I
A
E
S
I
n
t
e
r
n
at
io
n
al
Jou
r
n
al
of
A
r
t
if
ic
ia
l
I
n
t
e
ll
ig
e
n
c
e
(
I
J
-
AI
)
V
ol
.
10
, N
o.
4
,
D
e
c
e
m
be
r
202
1
, pp.
1060
~
1068
I
S
S
N
:
2252
-
8938
,
D
O
I
:
10.11591/
ij
a
i.
v
10
.i
4
.pp
1060
-
1068
1060
Jou
r
n
al
h
om
e
page
:
ht
tp
:
//
ij
ai
.
ia
e
s
c
or
e
.c
om
A
n
ove
l
on
t
ol
ogy f
r
am
e
w
or
k
su
p
p
or
t
i
n
g m
o
d
e
l
-
b
ase
d
t
ou
r
i
sm
r
e
c
om
m
e
n
d
e
r
H
o Q
u
oc
D
u
n
g
1
, L
ie
n
T
h
i
Q
u
yn
h
L
e
2
, N
gu
ye
n
H
u
u
H
oan
g
T
h
o
3
, T
r
i
Q
u
oc
T
r
u
on
g
4
,
C
u
on
g H
. N
gu
ye
n
-
D
in
h
5
1
,5
School o
f Engin
eerin
g and
Techn
ology, H
ue Uni
versity
, Huế,
Viet N
am
2
University of Econo
mics, Hue Universit
y, Huế, Viet Nam
3
Tomas Bata University, Zlín, C
zech Republic
4
Faculty
of En
gineer
ing, Va
n Lang
Unive
rsity, H
o Chi Min
h, Vie
t Nam
A
r
t
ic
le
I
n
f
o
A
B
S
T
R
A
C
T
A
r
ti
c
le
h
is
to
r
y
:
R
e
c
e
iv
e
d
D
e
c
18
,
2020
R
e
vi
s
e
d
A
ug
12
,
2021
A
c
c
e
pt
e
d
A
ug
20
,
2021
In
this
paper,
we
present
a
tourism
recomm
ender
framework
based
on
the
cooperatio
n
of
ontolo
gical
knowledg
e
base
and
supervis
ed
learning
models.
Specif
ically,
a
new
tourism
ontology
,
which
not
only
captur
es
domain
knowledge
but
also
specifies
knowledge
ent
ities
in
numerical
vector
s
pace,
is
presented.
The
recommendation
making
process
enables
machine
learning
models
to
work
directly
with
the
ontological
knowledge
base
from
training
step
to
deployment
step.
This
knowledge
base
can
work
we
ll
with
classifi
cati
on
models
(e.g.,
k
-
n
earest
neighbours,
support
vector
ma
c
hines,
or
naıve
bayes
).
A
prototype
of
the
framework
is
developed
and
expe
rimental
results con
firm the feasi
bility
of the propo
sed framework.
K
e
y
w
o
r
d
s
:
O
nt
ol
ogy
S
e
m
a
nt
ic
s
im
il
a
r
it
y
S
e
m
a
nt
ic
ve
c
to
r
S
upe
r
vi
s
e
d l
e
a
r
ni
ng mode
ls
T
our
is
m
r
e
c
om
m
e
nde
r
This is an
open
acce
ss artic
le unde
r
the
CC BY
-
SA
license
.
C
or
r
e
s
pon
di
n
g A
u
th
or
:
C
uong H. N
guye
n
-
D
in
h
S
c
hool
of
E
ngi
ne
e
r
in
g a
nd T
e
c
hnol
ogy
H
ue
U
ni
ve
r
s
it
y
1 D
ie
n B
ie
n P
hu S
t.
, H
ue
C
it
y, V
ie
t
N
a
m
E
m
a
il
:
ndhc
uong@
hue
uni
.e
du.vn
1.
I
N
T
R
O
D
U
C
T
I
O
N
R
e
c
om
m
e
nde
r
s
ys
te
m
s
m
a
ke
us
e
of
m
a
c
hi
n
e
le
a
r
ni
ng
m
od
e
ls
in
th
e
ir
d
e
c
is
io
n
m
a
ki
ng
pr
oc
e
s
s
.
T
he
s
e
m
ode
l
-
ba
s
e
d r
e
c
om
m
e
nde
r
s
ys
t
e
m
s
of
te
n us
e
t
he
ve
c
to
r
-
ba
s
e
d r
e
c
om
m
e
nde
r
da
ta
s
e
ts
(
e
.g., M
ovi
e
L
e
ns
[
1]
,
book
-
c
r
os
s
in
g
[
2]
)
f
or
m
e
a
s
ur
in
g
pe
r
f
or
m
a
nc
e
s
in
e
xpe
r
im
e
n
ts
.
W
hi
le
th
e
s
e
da
ta
s
e
ts
a
r
e
li
m
it
e
d
in
s
e
ve
r
a
l
dom
a
in
s
(
e
.g.,
m
ovi
e
s
,
books
)
,
th
e
gr
a
ph
-
ba
s
e
d
ope
n
li
nke
d
da
ta
(
e
.g.,
D
B
pe
di
a
[
3]
)
pr
ovi
de
da
ta
in
m
a
ny
f
ie
ld
s
a
nd
ha
ve
be
e
n
us
e
d
a
s
a
s
uppl
e
m
e
nt
a
r
y
da
ta
s
ou
r
c
e
in
r
e
c
e
nt
r
e
c
om
m
e
nde
r
r
e
s
e
a
r
c
h
[
4]
,
[
5]
.
H
ow
e
v
e
r
, t
he
gr
a
ph na
tu
r
e
of
ope
n l
in
ke
d da
ta
m
a
ke
s
i
t
di
f
f
ic
ul
t
to
be
c
ons
um
e
d by ma
c
hi
ne
l
e
a
r
ni
ng
m
ode
ls
a
nd
a
f
e
w
dom
a
in
s
of
r
e
c
om
m
e
nd
e
r
da
ta
s
e
ts
a
r
e
not
e
nough
t
o
bui
ld
r
e
a
l
-
li
f
e
s
pe
c
if
ic
r
e
c
om
m
e
nde
r
s
(
e
.g.,
to
ur
is
m
r
e
c
om
m
e
nde
r
s
)
.
I
n
or
de
r
to
f
il
l
th
is
ga
p,
our
s
tu
dy
f
oc
us
e
s
on
c
ons
tr
uc
ti
ng
ve
c
to
r
-
ba
s
e
d
da
t
a
f
or
ont
ol
ogi
c
a
l
knowle
dge
ba
s
e
a
nd ge
ne
r
a
ti
ng t
our
is
m
r
e
c
om
m
e
n
da
ti
on i
te
m
s
ba
s
e
d on the
u
s
e
of
t
he
s
e
ve
c
to
r
s
.
I
n
th
is
pa
pe
r
,
w
e
in
tr
oduc
e
a
nove
l
ont
ol
ogi
c
a
l
f
r
a
m
e
w
or
k
th
a
t
s
uppor
ts
m
ode
l
-
ba
s
e
d
to
ur
is
m
r
e
c
om
m
e
nde
r
in
ge
ne
r
a
ti
ng
to
p
-
K
pe
r
s
ona
li
z
e
d
r
e
c
om
m
e
nda
ti
ons
.
T
o
be
m
or
e
s
pe
c
if
ic
,
w
e
de
s
ig
n
a
to
ur
is
m
ont
ol
ogy
f
or
m
a
c
hi
ne
le
a
r
ni
ng
s
o
-
c
a
ll
e
d
t
our
is
m
ont
ol
ogy
f
o
r
m
a
c
hi
ne
le
a
r
ni
ng
(
T
O
M
L
)
w
hi
c
h
c
a
pt
ur
e
s
knowle
dge
of
to
ur
is
m
dom
a
in
a
nd
a
ls
o
in
te
gr
a
te
s
w
it
h
o
ut
s
o
ur
c
e
knowle
dge
ba
s
e
s
(
e
.g.
D
B
pe
di
a
or
lo
c
a
l
da
ta
ba
s
e
s
)
.
F
ur
th
e
r
m
or
e
,
w
e
c
ons
tr
uc
t
th
e
s
e
m
a
nt
ic
ve
c
to
r
c
la
s
s
to
e
nc
ode
e
ve
r
y
e
nt
it
y’
s
pr
ope
r
ti
e
s
in
num
e
r
ic
a
l
ve
c
to
r
s
pa
c
e
.
A
lg
or
it
hm
s
a
r
e
pr
opos
e
d
to
qua
nt
i
f
y
di
m
e
ns
io
na
l
va
lu
e
s
f
or
e
a
c
h
in
s
ta
nc
e
of
s
e
m
a
nt
ic
ve
c
to
r
.
T
he
r
e
c
om
m
e
nda
ti
on
e
ngi
n
e
is
de
s
ig
n
e
d
to
g
e
ne
r
a
te
to
p
-
K
r
e
c
om
m
e
nd
a
ti
ons
ba
s
e
d
on
th
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
A
nov
e
l
ont
ol
ogy
f
r
am
e
w
or
k
s
uppo
r
ti
ng m
ode
l
-
bas
e
d t
our
i
s
m
r
e
c
om
m
e
nde
r
(
H
o Q
uoc
D
ung
)
1061
c
a
lc
ul
a
ti
on
of
s
e
m
a
nt
ic
s
im
il
a
r
it
y
o
r
th
e
us
e
o
f
s
upe
r
vi
s
e
d
le
a
r
ni
ng
m
ode
ls
.
T
w
o
e
xpe
r
im
e
nt
s
a
r
e
c
onduc
te
d
a
nd t
he
e
xpe
r
im
e
nt
a
l
r
e
s
ul
ts
c
onf
ir
m
t
he
f
e
a
s
ib
il
it
y o
f
our
pr
opos
e
d f
r
a
m
e
w
or
k.
T
he
r
e
s
t
of
th
is
pa
p
e
r
is
or
ga
ni
z
e
d
a
s
:
S
e
c
ti
on
2
de
s
c
r
ib
e
s
th
e
r
e
la
te
d
w
or
k.
I
n
s
e
c
ti
on
3,
th
e
T
O
M
L
,
th
e
a
r
c
hi
te
c
tu
r
e
of
T
O
M
L
-
ba
s
e
d
to
ur
is
m
r
e
c
om
m
e
nde
r
a
nd
it
s
de
c
is
io
n
-
m
a
ki
ng
pr
oc
e
s
s
a
r
e
pr
e
s
e
nt
e
d.
S
e
c
ti
on
4
dr
a
w
s
th
e
e
xpe
r
im
e
nt
s
a
nd
di
s
c
us
s
e
s
th
e
r
e
s
ul
ts
.
F
in
a
ll
y,
s
e
c
ti
on
5
gi
ve
s
th
e
c
onc
lu
s
io
n
a
nd
s
ta
te
s
th
e
f
ut
ur
e
w
or
k.
2.
RE
L
A
T
E
D
W
O
R
K
I
n
th
is
s
e
c
ti
on,
w
e
a
na
ly
z
e
th
e
r
e
c
e
nt
m
e
th
ods
of
to
ur
is
m
r
e
c
o
m
m
e
nde
r
s
in
c
lu
di
ng
m
a
c
hi
ne
le
a
r
ni
ng
a
nd
s
e
m
a
nt
ic
w
e
b
ba
s
e
d
a
ppr
oa
c
he
s
.
T
he
r
e
vi
e
w
s
of
r
e
c
om
m
e
nde
r
s
ys
te
m
s
a
nd
to
ur
is
m
r
e
c
om
m
e
nde
r
s
a
r
e
out
of
th
e
s
c
ope
of
th
is
s
tu
dy
a
nd
c
a
n
be
f
ound
in
th
e
f
ol
lo
w
in
g
s
ur
ve
ys
[
6]
,
[
7]
,
r
e
s
pe
c
ti
ve
ly
.
T
r
a
di
ti
ona
ll
y,
c
ol
la
bor
a
ti
ve
f
il
te
r
in
g,
c
ont
e
nt
ba
s
e
d
f
il
te
r
in
g
a
nd
hybr
id
m
e
th
ods
a
r
e
dom
in
a
nt
a
ppr
oa
c
he
s
to
r
e
c
om
m
e
nd
e
r
s
ys
te
m
s
.
T
he
s
tr
e
ngt
h
a
nd
w
e
a
kne
s
s
of
th
e
s
e
m
e
th
ods
a
r
e
a
na
l
yz
e
d
in
[
6]
.
B
e
s
id
e
s
,
m
a
c
hi
ne
le
a
r
ni
ng
is
a
ls
o
a
ppl
ie
d
in
r
e
c
om
m
e
nde
r
s
f
or
gi
vi
ng
pe
r
s
ona
li
z
e
d
r
e
c
om
m
e
nda
ti
ons
.
S
pe
c
if
ic
a
ll
y,
c
la
s
s
if
ic
a
ti
on
m
e
th
ods
th
a
t
a
r
e
w
id
e
ly
us
e
d
in
m
a
ki
ng
r
e
c
om
m
e
nda
ti
ons
a
r
e
s
uppor
t
ve
c
to
r
m
a
c
hi
ne
(
S
V
M
)
[
8]
,
k
-
ne
a
r
e
s
t
ne
ig
hbor
s
(
kN
N
)
[
9]
,
a
r
ti
f
ic
ia
l
ne
ur
a
l
ne
twor
k
(
A
N
N
)
[
10]
,
de
c
is
io
n
tr
e
e
[
11]
or
e
ns
e
m
bl
e
m
e
th
od
[
12]
to
na
m
e
a
f
e
w
.
I
n
th
e
dom
a
in
of
to
u
r
is
m
r
e
c
om
m
e
nde
r
s
,
tr
a
di
ti
ona
l
m
e
th
ods
[
13]
a
nd
m
a
c
hi
ne
le
a
r
ni
ng
m
e
th
ods
[
14]
a
r
e
a
ls
o i
nt
r
oduc
e
d t
o t
he
li
te
r
a
tu
r
e
. B
ot
h
t
r
a
di
ti
ona
l
r
e
c
om
m
e
nde
r
m
e
th
ods
a
nd ma
c
hi
n
e
l
e
a
r
ni
ng
-
ba
s
e
d m
e
th
ods
a
r
e
da
ta
de
p
e
nde
nt
.
T
hi
s
m
e
a
n
s
th
a
t
th
e
qua
nt
it
y
a
nd
th
e
qua
li
ty
of
da
ta
de
c
id
e
th
e
pe
r
f
or
m
a
nc
e
of
r
e
c
om
m
e
nde
r
s
ys
te
m
s
.
H
ow
e
ve
r
,
th
e
la
c
k
of
da
ta
of
te
n
o
c
c
ur
s
in
r
e
c
om
m
e
nde
r
s
tu
di
e
s
.
T
hi
s
is
th
e
r
oot
of
th
e
c
ol
d
-
s
ta
r
t
pr
obl
e
m
of
r
e
c
om
m
e
nde
r
[
15]
.
I
n
or
de
r
to
s
uppor
t
r
e
c
om
m
e
nde
r
s
in
bui
ld
in
g
it
s
pr
e
di
c
ti
on
m
ode
l,
r
e
s
e
a
r
c
he
r
s
ha
v
e
us
e
d
s
uppl
e
m
e
nt
a
r
y
da
t
a
s
e
t
s
to
ove
r
c
om
e
th
is
di
f
f
ic
ul
ty
.
I
n
w
hi
c
h,
ope
n
li
nke
d
da
ta
ha
s
be
e
n a
dopt
e
d a
s
t
he
m
ode
r
n a
ppr
oa
c
h [
15]
. T
he
us
e
of
ope
n l
in
ke
d da
t
a
a
nd r
e
a
s
oni
ng t
e
c
hni
que
s
of
s
e
m
a
nt
ic
w
e
b
te
c
hnol
ogy
a
r
e
a
ls
o
f
ound
in
to
ur
is
m
r
e
c
om
m
e
nde
r
s
[
4]
a
nd
in
ot
he
r
ki
nds
of
r
e
c
om
m
e
nde
r
s
[
5]
.
A
s
a
r
e
s
ul
t,
c
om
bi
ni
ng
m
a
c
hi
ne
le
a
r
ni
ng
w
it
h
ope
n
li
nke
d
da
ta
a
nd
s
e
m
a
nt
ic
w
e
b
te
c
hnol
ogy
ha
s
be
c
om
e
a
r
is
in
g
tr
e
nd
in
r
e
c
om
m
e
nde
r
s
tu
di
e
s
.
I
n
th
is
pa
pe
r
,
our
ta
r
ge
t
not
on
ly
pr
ovi
de
s
a
ne
w
hyb
r
id
f
r
a
m
e
w
or
k
but
a
ls
o
pr
e
s
e
nt
s
a
ne
w
ont
ol
ogy
to
to
ur
is
m
r
e
c
om
m
e
nde
r
.
T
h
e
r
e
s
t
of
th
is
s
e
c
ti
on
r
e
vi
e
w
s
th
e
r
e
c
e
nt
s
tu
di
e
s
w
it
h
a
f
oc
us
on:
i)
ont
ol
ogy
e
ngi
ne
e
r
in
g
m
e
th
odol
ogi
e
s
;
ii
)
ont
ol
ogi
e
s
f
or
th
e
to
ur
is
m
in
dus
tr
y;
a
nd
i
ii
)
di
s
c
us
s
io
n
a
bout
t
he
di
s
ti
nc
t
c
ha
r
a
c
te
r
is
ti
c
s
of
our
pr
opos
e
d a
ppr
oa
c
h.
2.1. On
t
ol
ogy e
n
gi
n
e
e
r
in
g m
e
t
h
od
ol
ogi
e
s
I
n
2001,
B
e
r
ne
r
-
L
e
e
e
t
al
.
[
16]
p
r
opos
e
d
th
e
S
e
m
a
nt
ic
W
e
b
in
it
ia
ti
ve
w
hi
c
h
hi
ghl
ig
ht
s
th
e
ke
y
r
ol
e
of
ont
ol
ogy
a
s
a
n
e
f
f
ic
ie
nt
w
a
y
to
c
a
pt
ur
e
dom
a
in
kno
w
le
dge
i
n
m
a
c
hi
ne
und
e
r
s
ta
nda
bl
e
f
or
m
a
t.
S
in
c
e
th
e
n,
th
e
r
e
s
e
a
r
c
h
tr
e
nd
na
m
e
d
ont
ol
ogy
e
ngi
ne
e
r
in
g,
w
hi
c
h
f
oc
u
s
e
s
on
m
e
th
ods
of
de
ve
lo
pi
ng
dom
a
in
ont
ol
ogy,
ha
s
be
e
n r
a
is
e
d. I
n t
hi
s
r
e
s
e
a
r
c
h t
r
e
nd, t
h
e
t
ut
or
ia
l
of
N
oy
a
nd
M
c
G
ui
ne
s
s
[
17]
c
a
n be
s
e
e
n a
s
one
of
t
he
m
os
t
popula
r
m
e
th
ods
of
ont
ol
ogy
bui
ld
in
g.
T
he
a
ut
hor
s
pr
opos
e
d
7 s
te
p
m
e
th
od
in
c
lu
di
ng:
i)
de
te
r
m
in
e
th
e
s
c
op
e
of
th
e
dom
a
in
ont
ol
ogy;
ii
)
r
e
us
e
e
xi
s
ti
ng
ont
ol
ogi
e
s
;
ii
i)
e
num
e
r
a
te
dom
a
in
c
onc
e
pt
s
;
iv
)
c
on
s
tr
uc
t
th
e
c
la
s
s
a
nd
th
e
c
la
s
s
hi
e
r
a
r
c
hy;
v)
de
f
in
e
th
e
pr
ope
r
ti
e
s
of
th
e
c
la
s
s
-
s
lo
ts
;
vi
)
de
f
in
e
th
e
f
a
c
e
ts
of
th
e
s
lo
ts
;
a
nd
vi
i)
c
r
e
a
te
in
s
ta
nc
e
s
.
A
lt
hough
th
is
m
e
th
od
is
e
f
f
ic
ie
nt
,
it
f
a
c
e
s
th
e
di
f
f
ic
ul
ti
e
s
in
ont
ol
ogy
e
vol
ut
io
n
a
nd
c
ol
la
bor
a
ti
ve
bui
ld
in
g
of
ont
ol
ogy.
T
he
r
e
f
or
e
,
di
f
f
e
r
e
nt
on
to
lo
gy
e
ngi
ne
e
r
in
g
m
e
th
ods
ha
ve
be
e
n
pr
e
s
e
nt
e
d.
F
or
e
xa
m
pl
e
,
F
e
r
na
nde
z
-
L
ope
z
e
t
al
.
[
18]
f
oc
us
e
d
on
th
e
m
a
jo
r
s
ubt
a
s
k
s
to
d
e
ve
lo
p
n
e
w
ont
ol
ogi
e
s
a
nd
th
e
e
vol
ut
io
n of
ont
ol
ogy thr
oughout i
ts
l
i
f
e
ti
m
e
. I
n
a
not
he
r
a
ppr
oa
c
h, S
ur
e
e
t
al
. [
19]
pr
e
s
e
nt
e
d on
-
to
knowle
dge
m
e
th
odol
ogy
(
O
T
K
M
)
w
hi
c
h
ta
ke
s
a
c
c
ount
of
th
e
knowle
dge
pr
oc
e
s
s
e
s
a
nd
th
e
knowl
e
dge
m
e
ta
pr
oc
e
s
s
e
s
.
T
he
f
or
m
e
r
pr
oc
e
s
s
r
e
la
te
s
to
th
e
us
a
ge
of
ont
ol
ogi
e
s
,
w
hi
le
t
he
la
tt
e
r
pr
oc
e
s
s
m
a
k
e
s
in
it
ia
l
s
e
tu
p.
O
T
K
M
in
tr
oduc
e
s
th
e
w
a
ys
of
in
te
gr
a
ti
ng
ont
ol
ogy
in
knowle
dge
m
a
na
ge
m
e
nt
a
ppl
ic
a
ti
on
s
.
T
he
N
e
O
N
m
e
th
odol
ogy
[
20]
is
di
f
f
e
r
e
nt
f
r
om
pr
e
vi
ous
m
e
th
odol
ogi
e
s
.
W
hi
le
pr
e
vi
ous
s
tu
di
e
s
bui
ld
s
ta
nd
a
lo
ne
ont
ol
ogi
e
s
,
th
e
N
e
O
N
m
e
th
odol
ogy
c
on
s
tr
uc
ts
a
n
ont
ol
og
y
ne
twor
k
by
c
onne
c
ti
ng
di
f
f
e
r
e
nt
e
xi
s
ti
ng
ont
ol
ogi
e
s
t
hr
ough the
ir
r
e
la
ti
ons
hi
ps
.
2.2. On
t
ol
ogi
e
s
f
or
t
ou
r
is
m
i
n
d
u
s
t
r
y
R
e
c
e
nt
ly
,
S
e
m
a
nt
ic
W
e
b
te
c
hnol
ogy
a
nd
ont
ol
ogy
ha
ve
be
e
n
a
ppl
ie
d
to
to
ur
is
m
r
e
c
om
m
e
nde
r
s
in
m
a
ny
a
s
pe
c
ts
.
T
o
be
m
or
e
s
pe
c
if
ic
,
A
nt
oni
o
M
or
e
no
e
t
al
.
[
21]
us
e
d
ont
ol
ogy
to
c
a
pt
ur
e
knowle
dge
of
to
ur
is
m
obj
e
c
ts
a
nd
popula
te
d
th
e
ont
ol
ogi
c
a
l
in
s
ta
n
c
e
s
w
it
h
s
c
o
r
e
s
.
T
he
s
e
s
c
or
e
s
w
e
r
e
th
e
in
put
s
of
th
e
r
e
c
om
m
e
nda
ti
on
a
lg
or
it
hm
.
L
in
S
hi
e
t
al
.
[
22]
pr
ovi
de
d
to
ur
is
m
r
e
c
om
m
e
nda
ti
ons
ba
s
e
d
on
th
e
u
s
e
r
’
s
c
ont
e
xt
.
I
n
w
hi
c
h,
ont
ol
ogy
w
a
s
us
e
d
to
de
s
c
r
ib
e
a
nd
in
te
gr
a
te
to
ur
is
m
r
e
s
our
c
e
s
.
B
a
s
e
d
on
th
i
s
knowle
dge
f
ounda
ti
on
,
th
e
r
e
a
s
oni
ng
pr
oc
e
s
s
w
a
s
im
pl
e
m
e
nt
e
d
to
m
a
ke
p
e
r
s
ona
li
z
e
d
r
e
c
om
m
e
nd
a
ti
ons
.
G
r
un
e
t
al
.
[
4]
in
tr
oduc
e
d
a
n
ont
ol
ogy
-
ba
s
e
d
m
e
th
od
to
s
uppor
t
to
ur
is
ts
’
de
c
is
io
n
-
m
a
ki
ng
dur
in
g
th
e
i
r
pr
e
-
t
r
ip
pha
s
e
.
T
he
a
ut
hor
s
m
a
tc
he
d
to
ur
is
ts
’
pr
of
il
e
s
w
it
h
c
ha
r
a
c
te
r
is
ti
c
s
of
to
ur
i
s
m
obj
e
c
ts
th
r
ough
ve
c
to
r
s
pa
c
e
w
he
r
e
e
a
c
h
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
.
10
, N
o.
4
,
D
e
c
e
m
be
r
202
1
:
1060
-
1068
1062
di
m
e
ns
io
n
is
a
to
ur
is
t
f
a
c
to
r
.
I
n
a
not
he
r
a
ppr
oa
c
h,
P
.
F
e
r
r
a
r
o
a
nd
L
.
R
.
G
iu
s
e
ppe
[
23]
pr
opos
e
d
a
n
a
r
c
hi
te
c
tu
r
e
of
a
s
e
m
a
nt
ic
a
ll
y
a
da
pt
iv
e
r
e
c
om
m
e
nd
e
r
s
ys
te
m
a
s
s
is
ti
ng
us
e
r
s
in
th
e
tr
a
ve
l
pl
a
nni
n
g
pha
s
e
a
nd
in
on
-
s
it
e
pha
s
e
.
H
ybr
id
m
e
th
od
of
to
ur
is
m
r
e
c
om
m
e
nde
r
w
a
s
a
ls
o
in
tr
oduc
e
d
to
th
e
li
te
r
a
tu
r
e
in
th
e
r
e
s
e
a
r
c
h
of
Y
a
n C
hu
e
t
al
.
[
24]
.
F
ir
s
tl
y,
t
he
a
ut
hor
s
u
s
e
d a
s
s
oc
ia
ti
on r
ul
e
s
t
o f
in
d out
r
e
la
te
d us
e
r
s
a
nd unr
e
la
te
d us
e
r
s
.
S
e
c
ondl
y,
f
or
e
a
c
h
gr
oup
of
us
e
r
s
,
th
e
y
a
ppl
ie
d
di
f
f
e
r
e
nt
c
ol
la
bor
a
ti
ve
f
il
te
r
in
g
a
lg
or
it
hm
s
to
m
a
ke
r
e
c
om
m
e
nda
ti
ons
. F
in
a
ll
y, t
he
r
e
c
om
m
e
nda
ti
ons
w
e
r
e
e
xpa
nde
d by us
in
g a
t
our
is
m
ont
ol
og
y.
2.3. Dis
c
u
s
s
io
n
B
ot
h
r
e
c
om
m
e
nde
r
s
a
nd
m
a
c
hi
ne
le
a
r
ni
ng
m
ode
ls
r
e
qui
r
e
da
ta
w
hi
c
h
is
of
te
n
in
num
e
r
ic
a
l
ve
c
to
r
f
or
m
a
t.
H
ow
e
ve
r
,
th
is
ki
nd
of
da
ta
is
not
a
lw
a
ys
a
va
il
a
bl
e
,
e
s
pe
c
ia
ll
y
in
th
e
r
e
s
e
a
r
c
h
li
ne
of
to
ur
is
m
r
e
c
om
m
e
nde
r
.
O
n
th
e
ot
he
r
ha
nd,
th
e
r
e
a
r
e
m
a
ny
va
lu
a
bl
e
op
e
n
li
nke
d
da
ta
s
our
c
e
s
(
e
.g.
D
B
pe
di
a
)
,
w
hi
c
h
r
e
s
id
e
unde
r
gr
a
ph
-
ba
s
e
d
f
or
m
a
ts
,
c
a
n
e
f
f
ic
ie
nt
ly
s
uppor
t
th
e
r
e
c
om
m
e
nda
ti
on
m
a
ki
ng
pr
oc
e
s
s
. T
he
pr
obl
e
m
is
to
tr
a
ns
f
e
r
di
r
e
c
tl
y
th
e
gr
a
ph
-
ba
s
e
d
da
ta
to
num
e
r
ic
a
l
ve
c
to
r
s
in
or
de
r
to
s
e
r
ve
d
i
f
f
e
r
e
nt
m
a
c
hi
ne
le
a
r
ni
ng
m
ode
ls
i
n pr
e
di
c
ti
ng us
e
r
’
s
pr
e
f
e
r
e
nc
e
s
or
ge
ne
r
a
ti
ng t
op
-
K
pe
r
s
ona
li
z
e
d r
e
c
om
m
e
nda
ti
on l
is
ts
. T
o s
ol
ve
t
hi
s
pr
obl
e
m
,
our
pr
opos
e
d
f
r
a
m
e
w
or
k
is
di
f
f
e
r
e
nt
f
r
om
th
e
a
f
or
e
m
e
nt
io
ne
d
r
e
s
e
a
r
c
h
in
th
e
f
ol
lo
w
in
g
th
r
e
e
a
s
pe
c
ts
.
F
ir
s
tl
y,
w
e
in
tr
oduc
e
a
ne
w
to
ur
is
m
ont
ol
ogy
ba
s
e
d
o
n
dom
a
in
e
xpe
r
t
c
ol
la
bor
a
ti
on
a
nd
out
s
our
c
e
knowle
dge
in
te
gr
a
ti
on.
S
e
c
ondl
y,
a
S
e
m
a
nt
ic
ve
c
t
or
c
onc
e
pt
is
us
e
d
to
de
s
c
r
ib
e
e
ve
r
y
e
nt
it
y
of
th
e
ont
ol
ogy
in
a
ve
c
to
r
s
pa
c
e
m
ode
l.
T
hi
s
c
om
pone
nt
pr
ovi
de
s
s
e
m
a
nt
ic
a
ll
y
nu
m
e
r
ic
a
l
da
ta
f
or
a
ll
m
a
c
hi
ne
l
e
a
r
ni
ng t
a
s
ks
in
c
lu
di
ng
c
la
s
s
if
ic
a
ti
on
a
nd
c
lu
s
te
r
in
g.
T
hi
r
dl
y,
w
e
pr
e
s
e
nt
a
lg
or
it
hm
s
f
or
th
e
r
e
c
om
m
e
nda
ti
on
e
ngi
ne
w
hi
c
h
us
e
di
r
e
c
tl
y
th
e
s
e
m
a
nt
ic
num
e
r
ic
a
l
da
ta
in
th
e
r
e
c
om
m
e
nd
a
ti
on
m
a
ki
ng
pr
oc
e
s
s
.
T
hi
s
a
ppr
oa
c
h
is
di
f
f
e
r
e
n
t
f
r
om
t
he
pr
e
vi
ous
us
e
of
ot
he
r
ont
ol
ogi
e
s
i
n t
he
t
our
is
m
doma
in
.
3.
T
O
M
L
-
B
A
S
E
D
R
E
C
O
M
M
E
N
D
E
R
F
R
A
M
E
WO
R
K
I
n
th
is
s
e
c
ti
on,
w
e
de
s
c
r
ib
e
our
ont
ol
ogi
c
a
l
a
ppr
oa
c
h
to
th
e
t
our
is
m
r
e
c
om
m
e
nde
r
na
m
e
d
T
O
M
L
.
T
O
M
L
-
ba
s
e
d
r
e
c
om
m
e
nd
e
r
f
r
a
m
e
w
or
k
ha
s
th
r
e
e
m
a
jo
r
pa
r
ts
in
c
lu
d
in
g
T
O
M
L
ont
ol
ogy,
m
e
th
ods
of
popula
ti
ng
T
O
M
L
knowle
dge
ba
s
e
a
nd
T
O
M
L
-
ba
s
e
d
r
e
c
om
m
e
nda
ti
on
e
ngi
ne
.
F
ig
ur
e
1
s
how
s
th
e
ove
r
a
ll
a
r
c
hi
te
c
tu
r
e
of
th
is
f
r
a
m
e
w
or
k.
F
ig
ur
e
1
. T
he
ove
r
a
ll
a
r
c
hi
te
c
tu
r
e
of
T
O
M
L
-
ba
s
e
d
r
e
c
om
m
e
nd
e
r
f
r
a
m
e
w
or
k
I
n
th
is
f
r
a
m
e
w
or
k,
th
e
T
O
M
L
ont
ol
ogy
w
a
s
d
e
s
ig
ne
d
th
r
ough
th
e
pr
opos
e
d
s
ix
-
s
te
p
pr
oc
e
s
s
w
hi
c
h
is
pr
e
s
e
nt
e
d
in
th
e
s
ub
s
e
c
ti
on
3.1.
T
h
e
T
O
M
L
knowle
dg
e
ba
s
e
w
a
s
e
nr
ic
h
e
d
by
di
f
f
e
r
e
nt
w
a
y
s
li
ke
im
por
ti
ng
f
r
om
D
B
pe
di
a
,
lo
c
a
l
da
ta
ba
s
e
s
a
n
d
to
ur
is
ts
’
pr
e
f
e
r
e
nc
e
s
da
ta
.
T
he
e
nr
ic
hi
ng
m
e
th
ods
a
r
e
di
s
c
us
s
e
d
in
th
e
s
ubs
e
c
ti
on 3.2. S
ubs
e
c
ti
on 3.3
in
tr
oduc
e
s
t
he
r
e
c
om
m
e
nd
a
ti
on e
ngi
ne
of
t
hi
s
f
r
a
m
e
w
or
k.
3.1. T
O
M
L
on
t
ol
ogy
I
n ge
ne
r
a
l,
a
doma
in
ont
ol
ogy c
a
n be
de
f
in
e
d a
s
i
n D
e
f
in
it
io
n 1.
−
D
e
f
in
it
io
n
1
T
he
ont
ol
ogy
of
th
e
dom
a
in
de
not
e
d
a
s
is
a
tr
ip
le
=
<
;
;
>
w
he
r
e
is
th
e
s
e
t
of
dom
a
in
c
onc
e
pt
s
,
a
nd
a
r
e
th
e
s
e
ts
of
dom
a
in
pr
ope
r
ti
e
s
(
r
e
la
ti
ons
)
a
nd
dom
a
i
n
in
s
ta
nc
e
s
, r
e
s
pe
c
ti
ve
ly
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
A
nov
e
l
ont
ol
ogy
f
r
am
e
w
or
k
s
uppo
r
ti
ng m
ode
l
-
bas
e
d t
our
i
s
m
r
e
c
om
m
e
nde
r
(
H
o Q
uoc
D
ung
)
1063
I
n
or
de
r
to
bui
ld
T
O
M
L
ont
ol
ogy,
w
e
in
vi
te
d
to
ur
is
m
e
xpe
r
ti
s
e
s
a
nd
knowle
dge
e
ngi
ne
e
r
s
to
w
or
k
to
ge
th
e
r
s
.
T
he
w
or
ki
ng
pr
oc
e
s
s
of
th
is
gr
oup
in
c
lu
de
s
s
ix
s
t
e
p
s
:
A
t
f
ir
s
t,
w
e
a
da
pt
e
d
th
e
m
e
th
od
of
[
17]
f
or
c
r
e
a
ti
ng
th
e
f
ir
s
t
dr
a
f
t
of
th
e
knowle
dge
ba
s
e
.
S
pe
c
if
ic
a
ll
y,
e
x
pe
r
ti
s
e
s
e
num
e
r
a
te
d
th
e
c
on
c
e
pt
s
a
nd
r
e
la
ti
ons
of
t
he
to
ur
is
m
dom
a
in
.
T
he
n,
knowle
dge
e
ngi
ne
e
r
s
tr
a
ns
f
e
r
r
e
d
th
e
s
e
in
f
or
m
a
ti
on
to
ont
ol
ogy
s
tr
uc
tu
r
e
us
in
g
P
r
ot
é
gé
[
25]
s
of
twa
r
e
.
S
e
c
ondl
y,
th
e
f
ir
s
t
s
te
p
w
a
s
r
e
pe
a
te
d
unt
il
a
ll
of
th
e
e
xpe
r
ti
s
e
s
a
nd
e
ngi
ne
e
r
s
r
e
a
c
he
d
th
e
ir
c
ons
e
ns
us
.
T
hi
r
dl
y,
f
ur
th
e
r
s
pe
c
if
ic
de
s
c
r
ip
ti
ons
w
e
r
e
a
d
de
d
to
th
e
ont
ol
ogy
(
e
.g.
th
e
S
e
m
a
nt
ic
V
e
c
to
r
c
la
s
s
)
.
F
our
th
ly
,
w
e
e
nr
ic
he
d
th
e
ont
ol
ogi
c
a
l
in
s
ta
nc
e
s
by
u
s
in
g
our
lo
c
a
l
da
ta
ba
s
e
a
nd
im
por
ti
ng
da
ta
f
r
om
ope
n
knowle
dge
-
ba
s
e
(
e
.g.
D
B
pe
di
a
)
th
r
ough
m
a
ppi
ng
ope
r
a
ti
ons
.
F
if
th
ly
,
w
e
it
e
r
a
te
d
ove
r
e
a
c
h
e
nt
it
y
o
f
T
O
M
L
knowle
dge
b
a
s
e
a
nd
c
om
put
e
d
it
s
c
or
r
e
s
ponde
nt
s
e
m
a
nt
ic
ve
c
to
r
by
us
in
g
our
pr
opos
e
d
a
lg
or
it
hm
s
.
F
in
a
ll
y,
th
e
ont
ol
ogy
w
a
s
c
a
r
e
f
ul
ly
c
he
c
k
e
d
by
bot
h
th
e
e
xpe
r
ti
s
e
s
a
nd
th
e
e
ngi
ne
e
r
s
in
or
de
r
to
r
e
a
c
h
it
s
f
ir
s
t
ve
r
s
io
n. A
n e
xc
e
r
pt
of
T
O
M
L
i
s
s
how
n i
n F
ig
ur
e
2.
F
ig
ur
e
2
. A
n e
xc
e
r
pt
of
T
O
M
L
ont
ol
ogy
I
n
ge
ne
r
a
l,
T
O
M
L
ha
s
157
c
onc
e
pt
s
,
65
obj
e
c
t
pr
ope
r
ti
e
s
a
nd
24
da
ta
pr
ope
r
ti
e
s
.
D
ue
to
th
e
s
e
l
a
r
ge
num
be
r
s
of
c
onc
e
pt
s
a
nd
pr
ope
r
ti
e
s
,
w
e
d
e
s
c
r
ib
e
T
O
M
L
by
s
u
m
m
a
r
iz
in
g
it
s
c
ha
r
a
c
te
r
is
ti
c
s
a
nd
hi
ghl
ig
ht
our
ow
n
c
ont
r
ib
ut
io
n
in
s
pe
c
if
yi
ng
th
e
to
ur
is
m
dom
a
in
knowle
dg
e
.
F
ir
s
tl
y,
w
e
de
ve
lo
p
c
onc
e
pt
s
th
a
t
r
e
la
te
to
to
ur
is
t,
pl
a
c
e
,
s
e
r
vi
c
e
,
f
a
c
il
it
y
a
nd
a
c
ti
vi
ty
.
F
or
e
xa
m
pl
e
,
th
e
c
onc
e
pt
to
m
l:
T
our
is
t
is
in
he
r
it
e
d
f
r
om
f
oa
f
:P
e
r
s
on
c
onc
e
pt
a
nd
h
a
s
th
r
e
e
di
f
f
e
r
e
nt
obj
e
c
t
pr
ope
r
ti
e
s
w
it
h
to
m
l:
C
it
y
c
onc
e
pt
in
c
lu
di
ng
to
m
l:
ha
s
H
om
e
T
ow
n,
to
m
l:
vi
s
it
e
d
a
nd
to
m
l:
vi
s
it
s
.
T
he
to
m
l:
T
our
is
t
c
onc
e
pt
pl
a
ys
th
e
ke
y
r
ol
e
of
ou
r
ont
ol
ogy
in
c
a
pt
ur
in
g
th
e
knowle
dge
a
bout
to
ur
is
t’
s
pe
r
s
on
a
l
in
f
or
m
a
ti
on
(
e
.g.,
ge
nde
r
,
a
nd
na
m
e
)
,
to
ur
is
ts
’
pr
e
f
e
r
e
nc
e
s
th
r
ough the
r
e
la
ti
on w
it
h t
r
a
ve
l:
A
c
ti
vi
ty
a
nd i
ts
s
ubc
onc
e
pt
s
.
S
e
c
ondl
y,
w
e
e
l
a
bor
a
te
a
nd
s
pe
c
if
y
m
or
e
c
onc
e
pt
s
a
bout
t
our
is
t’
s
a
c
ti
vi
ty
li
ke
to
m
l:
P
ur
c
ha
s
e
,
to
m
l:
L
is
te
n
or
to
m
l:
F
e
s
ti
va
l
to
na
m
e
a
f
e
w
s
.
T
he
s
e
a
c
ti
vi
ty
c
onc
e
pt
s
a
r
e
e
f
f
ic
ie
nt
in
c
a
pt
ur
in
g
to
u
r
is
t’
s
pr
e
f
e
r
e
nc
e
s
.
A
nd
th
e
y
a
r
e
us
e
d
in
th
e
f
ir
s
t
pha
s
e
of
th
e
r
e
c
om
m
e
nda
ti
on
pr
oc
e
s
s
by
li
nki
ng
w
it
h
ot
he
r
c
onc
e
pt
s
t
hr
ough tom
l:
s
ugge
s
t
obj
e
c
t
pr
ope
r
ty
.
T
hi
r
dl
y,
e
ve
r
y
s
ub
-
c
onc
e
pt
of
to
m
l:
P
la
c
e
,
to
m
l:
P
r
oduc
ts
or
to
m
l:
S
e
r
vi
c
e
h
a
s
r
e
la
ti
on
w
it
h
to
m
l:
S
e
m
a
nt
ic
V
e
c
to
r
c
onc
e
pt
.
T
hi
s
c
onc
e
pt
pr
ovi
de
s
th
e
qu
a
nt
it
a
ti
ve
ve
c
to
r
f
or
e
ve
r
y
e
nt
it
y
of
th
e
r
e
la
te
d
c
onc
e
pt
.
T
hi
s
ve
c
to
r
is
th
e
ba
s
e
f
or
a
ny
f
ur
th
e
r
us
e
of
m
a
c
hi
ne
le
a
r
ni
ng
m
ode
ls
or
de
c
is
io
n
m
a
ki
ng
pr
oc
e
s
s
.
W
e
pr
opos
e
s
p
e
c
if
ic
a
lg
o
r
it
hm
s
t
o buil
d s
e
m
a
nt
ic
ve
c
to
r
s
f
or
e
v
e
r
y r
e
la
te
d e
nt
it
y of
T
O
M
L
knowle
dge
ba
s
e
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
.
10
, N
o.
4
,
D
e
c
e
m
be
r
202
1
:
1060
-
1068
1064
F
in
a
ll
y,
w
e
pr
opos
e
th
e
to
m
l:
R
e
c
om
I
te
m
c
onc
e
pt
to
c
a
pt
ur
e
o
ne
or
m
or
e
r
e
c
om
m
e
nde
d
th
in
gs
.
F
or
e
xa
m
pl
e
,
in
c
a
s
e
th
a
t
to
ur
is
ts
pr
e
f
e
r
to
buy
p
r
oduc
ts
,
a
nd
th
e
p
r
oduc
ts
a
r
e
f
oun
d
in
a
lo
c
a
l
m
a
r
ke
t
w
he
r
e
it
is
r
e
qui
r
e
d
to
us
e
th
e
publ
ic
tr
a
ns
por
t
s
e
r
vi
c
e
to
go
to
,
th
e
r
e
c
om
m
e
nde
d
it
e
m
s
f
or
to
ur
is
ts
s
houl
d
ta
ke
a
c
c
ount
of
not
onl
y
th
e
pr
oduc
t
it
s
e
lf
but
a
l
s
o
th
e
a
va
il
a
bl
e
publ
ic
tr
a
ns
por
t
s
e
r
vi
c
e
a
nd
r
out
e
gui
de
.
T
hi
s
i
s
th
e
di
f
f
e
r
e
nt
c
ha
r
a
c
te
r
is
ti
c
of
to
u
r
is
m
r
e
c
om
m
e
nda
ti
on
in
c
om
pa
r
is
on
w
it
h
ot
he
r
ki
nds
o
f
r
e
c
om
m
e
nde
r
s
li
ke
books
or
m
ovi
e
s
.
3.2. E
n
r
ic
h
in
g T
O
M
L
k
n
ow
le
d
ge
b
as
e
I
n
or
de
r
to
popula
te
th
e
T
O
M
L
knowle
dge
ba
s
e
,
w
e
im
por
te
d
r
e
le
va
nt
da
ta
f
r
om
ope
n
li
nke
d
da
ta
s
our
c
e
s
(
e
.
g., D
B
pe
di
a
)
a
nd l
oc
a
l
da
ta
ba
s
e
s
t
o t
he
T
O
M
L
know
le
dge
ba
s
e
.
T
he
i
m
por
ti
ng pr
oc
e
s
s
de
pe
nd
s
on
th
e
m
a
ppi
ng
m
e
th
ods
of
c
la
s
s
a
nd
pr
ope
r
ty
.
I
n
w
hi
c
h,
c
or
r
e
s
p
onde
nt
c
onc
e
pt
s
be
twe
e
n
D
bp
e
di
a
a
nd
T
O
M
L
w
e
r
e
f
ig
ur
e
d
out
.
S
im
il
a
r
ly
,
th
e
m
a
ppi
ng
r
ul
e
s
be
twe
e
n
da
ta
b
a
s
e
ta
bl
e
s
a
nd
T
O
M
L
ont
ol
ogy
w
e
r
e
de
f
in
e
d.
T
he
n,
r
e
le
va
nt
D
B
p
e
di
a
e
nt
it
ie
s
a
nd
th
e
ir
pr
ope
r
ti
e
s
w
e
r
e
s
e
le
c
te
d
by
S
P
A
R
Q
L
que
r
ie
s
a
nd
w
e
r
e
e
xpor
te
d
to
R
D
F
/J
S
O
N
f
or
m
a
t.
I
n c
a
s
e
of
i
nt
e
gr
a
ti
ng l
oc
a
l
da
ta
ba
s
e
s
i
nt
o
T
O
M
L
, t
he
r
e
le
va
nt
t
a
bl
e
r
e
c
or
ds
w
e
r
e
s
e
l
e
c
te
d
by
a
nd
e
xpor
te
d
to
R
D
F
/
J
S
O
N
f
il
e
s
.
F
in
a
ll
y,
th
e
s
e
ba
tc
h
f
il
e
s
w
e
r
e
im
por
te
d
di
r
e
c
tl
y
to
th
e
T
O
M
L
knowle
dge
ba
s
e
.
T
he
p
s
e
udo
c
od
e
s
of
im
por
ti
ng
da
ta
f
r
om
D
B
pe
di
a
a
nd
lo
c
a
l
da
ta
ba
s
e
a
r
e
s
how
n
in
F
ig
ur
e
s
3 a
nd 4, r
e
s
pe
c
ti
ve
ly
.
F
ig
ur
e
3. P
s
e
udo
-
a
lg
or
it
hm
of
r
e
t
r
ie
vi
ng r
e
le
va
nt
knowle
dge
f
r
om
D
B
pe
di
a
F
ig
ur
e
4. P
s
e
udo
-
a
lg
or
it
hm
of
popula
ti
ng T
O
M
L
ont
ol
ogy by l
oc
a
l
da
ta
ba
s
e
T
he
pr
im
a
r
y
pur
pos
e
of
T
O
M
L
knowle
dge
b
a
s
e
is
to
pr
ovi
d
e
da
ta
f
or
m
a
c
hi
ne
le
a
r
ni
ng
m
ode
ls
.
W
hi
le
m
a
c
hi
n
e
le
a
r
ni
ng
m
ode
ls
r
e
qui
r
e
in
put
s
a
s
num
e
r
ic
a
l
ve
c
to
r
s
,
ope
n
li
nke
d
da
ta
(
e
.g.,
D
B
pe
di
a
,
T
O
M
L
knowle
dge
ba
s
e
)
pr
ovi
de
da
ta
unde
r
gr
a
ph
-
ba
s
e
d
f
or
m
a
ts
(
e
.
g.,
R
D
F
,
O
W
L
)
.
W
e
tr
a
ns
f
e
r
r
e
d
th
e
pr
ope
r
ty
va
lu
e
of
a
n
e
nt
it
y
by
us
in
g
(
1
)
.
T
h
e
n,
our
s
ol
ut
io
n
to
bui
ld
in
g
num
e
r
ic
a
l
ve
c
to
r
s
ba
s
e
d
on
a
v
a
il
a
bl
e
li
nke
d
da
ta
f
or
e
ve
r
y
T
O
M
L
’
s
e
nt
it
y
a
ppl
ie
d
(
1
)
in
ps
e
udo
-
a
lg
or
it
hm
of
F
ig
ur
e
5.
E
a
c
h
pr
ope
r
ty
of
th
e
e
nt
it
y
now
pl
a
ys
t
he
r
ol
e
of
a
di
m
e
ns
io
n i
n t
he
s
e
m
a
nt
ic
ve
c
to
r
.
=
l
o
g
(
1
<
,
,
>
+
1
)
(
1)
w
he
r
e
<
,
,
>
tr
ip
le
s
i
s
t
he
t
ot
a
l
num
be
r
of
t
r
ip
le
s
w
hi
c
h ha
ve
t
he
s
a
m
e
s
ubj
e
c
t
c
onc
e
pt
(
c
la
s
s
)
-
c
, t
he
s
a
m
e
pr
ope
r
ty
-
p
a
nd t
he
s
a
m
e
pr
ope
r
ty
va
lu
e
-
e
.
B
y
im
pl
e
m
e
nt
in
g
th
e
a
lg
or
it
hm
s
how
n
in
F
ig
ur
e
5,
e
ve
r
y
e
nt
it
y
ha
s
it
s
ow
n
s
e
m
a
nt
ic
v
e
c
to
r
,
how
e
ve
r
,
s
om
e
pr
ope
r
ti
e
s
m
a
y
a
ppe
a
r
or
not
in
di
f
f
e
r
e
nt
e
nt
it
ie
s
.
I
n
ot
he
r
w
or
ds
,
di
f
f
e
r
e
nt
e
nt
it
ie
s
m
a
y
ha
ve
di
f
f
e
r
e
nt
ve
c
to
r
s
pa
c
e
s
.
T
he
r
e
f
or
e
,
bui
ld
in
g
th
e
c
om
m
on
ve
c
to
r
s
pa
c
e
f
or
a
ll
s
e
le
c
te
d
s
e
m
a
nt
ic
ve
c
to
r
s
is
ne
c
e
s
s
a
r
y. F
ir
s
tl
y
, a
ll
s
e
m
a
nt
ic
ve
c
to
r
s
r
e
la
te
d t
o t
he
r
e
c
om
m
e
n
da
ti
on t
a
s
k a
r
e
s
e
l
e
c
te
d by S
P
A
R
Q
L
S
E
L
E
C
T
que
r
y. T
he
n, a
ll
of
th
e
di
s
ti
nc
t
pr
ope
r
t
ie
s
a
r
e
f
ig
ur
e
d
out
a
nd a
r
e
s
or
te
d i
n a
s
c
e
ndi
ng or
de
r
o
f
pr
ope
r
ty
na
m
e
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
A
nov
e
l
ont
ol
ogy
f
r
am
e
w
or
k
s
uppo
r
ti
ng m
ode
l
-
bas
e
d t
our
i
s
m
r
e
c
om
m
e
nde
r
(
H
o Q
uoc
D
ung
)
1065
T
he
s
e
a
r
e
th
e
di
m
e
n
s
io
ns
of
th
e
ve
c
to
r
s
pa
c
e
.
F
in
a
ll
y,
f
or
e
a
c
h
s
e
m
a
nt
ic
ve
c
to
r
,
it
s
or
ig
in
a
l
va
lu
e
s
a
r
e
f
il
le
d
pr
ope
r
ly
in
to
c
or
r
e
s
ponding
di
m
e
ns
io
ns
.
T
h
e
r
e
s
t
of
di
m
e
ns
io
ns
,
w
hi
c
h
a
r
e
not
f
il
le
d,
r
e
c
e
iv
e
z
e
r
o
va
lu
e
s
.
T
hi
s
pr
oc
e
dur
e
i
s
e
xpr
e
s
s
e
d i
n F
ig
ur
e
6.
F
ig
ur
e
3
.
P
s
e
udo
-
a
lg
or
it
hm
of
c
a
lc
ul
a
ti
ng s
e
m
a
nt
ic
ve
c
to
r
3.3. T
O
M
L
-
b
as
e
d
r
e
c
o
m
m
e
n
d
at
io
n
e
n
gi
n
e
T
O
M
L
-
ba
s
e
d
r
e
c
om
m
e
nda
ti
on
s
tr
a
te
gi
e
s
w
e
r
e
de
s
ig
n
e
d
to
c
ope
w
it
h
th
e
two
popula
r
r
e
c
om
m
e
nda
ti
on
c
a
s
e
s
:
(
i)
w
it
h
th
e
a
v
a
il
a
bi
li
ty
of
to
ur
is
t
pr
e
f
e
r
e
nc
e
da
ta
;
a
nd
(
ii
)
w
it
hout
th
e
a
va
il
a
bi
li
ty
of
to
ur
is
t
pr
e
f
e
r
e
nc
e
da
ta
.
I
n
c
a
s
e
th
a
t
th
e
to
ur
is
t
pr
e
f
e
r
e
nc
e
da
ta
is
not
a
va
il
a
bl
e
,
th
e
r
e
c
om
m
e
nd
a
ti
on
s
tr
a
te
gy
is
a
s
:
A
s
s
um
in
g
th
a
t
to
ur
is
ts
w
a
nt
to
ge
t
a
to
p
-
K
r
e
c
om
m
e
nda
ti
on
li
s
t
a
bout
a
gi
ve
n c
onc
e
pt
(
e
.g.
pl
a
c
e
,
f
ood
o
r
pr
oduc
t)
.
F
ir
s
t,
a
n
e
nt
it
y
r
e
la
ti
ng
to
th
e
r
e
c
om
m
e
nd
e
d
c
onc
e
pt
is
r
a
ndoml
y
s
e
le
c
te
d
vi
a
a
S
P
A
R
Q
L
S
E
L
E
C
T
que
r
y. T
hi
s
e
nt
it
y
s
houl
d
be
s
pe
c
if
ie
d a
s
“
f
a
m
ous
”
in
t
he
knowle
dge
ba
s
e
.
W
e
us
e
th
is
e
nt
it
y
a
s
th
e
s
ta
r
ti
ng
poi
nt
a
nd
f
in
d
ot
he
r
(k
-
1)
ne
a
r
e
s
t
e
nt
it
i
e
s
by
c
a
lc
ul
a
ti
ng
s
e
m
a
nt
ic
s
im
il
a
r
it
y
be
twe
e
n
th
is
e
nt
it
y
a
nd
th
e
ot
he
r
e
nt
it
ie
s
w
it
hi
n
th
e
s
a
m
e
c
onc
e
pt
.
T
he
E
uc
li
de
a
n
di
s
ta
nc
e
is
a
c
c
e
pt
e
d
to
c
om
put
e
s
e
m
a
nt
ic
s
im
il
a
r
it
y. T
he
ps
e
udoc
od
e
of
t
hi
s
s
tr
a
te
gy i
s
s
how
n i
n F
ig
ur
e
7.
F
ig
ur
e
6. P
s
e
udo
-
al
gor
it
hm
of
c
ons
tr
uc
ti
ng c
om
m
on
ve
c
to
r
s
pa
c
e
F
ig
ur
e
7. P
s
e
udo
-
a
lg
or
it
hm
of
t
op
-
K
r
e
c
om
m
e
nda
ti
ons
ba
s
e
d on s
e
m
a
nt
ic
s
im
il
a
r
it
y
I
n c
a
s
e
t
ha
t
to
ur
is
ts
pr
ovi
de
pr
e
f
e
r
e
nc
e
da
ta
f
or
c
r
e
a
ti
ng l
a
be
le
d
da
ta
, t
he
s
upe
r
vi
s
e
d l
e
a
r
ni
ng mode
ls
a
r
e
a
ppl
ie
d
to
ge
ne
r
a
te
to
p
-
K
pe
r
s
ona
li
z
e
d
r
e
c
om
m
e
nda
ti
on
i
te
m
s
.
D
if
f
e
r
e
nt
c
la
s
s
if
ic
a
ti
on
m
ode
ls
c
a
n
be
pl
ugge
d
in
to
th
e
r
e
c
om
m
e
nda
ti
on
e
ngi
ne
vi
a
pa
r
a
m
e
te
r
in
put
.
A
nd
th
e
p
r
e
di
c
ti
on
s
c
or
e
s
w
e
r
e
us
e
d
to
r
a
nk
th
e
t
op
-
K
r
e
c
om
m
e
nda
ti
on l
is
t.
F
ig
ur
e
8 s
how
s
t
he
ps
e
udo
c
ode
of
t
hi
s
s
tr
a
te
gy.
B
a
s
e
d
on
th
e
to
p
-
K
r
e
c
om
m
e
nda
ti
on
li
s
t,
w
hi
c
h
is
ge
ne
r
a
te
d
by
a
lg
or
it
hm
s
in
e
it
he
r
F
ig
ur
e
7
or
F
ig
ur
e
8,
th
e
r
ou
te
pl
a
nni
ng
a
lg
or
it
hm
is
a
ppl
ie
d
to
f
in
d
th
e
s
h
or
te
s
t
pa
th
f
r
om
th
e
to
ur
is
t
’
s
c
ur
r
e
nt
lo
c
a
ti
on
to
a
ll
lo
c
a
ti
ons
of
k
s
ugg
e
s
te
d
it
e
m
s
.
T
he
lo
c
a
ti
on
d
a
ta
w
a
s
s
to
r
e
d
in
T
O
M
L
knowle
dge
ba
s
e
a
nd
G
oogl
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
.
10
, N
o.
4
,
D
e
c
e
m
be
r
202
1
:
1060
-
1068
1066
m
a
p
A
P
I
w
a
s
us
e
d
to
f
in
d
th
e
lo
c
a
ti
on
-
to
-
lo
c
a
ti
on
r
out
e
.
F
ig
u
r
e
9
s
how
s
th
e
ps
e
udoc
ode
of
r
out
e
pl
a
nni
ng
r
e
c
om
m
e
nda
ti
on.
F
ig
ur
e
8. P
s
e
udo
-
a
lg
or
it
hm
of
t
op
-
K
r
e
c
om
m
e
nda
ti
ons
ge
ne
r
a
te
d by c
l
a
s
s
if
ie
r
s
F
ig
ur
e
9. P
s
e
udo
-
a
lg
or
it
hm
of
r
ou
te
pl
a
nni
ng
r
e
c
om
m
e
nda
ti
on
4.
E
X
P
E
R
I
M
E
N
T
S
T
he
e
xpe
r
im
e
nt
s
w
e
r
e
c
onduc
te
d
to
e
va
lu
a
te
th
e
e
f
f
ic
ie
nc
y
of
T
O
M
L
knowle
dge
ba
s
e
a
nd
it
s
r
e
c
om
m
e
nda
ti
on
e
ngi
ne
.
W
e
de
ve
lo
pe
d
a
pr
ot
ot
ype
in
P
yt
hon
pr
ogr
a
m
m
in
g
la
ngua
ge
w
hi
c
h
im
pl
e
m
e
nt
s
a
ll
of
th
e
a
lg
or
i
th
m
s
pr
opos
e
d
in
s
e
c
ti
on
3.1.
T
he
te
s
ts
of
us
e
r
s
a
ti
s
f
a
c
ti
on
a
nd
th
e
f
e
a
s
ib
il
it
y
o
f
im
p
le
m
e
nt
in
g
m
a
c
hi
ne
le
a
r
ni
ng
m
ode
ls
w
it
h
T
O
M
L
knowle
dge
ba
s
e
w
e
r
e
pr
e
s
e
nt
e
d
in
s
ubs
e
c
ti
ons
4.1
a
nd
4.2,
r
e
s
pe
c
ti
ve
ly
.
4.1. E
xp
e
r
im
e
n
t
1:
b
u
il
d
in
g t
op
-
K
r
e
c
om
m
e
n
d
at
io
n
l
is
t
w
it
h
ou
t
u
s
e
r
p
r
e
f
e
r
e
n
c
e
I
n
th
is
e
xpe
r
im
e
nt
,
to
ur
is
ts
’
pr
e
f
e
r
e
nc
e
da
ta
is
not
a
va
il
a
bl
e
.
T
hi
s
s
it
ua
ti
on
c
a
us
e
s
th
e
c
ol
d
-
s
ta
r
t
pr
obl
e
m
of
th
e
r
e
c
om
m
e
nda
ti
on
r
e
s
e
a
r
c
h
f
ie
ld
.
I
n
r
e
a
l
w
or
d,
th
e
to
ur
gui
de
s
of
te
n
pr
ovi
de
s
ugge
s
ti
on
s
w
it
hout
ha
vi
ng
to
ur
is
ts
’
pr
e
f
e
r
e
nc
e
s
.
T
he
r
e
f
or
e
,
w
e
de
c
id
e
to
c
om
pa
r
e
th
e
to
p
-
K
r
e
c
om
m
e
nda
ti
on
li
s
t
s
yi
e
ld
e
d by T
O
M
L
-
ba
s
e
d pr
ot
ot
ype
t
o t
hos
e
of
t
our
gui
de
s
.
T
he
e
xp
e
r
im
e
nt
w
a
s
de
s
ig
n
e
d
a
s
:
Q
ue
s
ti
onna
ir
e
s
w
e
r
e
s
e
nt
to
to
ur
gui
de
s
of
5
di
f
f
e
r
e
nt
lo
c
a
l
to
u
r
is
t
c
om
pa
ni
e
s
.
T
h
e
s
ur
ve
y
c
lo
s
e
d
a
f
te
r
2
m
ont
hs
a
nd
th
e
r
e
w
e
r
e
3
2
to
ur
gui
de
s
w
ho
c
om
pl
e
te
d
th
e
s
ur
ve
y.
T
he
to
ur
gui
de
s
w
e
r
e
a
s
ke
d t
o gi
ve
t
op
-
10 r
e
c
om
m
e
nda
ti
ons
f
or
f
oo
ds
, pl
a
c
e
s
a
nd pr
oduc
t
s
of
a
gi
ve
n c
it
y. D
ue
t
o
th
e
c
om
pl
e
xi
ty
of
c
ol
le
c
te
d
da
ta
,
w
e
s
um
m
a
r
iz
e
d
th
e
e
xp
e
r
im
e
nt
a
l
r
e
s
ul
ts
in
T
a
bl
e
1.
T
o
be
m
or
e
s
pe
c
if
ic
,
to
p
-
10
pl
a
c
e
r
e
c
om
m
e
nda
ti
ons
ge
ne
r
a
te
d
by
bot
h
T
O
M
L
a
nd
to
ur
gui
de
s
a
r
e
vi
s
ua
li
z
e
d
in
F
ig
u
r
e
10
f
o
r
be
tt
e
r
unde
r
s
ta
ndi
ng of
t
he
r
e
c
om
m
e
nde
d r
e
s
ul
ts
.
A
s
s
how
n
in
T
a
bl
e
1,
a
ll
of
th
e
p
-
va
lu
e
s
of
th
r
e
e
di
f
f
e
r
e
nt
gr
oups
of
to
p
-
10
r
e
c
om
m
e
nda
ti
ons
a
r
e
gr
e
a
te
r
th
a
n
0.05.
T
he
s
e
s
ta
ti
s
ti
c
a
l
r
e
s
ul
ts
im
pl
y
th
a
t
th
e
r
e
is
n
o
di
f
f
e
r
e
nc
e
in
pe
r
s
ona
li
z
e
d
r
e
c
om
m
e
nda
ti
on
li
s
ts
be
tw
e
e
n
to
ur
gui
de
s
a
nd
T
O
M
L
-
ba
s
e
d
pr
ot
ot
ype
.
I
n
o
th
e
r
w
or
ds
,
th
e
T
O
M
L
-
ba
s
e
d
pr
ot
ot
ype
c
a
n
pr
ovi
de
s
ugge
s
ti
ons
a
s
good
a
s
th
o
s
e
of
to
ur
gui
de
s
.
F
ur
th
e
r
m
or
e
,
th
e
s
ta
ti
s
ti
c
a
l
r
e
s
ul
ts
a
ls
o
in
di
c
a
te
th
a
t
T
O
M
L
knowle
dge
ba
s
e
ha
s
c
a
pt
ur
e
d e
xp
e
r
ts
’
doma
in
knowle
d
ge
e
f
f
ic
ie
nt
ly
.
T
a
bl
e
1. p
-
va
lu
e
s
r
e
s
ul
ts
of
t
op
-
10 r
e
c
om
m
e
nda
ti
ons
T
ot
a
l
r
e
c
om
m
e
nde
d i
t
e
m
s
p
-
va
l
ue
s
T
O
M
L
T
our
G
ui
de
s
P
l
a
c
e
10
15
0.57
F
ood
10
17
0.46
P
r
oduc
t
10
14
0.63
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
A
nov
e
l
ont
ol
ogy
f
r
am
e
w
or
k
s
uppo
r
ti
ng m
ode
l
-
bas
e
d t
our
i
s
m
r
e
c
om
m
e
nde
r
(
H
o Q
uoc
D
ung
)
1067
F
ig
ur
e
10. T
op
-
10 pla
c
e
s
r
e
c
om
m
e
nde
d by
T
O
M
L
a
nd by tour
gui
de
s
4.2. E
xp
e
r
im
e
n
t
2:
I
m
p
le
m
e
n
t
in
g var
io
u
s
c
la
s
s
if
i
e
r
s
w
it
h
T
O
M
L
k
n
ow
le
d
ge
b
as
e
T
he
pur
pos
e
of
th
is
e
xpe
r
im
e
nt
is
to
de
m
ons
tr
a
te
th
e
a
bi
li
ty
o
f
T
O
M
L
knowle
dge
ba
s
e
in
te
r
m
s
of
pr
ovi
di
ng
da
ta
f
or
m
a
c
hi
ne
le
a
r
ni
ng
m
ode
l
s
.
S
pe
c
if
ic
a
ll
y,
la
be
l
e
d
da
ta
is
r
e
qui
r
e
d
to
tr
a
in
s
upe
r
vi
s
e
d
le
a
r
ni
ng
m
ode
ls
f
or
pr
e
di
c
ti
ng w
hi
c
h e
nt
it
y s
houl
d be
pr
e
s
e
nt
e
d t
o t
our
is
ts
. H
ow
e
ve
r
, i
t
w
a
s
h
a
r
d t
o a
s
k t
our
is
ts
t
o j
oi
n
th
is
e
xpe
r
im
e
nt
.
H
e
nc
e
,
w
e
in
vi
te
d
to
ur
gui
de
s
jo
in
in
g
th
is
e
xpe
r
im
e
nt
unde
r
th
e
r
ol
e
o
f
to
u
r
is
ts
.
E
a
c
h
pa
r
ti
c
ip
a
nt
f
ig
ur
e
d
out
w
hi
c
h
th
in
gs
(
pl
a
c
e
s
,
f
oods
,
a
nd
pr
oduc
ts
.)
s
he
or
he
li
ke
s
or
di
s
li
ke
s
.
T
he
s
e
pr
e
f
e
r
e
nc
e
s
w
e
r
e
upda
te
d
to
th
e
c
or
r
e
s
ponde
nt
e
nt
it
ie
s
in
T
O
M
L
knowle
dge
ba
s
e
vi
a
th
e
da
ta
pr
ope
r
ty
to
m
l:
ha
s
P
r
e
f
e
r
e
nc
e
.
T
hi
s
pr
ope
r
ty
w
a
s
a
ls
o
a
dde
d
to
th
e
s
e
m
a
nt
ic
ve
c
to
r
a
s
th
e
la
be
l
di
m
e
n
s
io
n.
T
h
e
pr
e
f
e
r
e
nc
e
da
ta
w
e
r
e
a
s
s
oc
ia
t
e
d w
it
h t
he
c
onc
e
pt
t
om
l:
T
our
is
t
w
hi
c
h c
a
pt
ur
e
s
t
our
is
t
pr
of
il
e
s
.
T
he
r
e
w
e
r
e
6
to
ur
gui
de
s
w
ho
pa
r
ti
c
ip
a
te
d
in
th
is
e
xpe
r
im
e
nt
a
nd
c
ons
tr
uc
te
d
327
r
e
c
or
ds
of
th
e
i
r
pr
e
f
e
r
e
nc
e
s
.
W
e
us
e
d
th
r
e
e
popula
r
c
la
s
s
if
ic
a
ti
on
m
ode
ls
in
c
lu
di
ng
k
-
N
N
,
N
a
iv
e
B
a
ye
s
a
nd
S
V
M
to
pr
e
di
c
t
pe
r
s
ona
li
z
e
d
to
ur
is
m
r
e
c
om
m
e
nda
ti
ons
.
T
h
e
r
e
w
e
r
e
th
r
e
e
s
u
gge
s
ti
on
li
s
ts
a
bout
pl
a
c
e
,
f
ood
a
nd
pr
oduc
t.
E
a
c
h
pa
r
ti
c
ip
a
nt
e
va
lu
a
te
d
on
e
ve
r
y
s
ugge
s
t
e
d
it
e
m
th
a
t
s
h
e
or
he
s
a
ti
s
f
ie
d
or
not
.
T
a
bl
e
2
s
ho
w
s
th
i
s
e
xpe
r
im
e
nt
r
e
s
ul
ts
.
I
t
is
im
por
ta
nt
to
e
m
pha
s
iz
e
th
a
t
th
is
e
xpe
r
im
e
nt
doe
s
not
ta
r
ge
t
a
t
in
tr
oduc
in
g
ne
w
c
la
s
s
if
ie
r
s
w
it
h
hi
ghl
y
pr
e
di
c
te
d
c
a
p
a
bi
li
ti
e
s
but
d
e
m
ons
tr
a
ti
ng
th
a
t
m
a
c
hi
ne
le
a
r
ni
ng
m
ode
ls
c
a
n
w
or
k
w
e
ll
di
r
e
c
tl
y
w
it
h
T
O
M
L
knowle
dge
ba
s
e
.
A
s
in
di
c
a
te
d
in
T
a
bl
e
2,
th
e
a
ve
r
a
ge
s
of
s
a
ti
s
f
ie
d
r
a
ti
os
r
a
nge
f
r
om
40%
to
63.3%
,
w
hi
le
th
os
e
of
uns
a
ti
s
f
ie
d
r
a
ti
os
r
a
nge
f
r
om
36.7%
to
56.7
%
.
T
he
th
r
e
e
tr
a
di
ti
ona
l
c
la
s
s
if
ie
r
s
r
e
a
c
h
a
nd
ove
r
c
om
e
th
e
50%
th
r
e
s
hol
d
6
ti
m
e
s
in
to
ta
l.
T
he
s
e
r
e
s
ul
ts
c
onf
ir
m
th
e
e
f
f
ic
ie
nc
y
us
a
ge
of
T
O
M
L
-
ba
s
e
d
s
e
m
a
nt
ic
ve
c
to
r
s
i
n m
a
c
hi
ne
l
e
a
r
ni
ng mode
ls
.
T
a
bl
e
1
. U
s
e
r
s
r
e
a
c
t
to
r
e
c
om
m
e
nda
ti
on l
is
t
s
ge
ne
r
a
te
d by ma
c
hi
ne
l
e
a
r
ni
ng mode
ls
U
s
e
r
1
U
s
e
r
2
U
s
e
r
3
U
s
e
r
4
U
s
e
r
5
U
s
e
r
6
A
ve
r
a
ge
s
a
t
i
s
f
y
not
s
a
t
i
s
f
y
not
s
a
t
i
s
f
y
not
s
a
t
i
s
f
y
not
s
a
t
i
s
f
y
not
s
a
t
i
s
f
y
not
s
a
t
i
s
f
y
not
P
l
a
c
e
S
V
M
5
5
4
6
6
4
5
5
6
4
7
3
5.5
4.5
N
a
ï
ve
B
a
ye
s
3
7
4
5
4
5
5
5
2
8
6
4
4
5.67
k
-
NN
4
6
4
6
7
3
6
4
3
7
3
7
4.5
5.5
F
ood
S
V
M
7
3
5
5
4
6
7
3
8
2
7
3
6.33
3.67
N
a
ï
ve
B
a
ye
s
6
4
6
4
2
8
4
6
6
4
3
7
4.5
5.5
k
-
NN
6
4
9
1
3
7
7
3
6
4
6
4
6.17
3.83
P
r
oduc
t
S
V
M
1
9
9
1
7
3
1
9
5
5
7
3
5
5
N
a
ï
ve
B
a
ye
s
2
8
4
6
8
2
6
4
5
5
8
2
5.5
4.5
k
-
NN
5
5
4
6
1
9
8
2
5
5
8
2
5.17
4.83
5.
C
O
N
C
L
U
S
I
O
N
I
n
th
is
pa
pe
r
,
a
n
ont
ol
ogi
c
a
l
f
r
a
m
e
w
or
k
s
uppor
ti
ng
m
ode
l
-
ba
s
e
d
to
ur
is
m
r
e
c
om
m
e
nde
r
s
o
-
c
a
ll
e
d
T
O
M
L
ha
s
be
e
n
pr
e
s
e
nt
e
d.
T
he
r
e
c
om
m
e
nda
ti
on
m
a
ki
ng
pr
oc
e
s
s
m
a
ke
s
u
s
e
of
th
e
s
tr
e
ngt
h
of
bot
h
S
e
m
a
nt
ic
W
e
b
te
c
hnol
ogy
a
nd
s
upe
r
vi
s
e
d
le
a
r
ni
ng
m
ode
ls
.
E
s
pe
c
ia
ll
y
th
e
de
s
ig
n
of
T
O
M
L
ont
ol
ogy
c
om
pos
e
s
of
a
s
pe
c
if
ic
s
e
m
a
nt
ic
v
e
c
to
r
c
onc
e
pt
w
hi
c
h
e
na
bl
e
s
m
a
c
hi
ne
le
a
r
ni
ng
m
ode
ls
to
c
ons
um
e
da
ta
di
r
e
c
tl
y
in
gr
a
ph
-
ba
s
e
d
s
tr
uc
tu
r
e
of
ont
ol
ogi
c
a
l
knowle
dge
ba
s
e
.
T
w
o
e
xp
e
r
im
e
nt
s
w
e
r
e
c
ondu
c
te
d
to
va
li
da
te
th
e
e
f
f
ic
ie
nc
y
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
.
10
, N
o.
4
,
D
e
c
e
m
be
r
202
1
:
1060
-
1068
1068
a
nd
th
e
pr
om
is
in
g
us
a
ge
of
T
O
M
L
-
ba
s
e
d
f
r
a
m
e
w
or
k.
W
hi
le
th
e
r
e
s
ul
ts
obt
a
in
e
d
f
r
om
e
xpe
r
im
e
nt
1
in
di
c
a
te
th
a
t
T
O
M
L
knowle
dge
ba
s
e
ha
s
c
a
pt
ur
e
d
e
xpe
r
ts
’
dom
a
in
knowle
dge
e
f
f
ic
ie
nt
ly
,
th
os
e
ga
in
e
d
f
r
om
e
xpe
r
im
e
nt
2
c
onf
ir
m
th
e
e
f
f
ic
ie
nc
y
us
a
ge
of
T
O
M
L
-
ba
s
e
d
s
e
m
a
nt
ic
v
e
c
to
r
s
in
m
a
c
hi
ne
le
a
r
ni
ng
m
ode
ls
.
T
he
f
ut
ur
e
w
or
k
of
th
is
s
tu
dy
w
il
l
f
oc
us
on
bui
ld
in
g
T
O
M
L
-
ba
s
e
d
w
e
b
s
e
r
vi
c
e
a
nd
in
te
gr
a
ti
ng
T
O
M
L
ont
ol
ogy
w
it
h
ot
he
r
to
ur
is
m
ont
ol
ogi
e
s
in
or
de
r
to
e
nl
a
r
ge
th
e
knowle
dge
ba
s
e
a
nd
bui
ld
in
g
T
O
M
L
-
ba
s
e
d
f
r
a
m
e
w
or
k a
s
a
ba
c
kbon
e
of
t
our
is
m
r
e
c
om
m
e
nda
ti
on s
e
r
vi
c
e
.
R
E
F
E
R
E
N
C
E
S
[1]
F.
M.
Harper
and
J.
A.
Konstan,
“The
movielens
datasets:
history
an
d
context,”
ACM
Transactions
on
Interactiv
e
Intelligen
t Systems
, vol. 5, no. 4, pp. 1
-
19, 2015, doi: 10.1145/2827872.
[2]
C.
-
N.
Ziegler,
S.
M.
McNee,
J.
A.
Konstan,
and
G.
Lausen,
“Improving
re
commendat
ion
lists
through
topic
diversification,”
WWW
'
05:
Proceedings
of
the
14th
International
Co
nference
on
World
Wide
Web
,
2005,
pp.
22
-
32, doi: 10.1145/1060745.1060754
[3]
C.
Bizer
et
al.
,
“DBpedia
-
a
crystallization
point
for
the
web
of
data,”
Journal
of
Web
Semantics
,
vol.
7,
no.
3,
pp.
154
-
165, Sep. 2009, doi: 10.1016/j.websem.2009.07.002.
[4]
C.
Grün,
J.
Neidhardt,
and
H.
Werthner,
“Ontology
-
based
matchmaki
ng
to
provide
personalized
recommendations
for tourist
s,” in
Informati
on and Co
mmunicatio
n
Techno
logies
in Tourism
2017
, 2017, pp. 3
-
16.
[5]
K.
Bagherifard,
M.
Rahmani,
M.
Nilashi,
and
V.
Rafe, “Performance
i
mprovement
for
recommender
systems
using
ontology,”
Telema
tics an
d Infor
matics
, vol. 34, no. 8, pp. 1772
-
1792, 2017, doi: 10.1016/j.tele.2017.0
8.008.
[6]
J.
Bobadilla,
F.
Ortega,
A.
Hernando,
and
A.
Gutiérrez,
“Recom
mender
systems
survey,”
Knowledge
-
Based
Systems
, vol. 46, pp. 109
-
132, Jul. 2013, doi: 10.1016/j.knosys.2013.03.012.
[7]
J.
Borràs,
A.
Moreno,
and
A.
Valls,
“Intelligent
tourism
recommender
s
ystems:
a
survey,”
Expert
Systems
with
Applications
, vol. 41, no. 16, pp. 7370
-
7389, Nov. 2014, doi: 10.1016/j.eswa.2014.06.007.
[8]
L.
Ren
and
W.
Wang,
“An
SVM
-
based
collaborative
filtering
approa
ch
for
top
-
n
web
services
recommendation,”
Future Generation Co
mputer Systems
, vol. 78, pp. 531
-
543, Jan. 2018, doi: 10.1016/j.future.2017.07.027.
[9]
D.
K.
Chae,
S.
C.
Lee,
S.
Y.
Lee,
and
S.
W.
Kim,
“On
identifying
k
-
n
earest
neighbors
in
neighborhood
models
fo
r
efficient
and
effective
collaborat
ive
filterin
g,”
Neurocom
pu
ting
,
vol.
278,
pp.
134
–
143,
Feb.
2018,
doi: 10.1016/j.neucom.2017.06.081.
[10]
T.
K.
Paradarami,
N.
D.
Bastian,
and
J.
L.
Wightman,
“A
hybrid
recommender
system
using
artificial
neural
networks,”
Expert Systems with Applicatio
ns
, vol. 83, pp. 300
-
313, Oct. 20
17, doi: 10.1016/j.eswa.2017.04.046.
[11]
S.
Linda
and
K.
K.
Bharadwaj,
“A
decision
tree
-
based
context
-
aware
recommender
system,”
in
Intelligen
t
Human
Computer
Interac
tion,
2018, pp. 293
-
305
,
doi: 10.1007/978
-
3
-
030
-
04021
-
5_27.
[12]
J.
Liu
et
al.
,
“Representing
condi
tional
preference
by
boosted
regres
sion
trees
for
re
commendation,”
Informati
on
Sciences
, vol. 327, pp. 1
-
20, Jan. 2016, doi: 10.1016/j.ins.2015.08.001.
[13]
M.
Nilashi,
O.
bin
Ibrahim,
N.
Ithnin,
and
N.
H.
Sarmin,
“A
multi
-
criteria
collaborat
ive
filterin
g
recom
mender
system
for
the
tourism
domain
using
E
xpectation
maximi
zation
(EM)
and
PCA
–
ANFIS,”
Electronic
Commerce
Research and Applications
, vol. 14, no. 6, pp. 542
-
562, Oct. 2015, doi: 10.1016/j.elerap.2015.08.004.
[14]
M.
Nilashi,
K.
Bagherifard,
M.
Rahmani,
and
V
.
Rafe,
“A
recommen
der
system
for
tourism
industry
using
cluster
ensemble
and
predicti
on
machine
learning
techniqu
es,”
Computers
a
nd
Industri
al
Engineering
,
vol.
109,
pp.
357
-
368, Jul. 2017, doi: 10.1016/j.cie.2017.05.016.
[15]
S.
Natarajan,
S.
Vairavasundaram,
S.
Natarajan,
and
A.
H.
Gando
mi,
“Resolving
data
sparsity
and
cold
start
problem
in
collaborative
filtering
recommender
system
using
linked
o
pen
d
ata,”
Expert
Systems
with
Applications.
,
vol. 149, Jul. 2020, doi: 10.1016/j.eswa.2020.113248.
[16]
T
.
B
e
r
ne
r
s
-
Le
e
,
J
.
H
e
n
dl
e
r
,
a
n
d
O.
L
a
ss
i
l
a
,
“
T
h
e
s
e
m
an
t
i
c
we
b
,
”
S
c
i
e
nt
i
f
i
c A
m
e
r
i
c
an
,
v
ol
.
2
84
,
n
o.
5
,
p
p
.
28
-
3
7
,
2
0
0
1.
[17]
N.
F.
Noy
and
D.
L.
McGuinness,
“Ontology
development
101:
A
gu
ide
to
creating
your
first
ontology,”
Stanford
Knowledge Systems
Laboratory
,
Tech
.
Rep
.
KSL
-
01
-
05, Mar
.
2001
.
[18]
M.
Fernández
-
López,
A.
Gómez
-
Pér
ez,
and
N.
Juristo,
“Methontolo
gy:
from
ontological
art
towards
ontological
engineerin
g,” AAAI
,
Tech
.
Rep
.
,1997
.
[19]
Y.
Sure,
S.
Staab,
and
R.
Studer,
“On
-
to
-
knowledge
methodology
(OT
KM),”
in
Handbook
on
ontologies,
Springer
,
2004, pp. 117
-
132.
[20]
M.
C.
Suarez
-
Figuero
a
et
al.
,
“NeOn
methodology
for
building
contextualized
ontology
networks,”
NeO
n
Deliverable D
, vol. 5, pp
. 4
-
1, 2008.
[21]
A.
Moreno,
A.
Valls,
D.
Isern,
L.
Marin,
and
J.
Borràs,
“SigTur
/E
-
destination:
ontology
-
based
personalized
recommendatio
n
of
tourism
and
leisure
activities
,”
Engineering
Appli
cations
of
Artificial
Intelligence
,
vol.
26,
no
.
1, pp. 633
-
651, Jan
. 2013, doi: 10.1016/j.engappai.2012.02.014.
[22]
L.
Shi,
F.
Lin,
T.
Yang,
J.
Qi,
W.
Ma,
and
S.
Xu,
“Context
-
based
on
tology
-
driven
recommendation
strategies
for
tourism in ubiquitou
s computing,”
Wireless Per
sonal
Communicat
ions
.
, vol. 76, no. 4, pp.
731
–
745, Ju
n. 2014,
doi:
10.1007/s11277
-
013
-
1550
-
9.
[23]
P.
Ferraro
and
G.
Lo
Re,
“Designing
ontology
-
d
riven
recommender
systems
for
tourism,”
in
Advanc
es
onto
the
Internet of
Things
: How Onto
logies Make t
he Internet o
f Thing
s Meaningf
ul,
2014, pp. 339
-
352
.
[24]
Y. Chu, H.
Wang, L. Z
heng, Z.
Wang, and
K.
L. Tan, “TRSO:
A
touri
sm recommender system based on ontology,”
Int
.
Conf
.
on Knowledge Science, Enginee
.
Management
,
2016, pp. 567
–
579
,
doi:
10.1007/978
-
3
-
319
-
47650
-
6_45.
[25]
M. A. Musen, “The
protégé proje
ct: a look back a
nd a look forwa
rd,”
AI Matters
, vol. 1, no. 4, pp. 4
-
12, 2015.
Evaluation Warning : The document was created with Spire.PDF for Python.