
2
f
o
r
in
d
ex
in
g
a
n
d
r
et
r
iev
al
u
s
in
g
L
M
r
etr
iev
al
m
eth
o
d
.
T
h
en
,
we
u
s
e
a
s
ea
r
c
h
en
g
in
e
lib
r
a
r
y
L
u
ce
n
e
v
er
s
i
o
n
8
.
7
.
0
to
im
p
lem
en
t
in
d
e
x
in
g
an
d
r
etr
ie
v
al
u
s
in
g
VSM
r
etr
iev
al
m
eth
o
d
.
Fo
r
e
ac
h
m
eth
o
d
,
in
d
ex
is
b
u
ilt
f
o
r
titl
e
an
d
ab
s
tr
ac
t
f
ield
s
.
W
h
en
th
e
s
ea
r
ch
u
n
it
is
titl
e,
we
co
m
p
u
te
titl
e
s
im
ilar
ities
b
etwe
en
q
u
er
y
’
s
titl
e
an
d
d
o
c
u
m
en
ts
’
titl
e
in
th
e
in
d
e
x
.
W
h
en
th
e
s
ea
r
ch
u
n
it
is
ab
s
tr
ac
t,
we
co
m
p
u
te
ab
s
tr
ac
t
s
im
ilar
ities
b
etwe
en
q
u
er
y
’
s
ab
s
tr
ac
t
an
d
d
o
c
u
m
e
n
ts
’
ab
s
tr
ac
t
in
th
e
in
d
ex
.
W
h
en
th
e
s
ea
r
ch
u
n
it
i
s
titl
e+
ab
s
tr
ac
t,
th
en
we
co
m
p
u
te
b
o
th
titl
e
s
im
ilar
ities
an
d
ab
s
tr
ac
t
s
im
ilar
itie
s
b
etwe
en
q
u
er
y
an
d
d
o
cu
m
en
ts
in
th
e
in
d
ex
.
T
o
ex
am
i
n
e
wh
i
ch
o
n
e
is
m
o
r
e
i
m
p
o
r
tan
t
b
etwe
en
titl
e
r
elev
an
ce
an
d
ab
s
tr
ac
t
r
elev
an
ce
,
we
ex
p
er
im
en
ted
with
d
if
f
e
r
en
t
γ
v
alu
es
in
th
e
r
an
g
e
o
f
0
-
1
with
th
e
s
tep
o
f
1
.
I
n
d
r
i
q
u
er
y
o
p
e
r
ato
r
s
#
c
o
mb
in
e
a
n
d
#
w
eig
h
t
is
u
s
ed
f
o
r
co
m
p
u
tin
g
titl
e
an
d
ab
s
tr
ac
t
r
elev
an
ce
u
s
in
g
L
M.
#
co
m
b
in
e
o
p
er
at
o
r
is
u
s
ed
to
co
m
b
in
e
all
ter
m
s
in
t
h
e
titl
e
o
r
ab
s
tr
ac
t,
wh
ile
#
weig
h
t
o
p
e
r
ato
r
is
u
s
ed
to
g
iv
e
th
e
im
p
o
r
tan
ce
o
f
titl
e
r
elev
an
ce
an
d
ab
s
tr
ac
t
r
elev
an
ce
.
T
h
e
f
o
llo
win
g
is
I
n
d
r
i
q
u
e
r
y
f
o
r
m
u
la
f
o
r
f
ield
r
etr
iev
al:
#
w
eig
h
t(
1
-
γ
#
c
o
mb
in
e[ti
tle]
(
t
1,
title
t
2,
title
… t
m,
title
)
γ
#
co
mb
in
e[a
b
s
tr
a
ct](t
1,
abstract
t
2,
abstract
… t
n,
abstract
))
T
h
e
t
i,
j
in
d
icate
s
th
e
i
-
th
ter
m
i
n
th
e
j
f
ield
.
T
h
e
γ
v
al
u
e
d
en
o
tes
th
e
im
p
o
r
tan
ce
o
f
titl
e
r
ele
v
an
ce
an
d
ab
s
tr
ac
t
r
elev
an
ce
in
th
e
r
elev
an
ce
.
T
h
e
lo
wer
th
e
γ
v
alu
e,
t
h
e
m
o
r
e
im
p
o
r
tan
ce
th
e
titl
e
r
elev
an
ce
to
th
e
f
in
al
r
elev
an
ce
s
co
r
e.
W
h
en
γ
eq
u
a
ls
to
0
,
th
en
th
e
q
u
e
r
y
in
g
p
r
o
ce
s
s
o
n
ly
co
n
s
id
er
th
e
titl
e
r
elev
an
ce
.
Oth
er
wis
e,
wh
en
γ
eq
u
als to
1
,
th
e
n
we
o
n
ly
tak
e
in
to
ac
co
u
n
t th
e
ab
s
tr
ac
t r
elev
an
ce
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci,
Vo
l.
23
,
No
.
2
,
Au
g
u
s
t 2
0
2
1
:
1
0
4
9
-
1
0
5
8
1054
T
o
co
m
p
u
te
titl
e
an
d
ab
s
tr
ac
t
r
elev
an
ce
u
s
in
g
VSM
m
eth
o
d
,
we
u
s
e
B
o
o
s
tQu
ery
class
in
L
u
ce
n
e
lib
r
ar
y
.
I
t
n
ee
d
s
two
p
ar
a
m
eter
s
: in
s
tan
ce
o
f
Ter
mQu
ery
wh
i
ch
co
n
tain
s
th
e
q
u
er
y
ter
m
s
an
d
th
e
in
d
e
x
f
ield
i
n
wh
ich
th
e
s
ea
r
ch
will
b
e
p
er
f
o
r
m
ed
,
a
n
d
γ
v
alu
e
wh
ic
h
d
e
n
o
tes
th
e
weig
h
t
o
f
th
e
r
elev
an
ce
s
co
r
e
f
o
r
th
e
q
u
er
y
te
r
m
s
.
4
.
3
.
E
v
a
lua
t
io
n
E
v
alu
atio
n
is
p
er
f
o
r
m
e
d
b
y
c
o
m
p
ar
in
g
th
e
s
u
b
ject
h
ea
d
in
g
s
in
th
e
r
ec
o
m
m
en
d
atio
n
lis
t
with
th
e
g
r
o
u
n
d
tr
u
th
s
u
b
ject
h
ea
d
in
g
s
,
wh
ich
ar
e
o
b
tain
ed
f
r
o
m
th
e
ac
tu
al
s
u
b
ject
h
ea
d
in
g
s
o
f
th
e
q
u
er
ies.
R
ec
all
th
at
a
m
etad
ata
m
ay
h
av
e
s
ev
er
a
l
s
u
b
ject
h
ea
d
in
g
s
.
T
h
er
ef
o
r
e
,
if
s
u
b
ject
h
ea
d
in
g
s
r
et
r
iev
e
d
in
th
e
s
p
ec
if
ie
d
p
o
s
itio
n
o
f
th
e
r
ec
o
m
m
en
d
at
io
n
lis
t
co
n
tain
s
a
n
y
g
r
o
u
n
d
tr
u
th
s
u
b
ject
h
ea
d
in
g
s
o
f
th
e
q
u
er
ies,
t
h
en
w
e
co
n
s
id
er
s
u
ch
r
etr
iev
e
d
s
u
b
je
ct
h
ea
d
in
g
s
as
r
elev
an
t.
Fo
r
e
x
am
p
le:
g
iv
en
th
e
ac
t
u
al
s
u
b
ject
h
ea
d
in
g
s
f
o
r
a
q
u
er
y
is
“
B
r
ea
s
tfeed
in
g
;
Ma
t
ern
a
l
--
C
h
ild
N
u
r
s
in
g
”,
th
en
we
co
n
s
id
e
r
th
e
r
e
ar
e
th
r
ee
g
r
o
u
n
d
tr
u
t
h
s
u
b
ject
h
ea
d
in
g
s
f
o
r
th
is
q
u
er
y
:
“
b
r
ea
s
tfeed
in
g
”,
“
ma
tern
a
l
”,
an
d
“
c
h
ild
n
u
r
s
in
g
”.
I
f
s
u
b
ject
h
ea
d
i
n
g
s
r
etr
iev
ed
in
th
e
f
ir
s
t
r
an
k
o
f
th
e
r
ec
o
m
m
e
n
d
atio
n
is
“
B
r
ea
s
tfeed
in
g
--
S
o
cia
l
a
s
p
ec
ts
;
B
r
ea
s
t
m
ilk
”,
th
e
n
it
m
ea
n
s
th
at
it
c
o
n
s
is
ts
o
f
th
r
ee
s
u
b
ject
h
ea
d
i
n
g
s
:
“
b
r
ea
s
t
feed
in
g
”
,
“
s
o
cia
l
a
s
p
ec
t
s
”,
an
d
“
b
r
ea
s
t
milk
”.
S
in
ce
it
co
n
tain
s
o
n
e
g
r
o
u
n
d
tr
u
t
h
s
u
b
ject
h
ea
d
in
g
f
o
r
th
e
g
i
v
en
q
u
er
y
,
i.e
.
,
“
b
r
ea
s
t fe
ed
in
g
”,
th
e
n
it is
co
n
s
id
er
e
d
as r
elev
an
t.
So
m
e
r
etr
iev
al
m
ea
s
u
r
es
a
r
e
u
s
ed
to
ev
alu
ate
th
e
ef
f
ec
tiv
en
ess
o
f
L
M
an
d
V
SM
r
etr
iev
al
m
eth
o
d
s
to
r
etr
iev
e
th
e
k
to
p
-
r
an
k
ed
s
u
b
je
ct
h
ea
d
in
g
s
in
th
e
r
ec
o
m
m
e
n
d
atio
n
lis
t
.
I
n
o
u
r
ex
p
er
im
en
t,
we
s
et
k
v
alu
e
to
1
,
5
,
a
n
d
1
0
as we
f
o
c
u
s
o
n
th
e
s
u
b
ject
h
ea
d
in
g
s
e
x
tr
ac
ted
f
r
o
m
h
ig
h
ly
r
an
k
ed
d
o
cu
m
e
n
ts
.
F
o
un
d@
k
m
ea
s
u
r
e
is
p
r
o
p
o
s
ed
b
y
au
th
o
r
s
to
d
em
o
n
s
tr
ate
th
e
n
u
m
b
er
o
f
q
u
er
ies
th
at
th
e
r
elev
an
t
s
u
b
ject
h
ea
d
in
g
s
ca
n
b
e
f
o
u
n
d
in
th
e
r
ec
o
m
m
en
d
atio
n
lis
t
R
k
.
T
h
e
h
ig
h
e
r
th
e
Fo
u
n
d
@
k
s
co
r
e,
th
e
m
o
r
e
p
r
o
b
a
b
le
th
e
s
y
s
tem
to
r
et
r
iev
e
r
elev
a
n
t
s
u
b
ject
h
ea
d
in
g
s
f
o
r
q
u
er
ies.
T
h
is
m
ea
s
u
r
e
m
ay
s
u
g
g
est
th
e
ap
p
licab
ilit
y
o
f
t
h
e
s
y
s
tem
to
h
elp
in
th
e
s
u
b
ject
h
ea
d
in
g
ass
ig
n
m
en
t p
r
o
ce
s
s
.
@
=
#
ℎ
ℎ
ℎ
P
re
cisi
o
n@
k
m
ea
s
u
r
e
h
ig
h
lig
h
ts
h
o
w
ac
cu
r
ate
th
e
m
eth
o
d
t
o
p
u
t
m
an
y
r
elev
an
t
s
u
b
ject
h
ea
d
in
g
s
in
th
e
to
p
r
a
n
k
o
f
th
e
r
ec
o
m
m
en
d
atio
n
lis
t
R
k
.
T
h
is
is
a
s
tan
d
a
r
d
m
etr
ic
th
at
h
as b
ee
n
u
s
ed
in
m
an
y
in
f
o
r
m
atio
n
r
etr
iev
al
r
esear
ch
es in
p
r
ev
io
u
s
wo
r
k
.
@
=
#
ℎ
M
RR
m
ea
s
u
r
e
o
b
s
er
v
es
th
e
r
ec
ip
r
o
ca
l
r
an
k
o
f
th
e
p
o
s
itio
n
wh
er
e
th
e
f
ir
s
t
r
elev
an
t
s
u
b
ject
h
ea
d
in
g
is
f
o
u
n
d
.
T
h
e
h
ig
h
e
r
th
e
MRR
s
co
r
e
in
d
icate
s
th
at
th
e
r
elev
an
t
s
u
b
ject
h
ea
d
in
g
s
ca
n
b
e
f
o
u
n
d
in
th
e
ea
r
ly
p
o
s
itio
n
in
th
e
r
ec
o
m
m
e
n
d
ati
o
n
lis
t.
T
h
e
lo
wer
th
e
s
co
r
e
in
d
icate
s
th
at
th
e
r
elev
an
t
s
u
b
ject
h
ea
d
in
g
s
a
r
e
r
etr
iev
ed
in
th
e
lo
wer
r
an
k
in
t
h
e
r
ec
o
m
m
e
n
d
atio
n
lis
t,
wh
ich
ca
u
s
es
u
s
er
s
n
ee
d
to
s
cr
o
ll
d
o
wn
th
e
lis
t
f
u
r
th
e
r
in
o
r
d
er
to
f
i
n
d
th
e
r
elev
a
n
t
r
esu
lts
.
I
n
th
is
wo
r
k
,
we
lim
it
t
h
e
MRR
ca
lcu
latio
n
to
th
e
10
to
p
-
r
an
k
ed
s
u
b
ject
h
ea
d
in
g
s
as
th
is
is
th
e
h
ig
h
est
n
u
m
b
e
r
o
f
r
etr
iev
ed
d
o
c
u
m
e
n
ts
in
o
u
r
ex
p
er
im
e
n
t.
T
h
is
is
a
s
tan
d
ar
d
m
etr
ic
th
at
h
as b
ee
n
u
s
ed
in
m
an
y
in
f
o
r
m
atio
n
r
etr
iev
al
r
esear
ch
es i
n
p
r
ev
i
o
u
s
wo
r
k
.
=
1
ℎ
ℎ
5.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
5
.
1
.
E
f
f
ec
t
iv
eness
o
f
L
M
a
nd
V
S
M
m
et
ho
ds
in
g
ener
a
t
ing
s
u
bje
ct
hea
din
g
re
co
m
m
enda
t
i
o
n lis
t
T
h
e
F
o
u
n
d
@
k
s
co
r
es
f
o
r
L
M
an
d
VSM
m
eth
o
d
s
ar
e
s
h
o
wn
in
Fig
u
r
e
3
.
T
h
e
r
esu
lts
f
o
r
T
itle+Ab
s
tr
ac
t
ar
e
th
e
b
est
r
esu
lts
am
o
n
g
all
γ
v
al
u
e
b
etwe
e
n
0
-
1
with
t
h
e
s
tep
o
f
1
.
T
h
e
r
esu
lts
u
s
in
g
ea
ch
o
f
th
ese
γ
v
alu
es
ar
e
p
r
esen
ted
later
in
Sectio
n
5
.
2
.
Fo
r
b
o
th
m
eth
o
d
s
,
we
ca
n
s
ee
th
at
u
s
in
g
ab
s
tr
ac
t
as
a
s
ea
r
ch
u
n
it
p
r
o
d
u
ce
s
m
o
r
e
ef
f
ec
tiv
e
r
ec
o
m
m
en
d
atio
n
r
ath
er
t
h
an
u
s
in
g
titl
e
in
m
o
s
t
ca
s
es.
T
h
is
is
b
ec
au
s
e
tak
in
g
th
e
s
im
ilar
ities
b
etwe
en
ab
s
tr
ac
t
s
ar
e
m
o
r
e
ac
cu
r
ate
to
d
et
er
m
in
e
th
e
s
im
ilar
ity
b
etwe
en
two
m
etad
ata
d
o
cu
m
e
n
ts
,
as
o
p
p
o
s
ed
to
tak
i
n
g
th
e
s
im
ilar
ities
b
etwe
en
titl
es.
Sin
ce
ab
s
tr
ac
ts
ar
e
m
o
r
e
d
etail,
th
er
ef
o
r
e
two
m
etad
ata
with
h
ig
h
s
im
ilar
ity
in
ab
s
tr
ac
ts
ten
d
to
b
e
m
o
r
e
s
im
ilar
th
an
th
o
s
e
wi
th
h
ig
h
s
im
ilar
ity
in
ti
tle.
Usi
n
g
b
o
th
titl
e
an
d
ab
s
tr
ac
t
as
a
s
ea
r
ch
u
n
i
t
ca
n
s
lig
h
tly
im
p
r
o
v
e
th
e
r
esu
lts
o
f
u
s
in
g
ab
s
tr
ac
t
o
n
ly
.
T
h
is
in
d
icate
s
th
at
co
m
b
in
in
g
s
im
ilar
ities
b
etwe
en
titl
e
s
an
d
s
im
il
ar
ities
b
etwe
en
ab
s
tr
ac
ts
is
m
o
r
e
ac
cu
r
ate
to
f
in
d
r
elev
an
t
m
etad
ata.
W
h
en
u
s
in
g
titl
e
o
n
ly
as
th
e
s
ea
r
ch
u
n
it,
VSM
i
s
s
h
o
wn
to
b
e
litt
le
s
u
p
er
io
r
to
L
M
f
o
r
all
Fo
u
n
d
@
k
m
etr
ics
b
y
8
-
2
2
%.
T
h
is
ca
n
b
e
u
n
d
er
s
to
o
d
b
ec
a
u
s
e
L
M
u
s
es
p
r
o
b
ab
ilis
tic
co
n
ce
p
t
in
wh
ich
it
is
less
ac
cu
r
ate
wh
en
th
e
s
ize
o
f
d
ata
u
s
ed
to
p
r
o
d
u
ce
th
e
p
r
o
b
ab
ilit
y
s
co
r
e
is
s
m
all.
No
te
t
h
at
wh
en
th
e
s
ea
r
ch
u
n
it
is
titl
e
o
n
ly
,
th
en
th
e
lan
g
u
ag
e
m
o
d
el
u
s
ed
to
c
o
m
p
u
te
q
u
er
y
g
en
er
atio
n
p
r
o
b
ab
ilit
y
is
b
u
ilt
o
n
ly
u
s
in
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:
2502
-
4
7
5
2
Dete
r
min
in
g
s
u
b
ject
h
ea
d
in
g
s
o
f d
o
cu
men
ts
u
s
in
g
in
fo
r
ma
tio
n
r
etri
ev
a
l m
o
d
els
(
E
vi
Yu
lia
n
ti
)
1055
m
etad
ata
titl
es
in
th
e
co
llecti
o
n
.
W
h
en
th
e
s
ea
r
ch
u
n
it
is
ab
s
tr
ac
t
o
n
ly
o
r
titl
e+
ab
s
tr
ac
t,
b
o
th
L
M
an
d
VSM
m
eth
o
d
s
s
ee
m
to
b
e
co
m
p
ar
a
b
le.
T
h
is
r
esu
lt
d
em
o
n
s
tr
a
tes
th
at
L
M
an
d
VSM
ar
e
co
m
p
ar
ab
le
wh
en
u
s
in
g
lo
n
g
er
tex
t a
s
th
e
s
ea
r
ch
u
n
it,
b
u
t V
SM
is
p
r
ef
er
r
ed
wh
en
u
s
in
g
s
m
aller
tex
t.
Fig
u
r
e
3
.
T
h
e
n
u
m
b
er
o
f
q
u
er
i
es th
at
th
e
r
elev
an
t su
b
ject
h
ea
d
in
g
ca
n
b
e
f
o
u
n
d
in
th
e
r
ec
o
m
m
en
d
atio
n
lis
t
T
ab
le
4
d
escr
ib
e
th
e
ef
f
ec
tiv
e
n
ess
o
f
L
M
an
d
VSM
m
eth
o
d
s
b
ased
o
n
s
tan
d
a
r
d
r
et
r
iev
al
m
ea
s
u
r
es.
W
e
ca
n
s
ee
th
at
in
g
en
er
al,
th
is
r
esu
lt
is
co
n
s
i
s
ten
t
to
th
at
p
r
esen
ted
in
Fig
u
r
e
3
.
Fo
r
b
o
th
m
eth
o
d
s
,
u
s
in
g
ab
s
tr
ac
t
as
a
s
ea
r
ch
u
n
it
g
ain
s
h
ig
h
er
r
etr
iev
al
s
co
r
es
th
an
u
s
in
g
titl
e.
T
h
en
,
co
m
b
in
in
g
s
i
m
ilar
ities
b
etwe
en
titl
es
an
d
s
im
ilar
itie
s
b
etwe
en
ab
s
tr
ac
ts
ca
n
f
u
r
th
er
s
lig
h
tly
en
h
an
ce
th
e
r
esu
lts
.
An
im
p
o
r
tan
t
f
in
d
i
n
g
f
r
o
m
th
e
tab
le
is
th
at
b
ased
o
n
MRR
s
co
r
es,
o
n
av
er
a
g
e
th
e
f
ir
s
t
r
elev
an
t
s
u
b
ject
h
ea
d
in
g
is
f
o
u
n
d
a
t
th
e
3
rd
r
an
k
in
th
e
r
ec
o
m
m
en
d
atio
n
lis
t.
T
h
is
ca
n
b
en
ef
it
lib
r
ar
ian
s
as
th
ey
ca
n
f
in
d
r
elev
an
t
s
u
b
ject
in
th
e
to
p
r
an
k
o
f
th
e
r
ec
o
m
m
en
d
atio
n
lis
t,
s
o
it c
o
u
ld
s
av
e
th
eir
tim
e.
An
ad
d
itio
n
al
ex
p
e
r
im
en
t
wa
s
p
er
f
o
r
m
ed
b
y
p
e
r
f
o
r
m
in
g
s
to
p
p
in
g
a
n
d
s
tem
m
in
g
(
u
s
in
g
Sas
tr
aw
i
to
o
ls
)
in
th
e
p
r
e
-
p
r
o
ce
s
s
in
g
s
tep
to
u
n
d
er
s
tan
d
th
e
ef
f
ec
t
o
f
th
ese
s
tep
s
to
th
e
ef
f
ec
tiv
en
ess
o
f
s
u
b
ject
h
ea
d
in
g
s
r
ec
o
m
m
en
d
atio
n
r
es
u
lts
.
Ho
wev
er
,
we
f
o
u
n
d
th
at
t
h
e
ef
f
ec
t o
f
th
is
s
tep
s
to
th
e
r
e
s
u
lts
is
n
o
t c
lear
.
I
n
s
o
m
e
ca
s
es,
it
ca
n
s
lig
h
tly
in
c
r
ea
s
e
th
e
s
co
r
es
(
b
u
t
it
is
n
o
t
s
tatis
tically
s
ig
n
if
ican
t)
,
an
d
in
s
o
m
e
ca
s
es,
it
ca
n
d
ec
r
ea
s
e
th
e
s
co
r
e
.
B
ec
au
s
e
o
f
p
a
g
e
len
g
th
c
o
n
s
tr
ain
t,
th
e
r
esu
lts
o
f
t
h
is
ex
p
er
im
e
n
t
ar
e
n
o
t
s
h
o
wn
in
th
is
p
ap
er
.
T
ab
le
4
.
T
h
e
r
etr
iev
al
e
f
f
ec
tiv
en
ess
o
f
L
M
an
d
VSM
ac
co
r
d
in
g
to
s
tan
d
ar
d
r
etr
iev
al
m
ea
s
u
r
es
S
e
a
r
c
h
U
n
i
t
LM
V
S
M
M
R
R
P
@
1
P
@
5
P
@
1
0
M
R
R
P
@
1
P
@
5
P
@
1
0
Ti
t
l
e
0
.
3
2
4
0
.
2
2
0
0
.
1
6
4
0
.
1
5
4
0
.
3
5
9
0
.
2
4
7
0
.
2
0
0
0
.
1
6
7
A
b
st
r
a
c
t
0
.
4
0
7
0
.
3
2
3
0
.
2
1
5
0
.
1
8
4
0
.
3
9
1
0
.
3
1
0
0
.
2
1
9
0
.
1
8
4
Ti
t
l
e
+
A
b
s
t
r
a
c
t
0
.
4
1
0
0
.
3
3
0
0
.
2
2
9
0
.
1
9
9
0
.
4
0
9
0
.
3
2
0
0
.
2
3
2
0
.
1
9
7
5
.
2
.
E
f
f
ec
t
iv
eness
o
f
L
M
a
nd
V
S
M
m
et
ho
ds
o
n
diff
er
ent
γ
v
a
lues
T
o
ex
am
in
e
t
h
e
m
o
s
t
o
p
tim
al
γ
v
alu
e
f
o
r
L
M
a
n
d
VSM
m
eth
o
d
s
,
we
co
n
d
u
ct
an
ex
p
er
i
m
en
t
u
s
in
g
v
ar
y
in
g
γ
v
alu
es
b
etwe
en
0
an
d
1
with
a
s
tep
o
f
0
.
1
.
T
h
e
r
es
u
lts
ar
e
r
ep
o
r
ted
in
T
a
b
le
5
.
W
e
f
o
u
n
d
d
if
f
e
r
en
t
γ
v
alu
es
th
at
r
esu
lts
in
th
e
b
est
p
er
f
o
r
m
an
ce
f
o
r
L
M
an
d
VS
M
m
eth
o
d
s
.
T
h
e
m
o
s
t
o
p
tim
al
γ
v
alu
es
f
o
r
L
M
ar
e
b
etwe
en
0
.
5
an
d
0
.
9
,
w
h
ile
f
o
r
VSM
it
is
b
etwe
en
0
.
1
an
d
0
.
3
.
So
,
th
e
o
p
tim
al
γ
v
alu
e
f
o
r
L
M
ten
d
to
b
e
m
id
d
le
o
r
h
ig
h
,
a
n
d
it
te
n
d
s
t
o
b
e
l
o
w
f
o
r
VSM.
No
te
t
h
at
th
e
h
ig
h
e
r
th
e
γ
v
alu
es,
th
e
m
o
r
e
im
p
o
r
tan
ce
th
e
ab
s
tr
ac
t similar
ities
.
Oth
er
wi
s
e,
th
e
lo
wer
th
e
γ
v
alu
es,
th
e
m
o
r
e
i
m
p
o
r
tan
ce
th
e
titl
e
s
im
ilar
ities
.
T
h
e
o
p
tim
al
γ
v
alu
e
f
o
r
L
M
te
n
d
s
to
b
e
m
id
d
le
o
r
h
ig
h
b
ec
a
u
s
e
as
h
as
b
ee
n
s
h
o
wn
ea
r
lier
,
u
s
in
g
titl
e
s
im
ilar
ities
in
L
M
p
r
o
d
u
ce
s
m
u
ch
wo
r
s
e
r
esu
lts
th
an
u
s
in
g
ab
s
tr
ac
t
s
im
ilar
ities
.
T
h
is
is
b
ec
au
s
e
L
M
u
s
es
p
r
o
b
a
b
ilis
tic
co
n
ce
p
t
in
wh
ich
it
is
less
ac
cu
r
ate
wh
e
n
th
e
s
ize
o
f
d
ata
u
s
ed
to
p
r
o
d
u
ce
th
e
p
r
o
b
a
b
ilit
y
s
co
r
e
is
s
m
all.
T
h
er
ef
o
r
e
,
u
s
in
g
ab
s
tr
a
ct
s
im
ilar
ities
g
iv
es
m
u
ch
im
p
r
o
v
em
en
t
o
v
e
r
u
s
in
g
titl
e
s
im
ilar
ities
,
wh
ich
is
b
y
3
0
.
5
%
o
n
av
e
r
ag
e,
s
in
ce
th
e
q
u
er
y
g
e
n
er
atio
n
p
r
o
b
ab
ilit
y
is
b
u
ilt
u
s
in
g
b
ig
g
e
r
d
ata
.
T
h
en
f
o
r
VSM,
th
e
o
p
tim
al
γ
v
alu
e
f
o
r
L
M
ten
d
t
o
b
e
lo
w
b
ec
au
s
e
as h
as b
ee
n
s
h
o
wn
in
th
e
ea
r
lier
s
ec
tio
n
,
u
s
in
g
titl
e
s
im
ilar
itie
s
in
VSM
alr
ea
d
y
p
r
o
d
u
ce
d
f
air
ly
g
o
o
d
r
esu
lts
.
T
h
is
is
b
ec
au
s
e
VSM
ca
n
ca
p
tu
r
e
well
th
e
d
o
cu
m
en
t
s
im
ilar
ities
ev
en
b
y
u
s
in
g
s
m
all
tex
t
u
n
it,
s
in
ce
it
is
b
ased
o
n
ter
m
o
v
er
lap
b
etwe
en
two
d
o
cu
m
en
ts
th
at
is
co
m
p
u
ted
u
s
in
g
d
o
t
p
r
o
d
u
ct
o
f
th
e
d
o
cu
m
e
n
ts
’
v
ec
to
r
s
,
i.e
.
,
co
s
in
e
s
im
ilar
ity
.
T
h
er
e
f
o
r
e,
u
s
in
g
a
b
s
tr
ac
t
s
im
ilar
ities
o
n
ly
g
iv
e
s
lig
h
t im
p
r
o
v
em
e
n
t
o
v
e
r
u
s
in
g
titl
e
s
im
ilar
ities
,
wh
ich
is
b
y
1
4
.
6
%
o
n
av
er
ag
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci,
Vo
l.
23
,
No
.
2
,
Au
g
u
s
t 2
0
2
1
:
1
0
4
9
-
1
0
5
8
1056
T
ab
le
5
.
T
h
e
ef
f
ec
tiv
e
n
ess
o
f
L
M
an
d
VSM
o
n
d
if
f
er
e
n
t γ
v
alu
es
γ
F
o
u
n
d
@
1
F
o
u
n
d
@
5
F
o
u
n
d
@
1
0
M
R
R
P
@
1
P
@
5
P
@
1
0
LM
0
.
1
73
1
4
4
1
7
2
0
.
3
4
3
0
.
2
4
3
0
.
1
8
2
0
.
1
6
3
0
.
2
77
1
4
8
1
7
3
0
.
3
5
4
0
.
2
5
7
0
.
1
9
3
0
.
1
7
2
0
.
3
83
1
5
2
1
7
9
0
.
3
7
4
0
.
2
7
7
0
.
2
0
5
0
.
1
8
2
0
.
4
88
1
5
2
1
7
9
0
.
3
8
7
0
.
2
9
3
0
.
2
1
0
0
.
1
9
0
0
.
5
91
1
5
1
1
8
4
0
.
3
9
6
0
.
3
0
3
0
.
2
1
9
0
.
1
9
8
0
.
6
94
1
5
8
1
8
3
0
.
4
0
4
0
.
3
1
3
0
.
2
2
9
0
.
1
9
9
0
.
7
94
1
5
7
1
7
7
0
.
4
0
1
0
.
3
1
3
0
.
2
2
7
0
.
1
9
8
0
.
8
97
1
5
0
1
7
5
0
.
4
0
7
0
.
3
2
3
0
.
2
2
2
0
.
1
9
1
0
.
9
99
1
5
4
1
7
1
0
.
4
1
0
0
.
3
3
0
0
.
2
2
3
0
.
1
8
8
V
S
M
γ
F
o
u
n
d
@
1
F
o
u
n
d
@
5
F
o
u
n
d
@
1
0
M
R
R
P
@
1
P
@
5
P
@
1
0
0
.
1
88
1
6
4
1
8
4
0
.
3
9
7
0
.
2
9
3
0
.
2
2
7
0
.
1
9
6
0
.
2
92
1
6
1
1
8
0
0
.
4
0
4
0
.
3
0
7
0
.
2
3
1
0
.
1
9
5
0
.
3
96
1
5
9
1
8
0
0
.
4
0
9
0
.
3
2
0
0
.
2
3
2
0
.
1
9
7
0
.
4
93
1
5
5
1
7
7
0
.
3
9
8
0
.
3
1
0
0
.
2
2
5
0
.
1
9
6
0
.
5
94
1
5
3
1
7
5
0
.
3
9
7
0
.
3
1
3
0
.
2
2
3
0
.
1
9
4
0
.
6
93
1
5
5
1
7
1
0
.
3
9
5
0
.
3
1
0
0
.
2
2
7
0
.
1
9
2
0
.
7
93
1
5
2
1
7
3
0
.
3
9
6
0
.
3
1
0
0
.
2
2
3
0
.
1
8
8
0
.
8
93
1
5
3
1
6
9
0
.
3
9
3
0
.
3
1
0
0
.
2
2
1
0
.
1
8
8
0
.
9
93
1
5
0
1
7
0
0
.
3
9
2
0
.
3
1
0
0
.
2
1
8
0
.
1
8
5
5
.
3
.
Ana
ly
s
is
o
f
s
ub
j
ec
t
hea
din
g
s
o
v
er
la
ps
o
f
do
cum
ent
s
f
ro
m
diff
er
ent
f
a
cult
ies
I
t
is
in
ter
esti
n
g
to
ex
am
in
e
wh
eth
er
it
is
p
o
s
s
ib
le
th
at
a
q
u
er
y
h
as
r
elev
an
t
s
u
b
ject
h
e
ad
in
g
s
th
at
co
m
e
f
r
o
m
d
o
cu
m
en
ts
o
f
d
if
f
er
en
t
f
ac
u
lties
.
T
h
e
f
in
d
in
g
s
f
r
o
m
th
is
a
n
aly
s
is
m
ay
g
i
v
e
in
s
ig
h
ts
o
n
d
if
f
er
en
t
f
ac
u
lties
th
at
m
ay
s
h
ar
e
s
im
i
lar
to
p
ics
in
th
eir
s
tu
d
en
ts
’
t
h
esis
.
Fo
r
th
is
an
aly
s
is
,
we
e
x
am
in
e
th
e
s
u
b
ject
h
ea
d
in
g
s
r
ec
o
m
m
e
n
d
atio
n
g
en
er
ated
f
r
o
m
to
p
1
0
r
etr
iev
e
d
m
etad
ata
f
o
r
all
q
u
er
ies
.
W
e
an
aly
ze
th
e
f
ac
u
lty
o
f
q
u
er
ies
an
d
th
e
f
ac
u
lty
o
f
th
o
s
e
d
o
cu
m
en
ts
m
etad
ata
th
at
s
h
ar
e
s
im
ilar
s
u
b
ject
h
ea
d
in
g
s
.
As
th
er
e
ar
e
1
0
r
etr
iev
ed
m
etad
ata
a
n
d
3
0
0
q
u
er
ies,
th
er
ef
o
r
e
i
n
to
tal,
th
er
e
ar
e
3
0
0
0
f
ac
u
lty
p
ai
r
s
to
b
e
e
x
am
in
ed
.
W
e
f
o
u
n
d
th
er
e
ar
e
4
%
ca
s
es
wh
er
e
th
e
f
ac
u
lty
o
f
q
u
er
ies
is
d
if
f
er
en
t
to
th
e
f
ac
u
lty
o
f
to
p
1
0
r
etr
i
ev
ed
m
etad
ata,
b
u
t
th
ey
s
h
ar
e
s
im
ilar
s
u
b
ject
h
ea
d
in
g
s
.
Her
e
,
th
e
r
e
ar
e
1
3
3
ca
s
es (
u
s
in
g
L
M)
an
d
1
2
8
ca
s
es (
u
s
in
g
VSM)
.
B
o
t
h
L
M
a
n
d
V
S
M
a
g
r
e
e
o
n
th
e
t
o
p
f
i
v
e
p
a
i
r
s
o
f
f
a
c
u
lt
i
es
t
h
a
t
o
f
t
e
n
h
a
v
e
o
v
e
r
l
a
p
s
u
b
j
e
ct
h
e
a
d
i
n
g
s
as
s
h
o
w
n
i
n
T
a
b
l
e
6
.
T
h
is
f
i
n
d
i
n
g
c
a
n
t
h
e
r
e
f
o
r
e
e
n
c
o
u
r
a
g
e
t
h
e
p
o
s
s
i
b
il
i
t
y
o
f
r
e
s
e
a
r
c
h
c
o
l
l
a
b
o
r
a
t
i
o
n
b
e
t
w
e
e
n
t
h
es
e
f
a
c
u
l
t
i
es
.
T
h
e
p
a
i
r
o
f
f
a
c
u
l
t
i
es
t
h
a
t
i
s
m
o
s
t
o
f
t
e
n
t
o
h
a
v
e
s
i
m
il
a
r
s
u
b
j
e
c
t
h
e
a
d
i
n
g
s
a
r
e
F
ac
u
l
ty
o
f
E
c
o
n
o
m
y
a
n
d
B
u
s
i
n
e
s
s
a
n
d
t
h
e
F
a
c
u
l
t
y
o
f
S
o
c
i
a
l
a
n
d
P
o
l
it
i
c
al
S
c
i
e
n
c
e
.
W
e
a
n
a
l
y
z
e
t
h
at
a
lt
h
o
u
g
h
t
w
o
m
e
t
ad
a
t
a
f
r
o
m
t
h
es
e
tw
o
f
a
c
u
l
t
i
es
c
o
n
t
ai
n
s
i
m
i
l
a
r
t
o
p
i
cs,
b
u
t
t
h
e
y
d
is
c
u
s
s
t
h
e
m
o
n
d
i
f
f
e
r
e
n
t
p
e
r
s
p
e
ct
i
v
es
.
Fi
g
u
r
e
4
s
h
o
w
s
a
n
e
x
a
m
p
l
e
o
f
t
w
o
m
e
ta
d
a
t
a
f
r
o
m
F
a
c
u
lt
y
o
f
E
c
o
n
o
m
y
a
n
d
F
a
c
u
l
t
y
o
f
s
o
c
ia
l
a
n
d
p
o
l
it
i
c
al
s
c
i
e
n
c
e
t
h
a
t
s
h
a
r
e
s
i
m
i
l
a
r
s
u
b
j
ec
t
h
e
a
d
i
n
g
,
i
.
e
.
,
m
u
t
u
a
l
f
u
n
d
s
.
W
e
c
a
n
s
e
e
f
r
o
m
t
h
e
t
i
tl
es
o
f
t
h
es
e
m
e
ta
d
a
t
a
t
h
a
t
al
t
h
o
u
g
h
b
o
t
h
m
e
t
a
d
a
t
a
c
o
n
t
a
i
n
s
i
n
f
o
r
m
a
t
i
o
n
a
b
o
u
t
m
a
r
k
e
t
t
i
m
i
n
g
,
s
t
o
c
k
s
e
l
e
ct
i
o
n
,
a
n
d
e
q
u
it
y
f
u
n
d
s
,
b
u
t
t
h
e
f
i
r
s
t
m
e
t
a
d
at
a
d
i
s
c
u
s
s
t
h
e
m
o
n
e
c
o
n
o
m
y
p
e
r
s
p
e
c
t
i
v
es
,
w
h
i
le
th
e
s
e
c
o
n
d
m
e
t
a
d
a
ta
d
i
s
c
u
s
s
t
h
em
o
n
s
o
c
i
a
l
p
e
r
s
p
e
c
t
i
v
es
.
T
ab
le
6
.
Fiv
e
p
ai
r
s
o
f
d
if
f
er
en
t
f
ac
u
lties
th
at
ar
e
m
o
s
t o
f
te
n
t
o
h
av
e
s
im
ilar
s
u
b
ject
h
ea
d
in
g
s
in
o
u
r
r
esu
lts
LM
V
S
M
F
a
c
u
l
t
y
P
a
i
r
s
To
t
a
l
F
a
c
u
l
t
y
P
a
i
r
s
To
t
a
l
Ec
o
-
S
o
c
29
Ec
o
-
S
o
c
31
N
u
r
s
-
P
u
b
H
28
N
u
r
s
-
P
u
b
H
27
M
e
d
-
P
u
b
H
13
M
e
d
-
P
u
b
H
10
M
e
d
-
N
u
r
s
11
M
e
d
-
N
u
r
s
10
N
u
r
s
-
P
sy
7
N
u
r
s
-
P
sy
6
Fig
u
r
e
4
.
E
x
am
p
le
o
f
two
m
et
ad
ata
o
f
d
i
f
f
er
en
t
f
ac
u
lties
th
at
s
h
ar
e
s
im
ilar
s
u
b
ject
h
ea
d
in
g
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:
2502
-
4
7
5
2
Dete
r
min
in
g
s
u
b
ject
h
ea
d
in
g
s
o
f d
o
cu
men
ts
u
s
in
g
in
fo
r
ma
tio
n
r
etri
ev
a
l m
o
d
els
(
E
vi
Yu
lia
n
ti
)
1057
I
n
f
u
tu
r
e,
u
s
er
s
tu
d
ies
co
u
ld
b
e
p
er
f
o
r
m
ed
t
o
test
wh
eth
er
lib
r
ar
ian
s
ca
n
f
in
d
r
elev
a
n
t
s
u
b
ject
h
ea
d
in
g
s
f
r
o
m
th
e
r
ec
o
m
m
en
d
atio
n
lis
t
o
r
n
o
t.
W
e
ca
n
also
ask
th
eir
p
r
ef
er
en
ce
o
n
th
e
s
u
b
ject
h
ea
d
in
g
s
r
ec
o
m
m
en
d
atio
n
g
en
er
ated
u
s
in
g
L
M
an
d
VSM
m
eth
o
d
s
.
Dif
f
er
en
t
r
etr
iev
al
m
eth
o
d
s
,
s
u
ch
as:
p
r
o
b
ab
ilis
tic
B
M2
5
[
3
0
]
,
[
3
1
]
an
d
s
eq
u
en
ti
al
d
ep
en
d
en
ce
m
o
d
el
(
SDM)
[
3
2
]
,
[
3
3
]
ca
n
b
e
co
n
s
id
er
e
d
f
o
r
f
u
r
t
h
er
s
tu
d
y
.
T
h
e
p
o
s
s
ib
ilit
y
o
f
co
m
b
in
in
g
s
em
an
tic
in
f
o
r
m
atio
n
[
2
5
]
,
[
3
4
]
,
[
3
5
]
o
r
s
o
cial
m
ed
ia
[
3
6
]
,
[
3
7
]
in
to
th
e
r
etr
iev
al
m
o
d
el
to
a
d
d
r
ess
lex
ical
m
is
m
atch
p
r
o
b
lem
is
a
g
r
ea
t c
h
allen
g
e
to
b
e
e
x
p
lo
r
e
d
.
6.
CO
NCLU
SI
O
N
W
e
co
n
d
u
ct
a
n
o
v
el
s
tu
d
y
o
n
th
e
u
s
e
o
f
d
o
cu
m
e
n
t
r
etr
iev
al
m
eth
o
d
s
,
i.e
.
,
q
u
er
y
lik
elih
o
o
d
lan
g
u
ag
e
m
o
d
els
(
L
M)
an
d
v
ec
to
r
s
p
ac
e
m
o
d
els
(
VSM)
,
to
g
e
n
er
ate
s
u
b
ject
h
ea
d
in
g
s
r
ec
o
m
m
e
n
d
a
tio
n
lis
t
f
o
r
a
g
iv
e
n
d
o
cu
m
e
n
t.
T
h
e
aim
o
f
t
h
is
s
t
u
d
y
is
to
h
elp
lib
r
a
r
ian
s
to
ass
i
g
n
ap
p
r
o
p
r
iate
s
u
b
ject
h
ea
d
i
n
g
s
to
a
d
o
cu
m
en
t
m
o
r
e
ef
f
ec
tiv
el
y
an
d
ef
f
icien
t
ly
.
Ou
r
r
esu
lts
s
h
o
ws
th
at
in
f
o
r
m
atio
n
r
etr
ie
v
al
m
eth
o
d
s
ca
n
h
elp
th
e
s
u
b
ject
h
ea
d
in
g
ass
ig
n
m
en
t
p
r
o
ce
s
s
.
T
h
er
e
ar
e
a
h
ig
h
n
u
m
b
er
o
f
q
u
er
ies
(
u
p
to
6
1
%)
th
at
h
a
v
e
r
elev
an
t
s
u
b
ject
h
ea
d
in
g
s
in
th
e
ten
to
p
-
r
an
k
e
d
r
ec
o
m
m
e
n
d
atio
n
lis
t,
wh
ich
s
u
g
g
est
th
e
ap
p
licab
ilit
y
o
f
th
e
s
y
s
tem
.
Nex
t,
o
n
av
er
ag
e,
th
e
f
i
r
s
t
r
elev
an
t
s
u
b
ject
h
ea
d
in
g
is
f
o
u
n
d
in
th
e
3
rd
r
an
k
in
th
e
r
ec
o
m
m
e
n
d
at
io
n
lis
t.
T
h
is
is
an
in
ter
esti
n
g
f
in
d
in
g
th
at
in
d
icate
s
th
at
lib
r
ar
ian
s
ca
n
f
i
n
d
r
elev
an
t
s
u
b
ject
h
ea
d
in
g
s
q
u
ick
ly
as
th
ey
d
o
n
o
t n
ee
d
to
s
cr
o
ll
d
o
wn
to
o
f
ar
i
n
th
e
r
ec
o
m
m
en
d
atio
n
lis
t.
B
o
th
L
M
an
d
VSM
h
as
co
m
p
ar
a
b
le
ef
f
ec
tiv
en
ess
wh
en
u
s
in
g
lo
n
g
er
tex
t
as
th
e
s
ea
r
c
h
u
n
it.
Ho
wev
er
,
VSM
is
s
u
p
er
io
r
to
L
M
b
y
8
-
2
1
%
wh
en
u
s
in
g
s
m
aller
tex
t
as
th
e
s
ea
r
ch
u
n
it.
Ou
r
an
al
y
s
is
t
h
en
r
ev
ea
ls
th
r
ee
p
air
s
o
f
f
ac
u
lties
with
th
e
h
ig
h
est
n
u
m
b
er
o
f
s
u
b
ject
h
ea
d
in
g
s
o
v
er
lap
a
r
e:
i
)
Facu
lty
o
f
E
c
o
n
o
m
y
an
d
B
u
s
in
ess
-
Facu
lty
o
f
So
cial
an
d
Po
liti
ca
l
Scien
ce
s
,
ii
)
Facu
lty
o
f
Nu
r
s
in
g
-
Facu
lty
o
f
Pu
b
lic
He
alth
,
an
d
iii
)
Facu
lty
o
f
Me
d
icin
e
-
Facu
lty
o
f
Pu
b
lic
Hea
lth
.
T
h
is
f
in
d
in
g
m
ay
s
u
g
g
est th
e
p
o
ten
tial r
esea
r
ch
co
llab
o
r
atio
n
b
etwe
en
th
ese
f
a
cu
lties
.
ACK
NO
WL
E
DG
E
M
E
NT
S
T
h
is
r
esear
ch
is
s
u
p
p
o
r
ted
b
y
th
e
PUTI
Q3
g
r
an
t
n
u
m
b
e
r
NKB
-
4
3
7
9
/UN2
.
R
ST/HKP.
0
5
.
0
0
/
2
0
2
0
f
r
o
m
Un
iv
e
r
s
itas
I
n
d
o
n
esia.
RE
F
E
R
E
NC
E
S
[1
]
G
.
L.
Ho
ffm
a
n
,
Or
g
a
n
izi
n
g
L
ib
r
a
ry
Co
ll
e
c
ti
o
n
s:
T
h
e
o
ry
a
n
d
Pra
c
ti
c
e
,
Ro
wm
a
n
&
Li
tt
lefie
ld
,
2
0
1
9
.
[2
]
L.
Ho
o
v
e
r,
“
A
b
e
g
in
n
e
rs’
g
u
id
e
f
o
r
su
b
jec
t
a
n
a
ly
sis
o
f
th
e
se
s
a
n
d
d
isse
rtatio
n
s
i
n
t
h
e
h
a
rd
sc
ien
c
e
s,
”
Ca
ta
l
o
g
i
n
g
&
c
la
ss
if
ica
ti
o
n
q
u
a
rte
rly
,
v
o
l.
4
1
,
n
o
.
1
,
p
p
.
1
3
3
–
1
6
1
,
2
0
0
5
,
d
o
i:
1
0
.
1
3
0
0
/J1
0
4
v
4
1
n
0
1
_
0
7
.
[3
]
A.
J.
S
p
e
n
c
e
r
a
n
d
J.
D.
El
d
re
d
g
e
,
“
Ro
les
fo
r
li
b
ra
rian
s
in
sy
ste
m
a
ti
c
re
v
iew
s:
a
sc
o
p
i
n
g
re
v
iew
,
”
J
o
u
rn
a
l
o
f
th
e
M
e
d
ica
l
L
ib
r
a
ry
Asso
c
ia
t
io
n
:
J
M
L
A
,
v
o
l.
1
0
6
,
n
o
.
1
,
p
p
.
4
6
-
5
6
,
2
0
1
8
,
d
o
i:
1
0
.
5
1
9
5
/j
m
la.
2
0
1
8
.
8
2
.
[4
]
R.
S
a
p
o
n
-
Wh
it
e
,
“
S
u
b
jec
t
a
n
a
ly
sis
train
i
n
g
fo
r
c
a
talo
g
in
g
p
a
ra
p
ro
fe
s
sio
n
a
ls:
A
m
o
d
e
l
f
o
r
o
n
g
o
i
n
g
lea
rn
in
g
a
n
d
su
p
p
o
rt,
”
T
e
c
h
n
ica
l
S
e
rv
ice
s Qu
a
rte
rly
,
v
o
l.
2
6
,
n
o
.
3
,
p
p
.
1
8
3
–
1
9
3
,
2
0
0
9
,
d
o
i:
1
0
.
1
0
8
0
/0
7
3
1
7
1
3
0
8
0
2
5
2
0
0
1
3
.
[5
]
C.
-
A.
Ju
li
e
n
,
B.
As
a
d
i,
J.
D.
Di
n
n
e
e
n
,
a
n
d
F
.
S
h
u
,
“
Li
b
ra
r
y
o
f
c
o
n
g
re
ss
s
u
b
jec
t
h
e
a
d
i
n
g
(LCS
H
)
b
r
o
ws
in
g
a
n
d
n
a
tu
ra
l
lan
g
u
a
g
e
se
a
rc
h
in
g
,
”
Pr
o
c
e
e
d
in
g
s
o
f
t
h
e
Asso
c
ia
t
io
n
fo
r
In
fo
rm
a
ti
o
n
S
c
ie
n
c
e
a
n
d
T
e
c
h
n
o
lo
g
y
,
v
o
l.
5
3
,
n
o
.
1
,
p
p
.
1
–
4
,
2
0
1
6
,
d
o
i:
1
0
.
1
0
0
2
/p
ra
2
.
2
0
1
6
.
1
4
5
0
5
3
0
1
1
1
6
.
[6
]
C.
D.
Ba
tt
y
,
An
in
tro
d
u
c
ti
o
n
t
o
t
h
e
De
we
y
d
e
c
ima
l
c
la
ss
if
ica
ti
o
n
,
A
sia
P
u
b
l
ish
in
g
H
o
u
se
(1
9
6
6
),
2
0
1
7
.
[7
]
A.
M
.
F
e
rris,
“
Birt
h
o
f
a
S
u
b
jec
t
He
a
d
in
g
,
”
L
ib
ra
ry
Res
o
u
rc
e
s &
T
e
c
h
n
ica
l
S
e
rv
ice
s
,
v
o
l.
6
2
,
n
o
.
1
,
p
.
1
6
,
2
0
1
8
.
[8
]
J.
Lee
-
S
m
e
lt
z
e
r
a
n
d
D.
Ha
c
k
lem
a
n
,
“
Ac
c
e
ss
to
OSU
Th
e
se
s a
n
d
Di
ss
e
rtatio
n
s in
Ke
rr
Li
b
ra
ry
:
H
o
w
Th
e
y
Are
Us
e
d
o
r
Are
Th
e
y
?
”
T
e
c
h
n
ica
l
S
e
rv
ice
s Qu
a
rte
rly
,
v
o
l.
1
2
,
n
o
.
4
,
p
p
.
2
5
–
4
3
,
1
9
9
5
,
d
o
i:
1
0
.
1
3
0
0
/J1
2
4
v
1
2
n
0
4
_
0
3
.
[9
]
R.
E.
S
a
p
o
n
-
W
h
it
e
a
n
d
M
.
Ha
n
s
b
ro
u
g
h
,
“
T
h
e
imp
a
c
t
o
f
su
b
jec
t
h
e
a
d
in
g
a
ss
ig
n
m
e
n
t
o
n
c
ircu
latio
n
o
f
d
isse
rtatio
n
s
a
t
Virg
in
ia
Tec
h
,
”
L
i
b
ra
ry
re
so
u
rc
e
s
&
tec
h
n
ica
l
se
rv
ice
s
,
v
o
l.
4
2
,
n
o
.
4
,
p
p
.
2
8
2
–
2
9
1
,
1
9
9
8
,
d
o
i:
1
0
.
5
8
6
0
/l
rts.
4
2
n
4
.
2
8
2
.
[1
0
]
M
.
Ad
ler,
J.
T.
Hu
b
e
r,
a
n
d
A.
T.
Nix
,
“
S
ti
g
m
a
ti
z
i
n
g
d
isa
b
il
it
y
:
Li
b
ra
ry
c
las
sifica
ti
o
n
s
a
n
d
t
h
e
m
a
rk
in
g
a
n
d
m
a
rg
in
a
li
z
a
ti
o
n
o
f
b
o
o
k
s
a
b
o
u
t
p
e
o
p
le
with
d
isa
b
il
it
ies
,
”
T
h
e
L
ib
r
a
r
y
Qu
a
rte
rly
,
v
o
l
.
8
7
,
n
o
.
2
,
p
p
.
1
1
7
–
1
3
5
,
2
0
1
7
,
d
o
i:
1
0
.
1
0
8
6
/6
9
0
7
3
4
.
[1
1
]
Y.
-
L.
Lu
a
n
d
D.
W
.
Bian
c
h
i,
“
T
re
n
d
s
in
p
re
n
a
tal
d
iag
n
o
sis:
An
a
n
a
ly
sis
o
f
4
0
y
e
a
rs
o
f
M
e
d
ica
l
S
u
b
jec
t
He
a
d
in
g
(M
e
S
H) t
e
rm
s in
p
u
b
li
c
a
ti
o
n
s
,
”
P
re
n
a
ta
l
Dia
g
n
o
sis
,
v
o
l.
4
0
,
n
o
.
1
3
,
p
p
.
1
6
3
6
–
1
6
4
0
,
2
0
2
0
,
d
o
i:
1
0
.
1
0
0
2
/
p
d
.
5
8
7
1
.
[1
2
]
Y.
-
J.
Kim
,
“
C
o
m
p
a
riso
n
o
f
a
u
th
o
r
k
e
y
wo
r
d
s
a
n
d
M
e
d
ica
l
S
u
b
jec
t
He
a
d
in
g
term
s
i
n
th
e
Jo
u
rn
a
l
o
f
Ko
re
a
n
S
o
c
iety
o
f
De
n
tal
Hy
g
ien
e
fro
m
2
0
0
1
to
2
0
1
5
,
”
J
o
u
rn
a
l
o
f
K
o
re
a
n
so
c
iety
o
f
De
n
ta
l
Hy
g
ien
e
,
v
o
l.
1
8
,
n
o
.
6
,
p
p
.
1
0
4
7
-
1
0
5
5
,
2
0
1
8
,
d
o
i:
1
0
.
1
3
0
6
5
/j
k
sd
h
.
2
0
1
8
0
0
9
0
.
[1
3
]
F
.
S
h
u
,
J.
D.
Di
n
n
e
e
n
,
B.
As
a
d
i,
a
n
d
C.
-
A.
Ju
li
e
n
,
“
M
a
p
p
in
g
sc
ie
n
c
e
u
sin
g
li
b
ra
ry
o
f
c
o
n
g
re
ss
su
b
jec
t
h
e
a
d
in
g
s,”
J
o
u
rn
a
l
o
f
In
fo
rm
e
trics
,
v
o
l.
1
1
,
n
o
.
4
,
p
p
.
1
0
8
0
–
1
0
9
4
,
2
0
1
7
,
d
o
i:
0
.
1
0
1
6
/j
.
jo
i
.
2
0
1
7
.
0
8
.
0
0
8
.
[1
4
]
F
.
S
h
u
,
J.
Qi
u
,
a
n
d
V.
Lari
v
ière
,
“
M
a
p
p
in
g
t
h
e
Li
fe
S
c
ien
c
e
u
sin
g
M
e
d
ica
l
S
u
b
jec
t
He
a
d
in
g
s
(M
e
S
H),”
in
1
7
t
h
In
ter
n
a
t
io
n
a
l
C
o
n
fer
e
n
c
e
o
n
S
c
ien
to
me
trics
&
In
f
o
rm
e
trics
,
2
0
1
9
,
p
p
.
1
9
2
7
–
1
9
3
2
.
[1
5
]
J.
D.
Din
n
e
e
n
,
B.
As
a
d
i,
I.
F
ri
ss
e
n
,
F
.
S
h
u
,
a
n
d
C.
-
A.
J
u
li
e
n
,
“
Im
p
ro
v
in
g
e
x
p
lo
ra
ti
o
n
o
f
t
o
p
ic
h
iera
rc
h
ies
:
Co
m
p
a
ra
ti
v
e
tes
ti
n
g
o
f
sim
p
li
fie
d
li
b
ra
ry
o
f
c
o
n
g
re
ss
su
b
jec
t
h
e
a
d
in
g
str
u
c
tu
re
s,”
in
Pro
c
e
e
d
in
g
s
o
f
th
e
2
0
1
8
Co
n
fer
e
n
c
e
o
n
Hu
ma
n
I
n
fo
rm
a
ti
o
n
In
ter
a
c
ti
o
n
&
Retrie
v
a
l
,
2
0
1
8
,
p
p
.
1
0
2
–
1
0
9
,
d
o
i:
1
0
.
1
1
4
5
/3
1
7
6
3
4
9
.
3
1
7
6
3
8
5
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci,
Vo
l.
23
,
No
.
2
,
Au
g
u
s
t 2
0
2
1
:
1
0
4
9
-
1
0
5
8
1058
[1
6
]
C.
-
A.
Ju
li
e
n
,
P
.
Ti
ril
l
y
,
J
.
D.
Din
n
e
e
n
,
a
n
d
C.
G
u
a
sta
v
in
o
,
“
Re
d
u
c
in
g
su
b
jec
t
tree
b
ro
ws
in
g
c
o
m
p
lex
it
y
,
”
J
o
u
rn
a
l
o
f
th
e
Ame
ric
a
n
S
o
c
iety
fo
r
In
f
o
rm
a
ti
o
n
S
c
ien
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
v
o
l.
6
4
,
n
o
.
1
1
,
p
p
.
2
2
0
1
–
2
2
2
3
,
2
0
1
3
,
d
o
i:
1
0
.
1
0
0
2
/as
i.
2
2
9
1
5
.
[1
7
]
R.
Ka
v
u
lu
r
u
a
n
d
Z.
He
,
“
Un
s
u
p
e
rv
ise
d
m
e
d
ica
l
su
b
jec
t
h
e
a
d
in
g
a
ss
ig
n
m
e
n
t
u
sin
g
o
u
tp
u
t
lab
e
l
c
o
-
o
c
c
u
rre
n
c
e
sta
ti
stics
a
n
d
se
m
a
n
ti
c
p
re
d
ica
t
io
n
s,”
in
In
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
A
p
p
li
c
a
ti
o
n
o
f
Na
t
u
ra
l
L
a
n
g
u
a
g
e
t
o
In
fo
rm
a
t
io
n
S
y
ste
ms
,
2
0
1
3
,
p
p
.
1
7
6
–
1
8
8
,
d
o
i:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
6
4
2
-
3
8
8
2
4
-
8
_
1
5
.
[1
8
]
R.
Ka
v
u
l
u
ru
a
n
d
Y.
Lu
,
“
Lev
e
ra
g
in
g
o
u
tp
u
t
term
c
o
-
o
c
c
u
rre
n
c
e
f
re
q
u
e
n
c
ies
a
n
d
late
n
t
a
ss
o
c
iatio
n
s
in
p
re
d
ictin
g
m
e
d
ica
l
su
b
jec
t
h
e
a
d
i
n
g
s,”
Da
ta
&
Kn
o
wled
g
e
En
g
in
e
e
rin
g
,
v
o
l
.
9
4
,
p
p
.
1
8
9
-
2
0
1
,
2
0
1
4
,
d
o
i:
1
0
.
1
0
1
6
/j
.
d
a
tak
.
2
0
1
4
.
0
9
.
0
0
2
.
[1
9
]
K.
G
o
lu
b
,
J.
Ha
g
e
l
b
ä
c
k
,
a
n
d
A.
Ard
ö
,
“
Au
to
m
a
ti
c
Clas
sifica
ti
o
n
o
f
S
we
d
ish
M
e
ta
d
a
ta
Us
in
g
De
we
y
De
c
im
a
l
Clas
sifica
ti
o
n
:
A
C
o
m
p
a
riso
n
o
f
Ap
p
ro
a
c
h
e
s,”
J
o
u
rn
a
l
o
f
D
a
ta
a
n
d
In
fo
rm
a
t
io
n
S
c
ien
c
e
,
v
o
l.
5
,
n
o
.
1
,
p
p
.
1
8
–
3
8
,
2
0
2
0
,
d
o
i:
1
0
.
2
4
7
8
/j
d
is
-
2
0
2
0
-
0
0
0
3
.
[2
0
]
E.
S
v
e
n
o
n
iu
s
a
n
d
D
.
M
c
G
a
rry
,
“
Ob
jec
ti
v
it
y
i
n
e
v
a
lu
a
ti
n
g
su
b
jec
t
h
e
a
d
in
g
a
ss
ig
n
m
e
n
t
,
”
Ca
t
a
lo
g
i
n
g
&
c
la
ss
if
ica
ti
o
n
q
u
a
rte
rly
,
v
o
l.
1
6
,
n
o
.
2
,
p
p
.
5
–
4
0
,
1
9
9
3
,
d
o
i:
1
0
.
1
3
0
0
/J1
0
4
v
1
6
n
0
2
_
0
2
.
[2
1
]
K.
Lein
in
g
e
r,
“
In
terin
d
e
x
e
r
c
o
n
si
ste
n
c
y
i
n
P
sy
c
IN
F
O,”
J
o
u
r
n
a
l
o
f
L
ib
ra
ri
a
n
s
h
ip
a
n
d
I
n
fo
rm
a
ti
o
n
S
c
ien
c
e
,
v
o
l.
3
2
,
n
o
.
1
,
p
p
.
4
–
8
,
2
0
0
0
,
d
o
i:
1
0
.
1
1
7
7
/0
9
6
1
0
0
0
6
0
0
0
3
2
0
0
1
0
2
.
[2
2
]
Y.
S
u
,
Y.
Hu
a
n
g
,
a
n
d
C.
J.
K
u
o
,
“
Eff
icie
n
t
Tex
t
Clas
sifica
ti
o
n
Us
in
g
Tree
-
stru
c
tu
re
d
M
u
lt
i
-
l
in
e
a
r
P
rin
c
i
p
a
l
Co
m
p
o
n
e
n
t
A
n
a
ly
sis,”
2
0
1
8
2
4
t
h
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
P
a
tt
e
rn
Rec
o
g
n
it
io
n
(IC
PR
)
,
2
0
1
8
,
p
p
.
5
8
5
-
5
9
0
,
d
o
i:
1
0
.
1
1
0
9
/ICP
R.
2
0
1
8
.
8
5
4
5
8
3
2
.
[2
3
]
D.
A.
M
e
e
d
e
n
iy
a
a
n
d
A.
S
.
P
e
re
r
a
,
“
Ev
a
lu
a
ti
o
n
o
f
P
a
rti
ti
o
n
-
Ba
se
d
Tex
t
Clu
ste
rin
g
Tec
h
n
i
q
u
e
s
to
C
a
teg
o
rize
In
d
ic
Lan
g
u
a
g
e
Do
c
u
m
e
n
ts,”
2
0
0
9
I
EE
E
I
n
t
e
rn
a
t
io
n
a
l
Ad
v
a
n
c
e
C
o
mp
u
t
in
g
C
o
n
fer
e
n
c
e
,
2
0
0
9
,
p
p
.
1
4
9
7
-
1
5
0
0
,
d
o
i:
1
0
.
1
1
0
9
/IADCC.2
0
0
9
.
4
8
0
9
2
3
9
.
[2
4
]
J.
M
.
P
o
n
te
a
n
d
W.
B.
Cro
ft
,
“
A
lan
g
u
a
g
e
m
o
d
e
li
n
g
a
p
p
r
o
a
c
h
t
o
i
n
fo
rm
a
ti
o
n
re
tri
e
v
a
l,
”
in
S
IGIR
'
9
8
:
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
2
1
st
a
n
n
u
a
l
in
ter
n
a
ti
o
n
a
l
ACM
S
IGIR
c
o
n
fe
re
n
c
e
o
n
Res
e
a
rc
h
a
n
d
d
e
v
e
lo
p
me
n
t
i
n
in
f
o
rm
a
ti
o
n
re
triev
a
l
,
1
9
9
8
,
p
p
.
2
7
5
–
2
8
1
,
d
o
i:
1
0
.
1
1
4
5
/
2
9
0
9
4
1
.
2
9
1
0
0
8
.
[2
5
]
E.
Yu
li
a
n
ti
,
R.
-
C
.
C
h
e
n
,
F
.
S
c
h
o
l
e
r,
a
n
d
M
.
S
a
n
d
e
rso
n
,
“
Us
in
g
se
m
a
n
ti
c
a
n
d
c
o
n
tex
t
fe
a
tu
re
s
fo
r
a
n
sw
e
r
su
m
m
a
ry
e
x
trac
ti
o
n
,
”
in
Pro
c
e
e
d
in
g
s
o
f
th
e
2
1
s
t
Au
str
a
la
si
a
n
Do
c
u
m
e
n
t
Co
m
p
u
ti
n
g
S
y
mp
o
siu
m
,
2
0
1
6
,
p
p
.
8
1
–
8
4
,
d
o
i:
1
0
.
1
1
4
5
/3
0
1
5
0
2
2
.
3
0
1
5
0
3
1
.
[2
6
]
G
.
S
a
lt
o
n
,
A.
W
o
n
g
,
a
n
d
C.
-
S
.
Ya
n
g
,
“
A
v
e
c
to
r
sp
a
c
e
m
o
d
e
l
f
o
r
a
u
t
o
m
a
ti
c
in
d
e
x
in
g
,
”
Co
mm
u
n
ica
ti
o
n
s o
f
t
h
e
ACM
,
v
o
l.
1
8
,
n
o
.
1
1
,
p
p
.
6
1
3
–
6
2
0
,
1
9
7
5
,
d
o
i:
1
0
.
1
1
4
5
/3
6
1
2
1
9
.
3
6
1
2
2
0
.
[2
7
]
R.
C.
Be
lwa
l,
S
.
Ra
i,
a
n
d
A.
G
u
p
ta,
“
Tex
t
su
m
m
a
riza
ti
o
n
u
sin
g
to
p
ic
-
b
a
se
d
v
e
c
to
r
sp
a
c
e
m
o
d
e
l
a
n
d
se
m
a
n
ti
c
m
e
a
su
re
,
”
In
fo
rm
a
ti
o
n
Pr
o
c
e
ss
in
g
&
M
a
n
a
g
e
me
n
t
,
v
o
l.
5
8
,
n
o
.
3
,
2
0
2
1
,
d
o
i:
1
0
.
1
0
1
6
/j
.
ip
m
.
2
0
2
1
.
1
0
2
5
3
6
.
[2
8
]
D.
M
e
tzle
r,
T.
S
tr
o
h
m
a
n
,
H.
T
u
r
tl
e
,
a
n
d
W.
B.
Cr
o
ft,
“
I
n
d
ri
a
t
T
REC
2
0
0
4
:
Tera
b
y
te
trac
k
,
”
M
a
ss
a
c
h
u
se
tt
s
Un
iv
Am
h
e
rst
Ce
n
ter F
o
r
I
n
telli
g
e
n
t
I
n
fo
rm
a
ti
o
n
Re
tri
e
v
a
l,
2
0
0
4
.
[2
9
]
Y.
Wan
g
a
n
d
H.
F
a
n
g
,
“
Co
m
b
in
i
n
g
Term
-
b
a
se
d
a
n
d
C
o
n
c
e
p
t
-
b
a
se
d
Re
p
re
se
n
tatio
n
f
o
r
Cli
n
ica
l
Re
tri
e
v
a
l,
”
T
RE
C
,
2
0
1
7
.
[3
0
]
S
.
Ro
b
e
rtso
n
,
H.
Zara
g
o
z
a
,
a
n
d
M
.
Tay
lo
r,
“
S
imp
le
B
M
2
5
e
x
ten
sio
n
to
m
u
lt
i
p
le
we
ig
h
ted
fiel
d
s,”
in
Pro
c
e
e
d
in
g
s
o
f
th
e
th
irtee
n
t
h
AC
M
in
ter
n
a
ti
o
n
a
l
c
o
n
fer
e
n
c
e
o
n
I
n
fo
rm
a
ti
o
n
a
n
d
k
n
o
wled
g
e
m
a
n
a
g
e
me
n
t
,
2
0
0
4
,
p
p
.
4
2
–
4
9
,
d
o
i:
1
0
.
1
1
4
5
/1
0
3
1
1
7
1
.
1
0
3
1
1
8
1
.
[3
1
]
C.
Ka
m
p
h
u
is,
A.
P
.
d
e
Vrie
s,
L.
Bo
y
tso
v
,
a
n
d
J.
Li
n
,
“
Wh
ic
h
B
M
2
5
d
o
y
o
u
m
e
a
n
?
A
larg
e
-
sc
a
le
re
p
ro
d
u
c
ib
il
it
y
stu
d
y
o
f
sc
o
rin
g
v
a
rian
ts,” i
n
Eu
r
o
p
e
a
n
Co
n
fer
e
n
c
e
o
n
In
f
o
rm
a
ti
o
n
Retrie
v
a
l
,
2
0
2
0
,
p
p
.
2
8
–
3
4
,
d
o
i:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
030
-
4
5
4
4
2
-
5
_
4
.
[3
2
]
D.
M
e
tzle
r
a
n
d
W
.
B.
Cro
f
t,
“
A
M
a
rk
o
v
ra
n
d
o
m
f
ield
m
o
d
e
l
fo
r
term
d
e
p
e
n
d
e
n
c
ies
,
”
in
Pro
c
e
e
d
i
n
g
s
o
f
th
e
2
8
th
a
n
n
u
a
l
in
ter
n
a
ti
o
n
a
l
ACM
S
IGIR
c
o
n
fer
e
n
c
e
o
n
Res
e
a
rc
h
a
n
d
d
e
v
e
lo
p
me
n
t
i
n
in
f
o
rm
a
ti
o
n
re
triev
a
l
,
2
0
0
5
,
p
p
.
4
7
2
–
4
7
9
,
d
o
i:
1
0
.
1
1
4
5
/
1
0
7
6
0
3
4
.
1
0
7
6
1
1
5
.
[3
3
]
E.
Yu
li
a
n
ti
,
R.
-
C.
C
h
e
n
,
F
.
S
c
h
o
ler,
W.
B.
Cr
o
ft,
a
n
d
M
.
S
a
n
d
e
rso
n
,
“
Ra
n
k
in
g
d
o
c
u
m
e
n
ts
b
y
a
n
sw
e
r
-
p
a
ss
a
g
e
q
u
a
li
t
y
,
”
in
S
IGIR
'
1
8
:
T
h
e
4
1
st
I
n
ter
n
a
ti
o
n
a
l
AC
M
S
IGIR
Co
n
fer
e
n
c
e
o
n
Res
e
a
rc
h
&
De
v
e
lo
p
me
n
t
in
In
f
o
rm
a
ti
o
n
Retrie
v
a
l
,
2
0
1
8
,
p
p
.
3
3
5
–
3
4
4
,
d
o
i
:
1
0
.
1
1
4
5
/3
2
0
9
9
7
8
.
3
2
1
0
0
2
8
.
[3
4
]
T
.
V
.
R
a
m
p
i
s
e
l
a
a
n
d
E
.
Y
u
l
i
a
n
t
i
,
“
S
e
m
a
n
t
i
c
-
B
a
s
e
d
Q
u
e
r
y
E
x
p
a
n
s
i
o
n
f
o
r
A
c
a
d
e
m
i
c
E
x
p
e
r
t
F
i
n
d
i
n
g
,
”
2
0
2
0
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
A
s
i
a
n
L
a
n
g
u
a
g
e
P
r
o
c
e
s
s
i
n
g
(
I
A
L
P
)
,
2
0
2
0
,
p
p
.
3
4
-
3
9
,
d
o
i
:
1
0
.
1
1
0
9
/
I
A
L
P
5
1
3
9
6
.
2
0
2
0
.
9
3
1
0
4
9
2
.
[3
5
]
T.
V.
Ra
m
p
ise
la
a
n
d
E
.
Y
u
li
a
n
ti
,
“
Ac
a
d
e
m
ic
Ex
p
e
rt
F
in
d
in
g
i
n
In
d
o
n
e
sia
u
sin
g
Wo
r
d
Emb
e
d
d
i
n
g
a
n
d
Do
c
u
m
e
n
t
Emb
e
d
d
i
n
g
:
A
Ca
se
S
tu
d
y
o
f
F
a
silk
o
m
UI
,
”
2
0
2
0
8
t
h
I
n
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
I
n
fo
rm
a
t
io
n
a
n
d
Co
mm
u
n
ica
ti
o
n
T
e
c
h
n
o
l
o
g
y
(IC
o
ICT
)
,
2
0
2
0
,
p
p
.
1
-
6
,
d
o
i
:
1
0
.
1
1
0
9
/
ICo
ICT4
9
3
4
5
.
2
0
2
0
.
9
1
6
6
2
4
9
.
[3
6
]
E.
Yu
li
a
n
t
i,
R.
Ch
e
n
,
F
.
S
c
h
o
ler,
W.
B.
Cro
ft
a
n
d
M
.
S
a
n
d
e
rso
n
,
“
Do
c
u
m
e
n
t
S
u
m
m
a
riza
ti
o
n
fo
r
A
n
sw
e
rin
g
No
n
-
F
a
c
to
id
Qu
e
ries
,
”
in
IEE
E
T
ra
n
sa
c
ti
o
n
s
o
n
Kn
o
wled
g
e
a
n
d
Da
t
a
En
g
in
e
e
rin
g
,
v
o
l.
3
0
,
n
o
.
1
,
p
p
.
1
5
-
2
8
,
2
0
1
8
,
d
o
i:
1
0
.
1
1
0
9
/T
KD
E.
2
0
1
7
.
2
7
5
4
3
7
3
.
[3
7
]
E.
Yu
li
a
n
ti
,
S
.
Hu
s
p
i,
a
n
d
M
.
S
a
n
d
e
rso
n
,
“
Twe
e
t
-
b
ias
e
d
s
u
m
m
a
riza
ti
o
n
,
”
J
o
u
rn
a
l
o
f
t
h
e
Asso
c
ia
ti
o
n
fo
r
In
fo
rm
a
t
io
n
S
c
ien
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
6
7
,
n
o
.
6
,
p
p
.
1
2
8
9
–
1
3
0
0
,
2
0
1
6
,
d
o
i:
1
0
.
1
0
0
2
/as
i.
2
3
4
9
6
.
Evaluation Warning : The document was created with Spire.PDF for Python.