T
E
L
KO
M
N
I
KA
T
e
lec
om
m
u
n
icat
ion
,
Com
p
u
t
i
n
g,
E
lec
t
r
on
ics
an
d
Cont
r
ol
Vol.
18
,
No.
1
,
F
e
br
ua
r
y
2020
,
pp.
124
~
132
I
S
S
N:
1693
-
6930
,
a
c
c
r
e
dit
e
d
F
ir
s
t
G
r
a
de
by
Ke
me
nr
is
tekdikti
,
De
c
r
e
e
No:
21/E
/KP
T
/2018
DO
I
:
10.
12928/
T
E
L
KO
M
NI
KA
.
v18i1.
13006
124
Jou
r
n
al
h
omepage
:
ht
tp:
//
jour
nal.
uad
.
ac
.
id/
index
.
php/T
E
L
K
OM
N
I
K
A
A
devel
oped
G
P
S
t
raj
ect
ori
es
dat
a
m
anage
m
ent
s
ys
t
em
f
or
predi
ct
i
ng
t
ouri
s
t
s
'
P
O
I
Rul
a
Am
j
ad
Ham
id
1
,
M
u
aya
d
S
a
d
ik
Cr
ooc
k
2
1
Co
l
l
eg
e
o
f
B
u
s
i
n
e
s
s
In
f
o
rmat
i
cs
,
U
n
i
v
ers
i
t
y
o
f
In
f
o
rma
t
i
o
n
T
ec
h
n
o
l
o
g
y
an
d
Co
mmu
n
i
ca
t
i
o
n
s
,
Bag
h
d
ad
,
Iraq
2
Co
mp
u
t
er
E
n
g
i
n
eer
i
n
g
D
e
p
art
me
n
t
,
U
n
i
v
ers
i
t
y
o
f
T
ec
h
n
o
l
o
g
y
,
Iraq
Ar
t
icle
I
n
f
o
AB
S
T
RA
CT
A
r
ti
c
le
h
is
tor
y
:
R
e
c
e
ived
Apr
28
,
2019
R
e
vis
e
d
Ju
n
2
5
,
20
19
Ac
c
e
pted
Ju
l
12
,
20
19
O
n
e
o
f
t
h
e
areas
t
h
at
h
av
e
ch
a
l
l
en
g
es
i
n
t
h
e
u
s
e
o
f
i
n
t
er
n
et
o
f
t
h
i
n
g
s
(Io
T
)
i
s
t
h
e
fi
e
l
d
o
f
t
o
u
ri
s
m
an
d
t
rav
el
.
T
h
e
i
s
s
u
e
h
ere
i
s
h
o
w
t
o
em
p
l
o
y
t
h
i
s
t
ech
n
o
l
o
g
y
t
o
s
erv
e
t
h
e
t
o
u
ri
s
m
a
n
d
ma
n
ag
i
n
g
t
h
e
p
r
o
d
u
ced
d
at
a.
T
h
i
s
w
o
r
k
i
s
fo
cu
s
o
n
t
h
e
u
s
e
o
f
t
o
u
r
i
s
t
s
'
t
raj
ec
t
o
r
i
es
t
h
at
are
co
l
l
ect
e
d
fro
m
g
l
o
b
al
p
o
s
i
t
i
o
n
i
n
g
s
y
s
t
em
(G
PS)
mo
b
i
l
e
s
en
s
o
r
s
as
a
s
o
u
rce
o
f
i
n
f
o
rmat
i
o
n
.
T
h
e
ai
m
o
f
w
o
rk
i
s
t
o
p
re
d
i
c
t
p
referred
t
o
u
r
i
s
m
p
l
aces
f
o
r
t
o
u
ri
s
t
s
b
y
t
rac
k
i
n
g
t
o
u
ri
s
t
s
'
b
e
h
av
i
o
r
t
o
ex
t
ract
t
h
e
t
o
u
ri
s
m
p
l
ace
s
t
h
at
h
a
v
e
b
een
v
i
s
i
t
ed
b
y
s
u
ch
t
o
u
ri
s
t
s
.
D
e
n
s
i
t
y
b
a
s
ed
cl
u
s
t
eri
n
g
a
l
g
o
ri
t
h
m
i
s
mai
n
l
y
u
s
ed
t
o
ex
t
ract
s
t
ay
p
o
i
n
t
s
a
n
d
p
o
i
n
t
o
f
i
n
t
ere
s
t
(P
O
I).
By
p
r
o
j
ec
t
i
n
g
G
PS
l
o
cat
i
o
n
(fo
r
u
s
er
an
d
p
l
ace
s
)
o
n
t
h
e
G
o
o
g
l
e
ma
p
,
t
h
e
t
y
p
e
a
n
d
n
ame
o
f
p
l
ace
s
fav
o
red
b
y
t
h
e
t
o
u
ri
s
t
s
are
d
et
ermi
n
e
d
.
K
n
eares
t
n
e
i
g
h
b
o
r
(K
N
N
)
al
g
o
ri
t
h
m
w
i
t
h
h
av
er
s
i
n
e
d
i
s
t
a
n
ce
h
as
b
ee
n
ad
o
p
t
e
d
t
o
fi
n
d
t
h
e
n
eare
s
t
p
l
ace
s
fo
r
t
o
u
ri
s
t
s
.
T
h
e
ev
a
l
u
a
t
i
o
n
o
f
t
h
e
o
b
t
a
i
n
e
d
res
u
l
t
s
s
h
o
w
s
s
u
p
er
i
o
r
an
d
s
a
t
i
s
fact
o
ry
p
erfo
rma
n
ce
t
h
a
t
can
reac
h
t
h
e
o
b
j
ect
i
v
e
b
e
h
i
n
d
t
h
i
s
w
o
r
k
.
K
e
y
w
o
r
d
s
:
I
o
T
P
oint
of
int
e
r
e
s
t
S
tay
point
s
T
our
is
m
T
r
a
jec
tor
ies
Th
i
s
i
s
a
n
o
p
en
a
c
ces
s
a
r
t
i
c
l
e
u
n
d
e
r
t
h
e
CC
B
Y
-
SA
l
i
ce
n
s
e
.
C
or
r
e
s
pon
din
g
A
u
th
or
:
R
ula
Amjad
Ha
mi
d,
Unive
r
s
it
y
of
I
nf
or
mation
T
e
c
hnology
a
nd
C
omm
unica
ti
ons
,
I
r
a
q,
C
oll
e
ge
of
B
us
ines
s
I
nf
or
matics
B
a
ghda
d,
I
r
a
q.
E
mail:
e
ng_r
ula_a
mj
e
d@uoit
c
.
e
du.
iq
1.
I
NT
RODU
C
T
I
ON
I
t
is
f
r
e
que
ntl
y
thought
that
the
loca
l
a
uthor
it
ies
a
n
d
tour
is
m
a
ge
nc
ies
ha
ve
a
a
de
qua
te
unde
r
s
tanding
of
tou
r
is
t's
pr
e
f
e
r
e
nc
e
s
,
ne
e
ds
,
a
nd
how
loca
l
p
e
ople's
int
e
r
e
s
ts
c
a
n
be
in
tegr
a
ted
in
tou
r
is
m
plan
ning
[
1
]
.
A
c
hief
c
ha
ll
e
nge
in
mana
ging
tour
is
m
s
ys
tem
us
ing
I
o
T
is
how
to
t
r
a
c
k
us
e
r
be
ha
vior
s
a
nd
pr
e
f
e
r
e
nc
e
a
c
quis
it
ion
[
2]
.
T
h
e
r
e
is
a
ne
e
d
to
know
the
de
tails
inf
or
mation
of
p
r
e
c
is
e
loca
ti
ons
vis
it
e
d
by
tour
is
ts
,
the
a
tt
r
a
c
ted
loca
ti
ons
by
tour
is
t,
pe
r
s
ona
l
r
e
f
lec
ti
ons
on
tour
is
ts
’
e
xpe
r
ienc
e
s
a
nd
f
utur
e
tr
a
ve
l
be
ha
vior
a
l
int
e
nti
ons
[
3]
.
M
a
ny
s
tudi
e
s
e
mpl
oye
d
tour
is
t
GPS
tr
a
je
c
tor
ies
to
c
las
s
if
y
a
nd
f
or
e
c
a
s
t
the
be
ha
vior
s
of
tour
is
ts
that
vis
it
loca
ti
ons
by
c
oll
e
c
ti
ng
their
moveme
nts
,
c
hoice
s
a
nd
ne
e
ds
.
T
r
a
jec
tor
y
is
a
loca
ti
on
s
e
que
nc
e
(
s
pa
ti
a
l
–
tempor
a
l)
with
tr
a
ve
l
ti
mes
.
T
he
r
e
lati
on
a
mong
the
s
e
que
nc
e
s
de
pe
nd
s
on
the
ne
ig
h
bor
hood
f
unc
ti
on
a
nd
the
ti
me
tol
e
r
a
nc
e
[
4]
.
T
he
s
tudy
in
[
5]
de
s
igned
da
taflow
mi
ning
s
tr
uc
tur
e
f
or
us
e
r
’
s
mobi
le
be
ha
vior
tr
a
jec
tor
y
de
pe
nd
on
plac
e
s
e
r
vice
s
in
m
obil
e
.
T
he
a
im
of
the
s
tudy
o
f
[
5]
wa
s
to
ge
t
us
e
r
pa
th
da
ta
that
incor
por
a
tes
p
lac
e
inf
o
r
ma
ti
on
a
nd
s
oc
ial
in
f
o
r
mation.
Anothe
r
s
tudy
in
[
6
]
pr
opos
e
d
a
he
ur
is
ti
c
method
that
c
ombi
ne
s
dyna
mi
c
ti
me
wa
r
ping
a
nd
the
e
a
r
t
h
mover
's
dis
tanc
e
,
to
a
c
c
ur
a
tely
mea
s
ur
e
the
s
im
il
a
r
it
y
of
tour
is
t
tr
a
jec
tor
ies
.
T
he
s
tudy
of
[
7]
e
xpa
nde
d
the
a
ppli
c
a
ti
on
of
tour
is
t
moveme
nts
in
the
mobi
le
I
nt
e
r
ne
t
e
r
a
,
in
whic
h
moveme
nt
da
ta
(
us
ing
GPS
t
r
a
jec
tor
ies
)
c
ould
be
c
oll
e
c
ted
mor
e
e
a
s
il
y.
T
his
wa
s
done
by
p
r
opos
ing
a
method
that
im
pr
ove
s
pr
e
diction
a
c
c
ur
a
c
y
a
n
d
tr
a
de
-
of
f
be
twe
e
n
p
r
e
diction
a
c
c
ur
a
c
y
a
nd
e
f
f
icie
nc
y.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
A
de
v
e
loped
GP
S
tr
ajec
tor
ies
data
manage
me
nt
s
y
s
tem
for
pr
e
dicting
tour
is
ts
'
P
OI
(
R
ula
A
mjad
Ha
mid
)
125
T
he
s
tudy
in
[
8]
pr
opos
e
d
method
us
ing
ti
e
d
r
a
n
king
with
or
dinal
logi
s
ti
c
r
e
gr
e
s
s
ion
to
pr
e
dict
th
e
f
a
c
tor
s
that
c
a
n
a
f
f
e
c
t
the
tour
is
m
s
e
c
tor
in
I
r
a
q
a
nd
wha
t
a
r
e
the
f
a
c
tor
s
inf
luenc
ing
thi
s
f
ield
in
o
r
de
r
to
f
oc
us
on
the
de
ve
lopm
e
nt
of
thi
s
indus
tr
y
.
T
he
wo
r
k
in
[
9
]
e
xtr
a
c
ted
tour
is
m
P
OI
f
r
om
a
va
s
t
qua
nti
ty
of
a
ppli
c
a
nt
P
OI
s
ba
s
e
d
on
tour
is
t
pr
e
f
e
r
e
nc
e
s
.
T
he
r
e
s
e
a
r
c
h
in
[
10]
pr
e
s
e
nted
the
us
e
of
loca
ti
on
c
he
c
k
-
ins
,
w
hich
a
r
e
a
va
il
a
ble
on
mobi
le
s
oc
ial
media
platf
or
ms
,
a
s
a
n
e
xtr
a
da
ta
s
upply
to
r
e
vis
e
tour
i
s
t
be
ha
vior
s
.
I
n
ge
ne
r
a
l,
mos
t
e
xis
ti
ng
a
ppr
oa
c
he
s
a
r
e
not
c
a
pa
ble
to
tac
kle
the
c
ha
ll
e
nge
in
int
e
gr
a
ted
a
nd
wide
s
pr
e
a
d
manne
r
[
11]
.
T
he
a
im
of
thi
s
wor
k
is
to
e
xtr
a
c
t
GPS
point
s
f
r
om
tr
a
jec
tor
ies
da
ta,
a
na
lyze
be
ha
vior
pa
tt
e
r
ns
o
f
tour
is
t
pa
ths
,
a
nd
pr
e
d
ict
the
pr
e
f
e
r
r
e
d
plac
e
s
f
or
tour
is
ts
.
T
his
is
done
us
ing
the
google
map
inf
or
mation
to
de
ter
mi
ne
the
f
a
vo
r
it
e
plac
e
s
by
tou
r
is
t
us
ing
c
l
us
ter
ing
a
lgor
it
hms
.
T
he
s
e
a
lgor
it
hms
de
pe
nd
m
a
inl
y
on
the
de
ns
it
y
in
f
or
mation
a
nd
the
P
OI
f
or
e
a
c
h
tour
is
t.
T
he
c
oll
e
c
ted
in
f
or
mation
is
us
e
d
f
or
buil
ding
the
da
tas
e
t
f
or
s
ys
tem
pr
e
diction
.
2.
DA
T
A
DE
S
CR
I
P
T
I
ON
T
he
pr
opos
e
d
s
ys
tem
us
e
s
Ge
o
L
if
e
T
r
a
jec
tor
ies
da
tas
e
t.
T
his
GPS
da
tas
e
t
wa
s
c
ompos
e
d
in
(
M
icr
os
of
t
R
e
s
e
a
r
c
h
As
ia)
by
182
us
e
r
s
in
a
pe
r
iod
o
f
ove
r
f
ive
ye
a
r
s
(
2007
-
2012)
.
A
t
r
a
jec
tor
y
of
th
is
da
tas
e
t
is
de
noted
by
point
s
s
e
que
n
c
e
s
.
E
a
c
h
one
ha
s
the
inf
or
mation
of
(
latit
ude
,
longi
tude,
a
nd
a
lt
it
ude
)
.
T
he
Ge
oL
if
e
da
tas
e
t
wa
s
c
oll
e
c
ted
by
us
e
r
mobi
le
de
vice
s
ove
r
a
ti
me
pe
r
iod
of
f
ive
ye
a
r
s
.
I
t
r
e
pr
e
s
e
nts
us
e
r
s
’
moveme
nts
his
tor
y
li
ke
going
to
wor
k
,
r
e
tu
r
ning
to
ho
me,
a
nd
a
ll
k
inds
of
a
c
ti
vit
ies
in
the
da
y
li
f
e
o
f
unde
r
lyi
ng
us
e
r
s
[
12]
.
I
n
or
de
r
to
tes
t
the
pr
opos
e
d
a
lgor
it
hm
ove
r
di
f
f
e
r
e
nt
da
tas
e
t,
GPS
poin
ts
f
or
a
gr
oup
of
I
r
a
qi
tou
r
is
ts
a
r
e
us
e
d
to
e
xtr
a
c
t
their
be
ha
vior
dur
ing
their
vis
it
to
tour
is
m
plac
e
s
in
the
c
it
y
of
E
r
bil
.
T
his
c
it
y
wa
s
c
hos
e
n
be
c
a
us
e
it
is
c
ons
ider
e
d
the
mo
s
t
im
por
tant
I
r
a
qi
pr
ovince
s
in
ter
ms
of
the
div
e
r
s
it
y
of
tour
is
m
a
r
e
a
s
.
3.
P
ROP
OS
E
D
S
Y
S
T
E
M
As
mentioned
e
a
r
li
e
r
,
thi
s
wor
k
pr
oduc
e
s
a
tour
i
s
t
pr
e
diction
s
ys
tem
ba
s
e
d
on
the
P
OI
of
tour
is
ts
us
ing
dif
f
e
r
e
nt
tr
a
jec
tor
y
da
tas
e
ts
.
F
or
e
a
s
ing
the
r
e
a
ding
f
low
of
thi
s
pa
pe
r
,
the
p
r
opos
e
d
s
ys
tem
c
a
n
be
e
xplaine
d
a
c
c
or
ding
to
the
a
ppli
e
d
s
teps
a
s
f
oll
ows
:
3.
1.
Dat
a
c
leani
n
g
(
p
r
e
p
r
oc
e
s
s
in
g)
T
he
f
ir
s
t
s
tep
is
the
a
na
lys
is
a
nd
pr
e
pr
oc
e
s
s
ing
of
t
he
da
tas
e
t
to
r
e
move
pos
s
ibl
e
nois
e
f
r
om
the
da
ta.
Da
ta
c
lea
ning
is
a
tec
hnique
to
de
tec
t
a
nd
e
it
he
r
r
e
move
or
c
o
r
r
e
c
t
incons
is
tenc
ies
or
mi
s
s
ing
d
a
ta
in
a
da
tas
e
t
[
13]
.
S
uc
h
incons
is
tent
da
ta
may
a
f
f
e
c
t
the
r
e
s
ult
s
of
the
s
tudy.
Nois
e
in
da
ta
may
be
c
a
us
e
d
by
many
dif
f
e
r
e
nt
r
e
a
s
ons
,
s
uc
h
a
s
e
r
r
or
in
e
lec
tr
on
ic
de
vi
c
e
s
(
e
.
g.
GPS
logger
s
)
,
s
of
twa
r
e
e
r
r
o
r
o
r
human
mi
s
take
.
Af
ter
c
lea
ning,
the
da
ta
mus
t
be
c
ons
is
tent
with
th
e
other
s
im
il
a
r
da
ta
in
the
s
ys
te
m.
F
or
e
xa
mpl
e
,
it
a
ppe
a
r
s
that
ther
e
a
r
e
point
s
on
the
pa
th,
c
onf
li
c
ti
ng
f
r
om
the
pa
tt
e
r
n
of
the
pa
th.
At
thi
s
point
,
the
pe
r
s
on
s
udde
nly
take
s
a
ve
r
y
high
s
pe
e
d,
f
or
e
xa
mpl
e
mo
r
e
than
200
km/
s
,
in
les
s
than
5
s
e
c
onds
,
c
a
n
be
r
e
moved
f
or
invalidi
ty.
T
o
r
e
move
thi
s
type
of
nois
e
,
a
s
olut
i
on
ba
s
e
d
on
indi
vidual
ve
locity
is
a
dopted
a
long
the
pa
th.
T
his
is
done
by
c
a
lcula
te
the
ve
locity
take
n
f
r
om
t
he
indi
vidual
of
e
a
c
h
point
on
the
t
r
a
c
ks
,
a
nd
then
c
he
c
king
whe
ther
the
s
pe
e
d
of
the
indi
vidual
he
s
it
a
tes
to
a
h
igh
va
lue
be
twe
e
n
a
ny
two
point
s
.
3.
2.
F
e
a
t
u
r
e
e
xt
r
ac
t
ion
T
he
ne
xt
s
tep
is
f
indi
ng
a
nd
e
xtr
a
c
ti
ng
ne
w
f
e
a
tur
e
s
f
r
om
the
da
tas
e
t.
T
he
ne
w
e
xtr
a
c
ted
f
e
a
tur
e
s
a
r
e
s
tay
point
s
a
nd
P
OI
.
3.
2.
1.
E
xt
r
ac
t
in
g
s
t
ay
p
oi
n
t
s
S
tay
point
s
:
a
r
e
ge
ogr
a
phic
a
r
e
a
s
whe
r
e
the
indi
v
idual
ha
s
s
pe
nt
a
long
t
im
e
in
thei
r
s
ur
r
oundings
c
e
nter
point
.
S
tay
P
oint
a
r
e
e
xtr
a
c
ted
a
nd
gr
oupe
d
f
r
om
us
e
r
poin
ts
ba
s
e
d
on
the
ti
me
a
nd
dis
tanc
e
,
ta
ke
n
on
a
r
oute
to
a
ge
ogr
a
phic
a
r
e
a
.
S
tay
poin
ts
c
a
n
be
de
te
c
ted
a
utom
a
ti
c
a
ll
y
f
r
om
a
us
e
r
’
s
GPS
tr
a
jec
tor
y
by
s
e
e
king
the
s
pa
ti
a
l
r
e
gion
whe
r
e
the
us
e
r
s
pe
nt
a
pe
r
iod
e
xc
e
e
ding
a
c
e
r
tain
thr
e
s
hold
[
14]
.
I
n
thi
s
s
e
c
ti
on,
the
s
tay
point
s
a
r
e
de
tec
ted
f
r
om
us
e
r
s
'
mobi
li
ty
pa
ths
by
s
e
e
king
the
s
pa
ti
a
l
r
e
gion
whe
r
e
the
us
e
r
s
taye
d
f
or
while.
T
he
a
lgo
r
it
hm
that
ha
s
be
e
n
pr
opos
e
d
in
[
15]
wa
s
a
dopted
in
or
de
r
to
e
xtr
a
c
t
s
tay
point
s
a
s
s
hown
in
F
igur
e
1
.
I
n
the
p
r
opos
e
d
s
ys
tem,
if
the
tour
is
t
s
p
e
nt
mor
e
than
35
m
inut
e
s
withi
n
a
dis
tanc
e
of
20
0
mete
r
s
,
the
point
is
de
tec
ted
a
s
s
tay
po
int
.
I
n
other
wor
ds
,
a
c
lus
ter
is
de
tec
ted
a
nd
a
ll
oc
a
ted.
T
he
e
xtr
a
c
ted
s
t
a
y
point
inf
or
mation
c
ontains
mea
n
c
oor
dinate
s
,
a
r
r
ival
t
im
e
(
S
.
a
r
vT
)
a
nd
lea
ving
ti
me
(
S
.
levT
)
f
or
e
a
c
h
tour
is
t,
indi
vidually.
At
the
other
point
,
the
th
r
e
s
hold
is
s
e
lec
ted
ba
s
e
d
on
the
a
ve
r
a
ge
ti
me,
c
omput
e
d
f
r
om
[
16]
a
nd
[
17]
.
T
he
f
i
r
s
t
one
take
s
ti
me
thr
e
s
hold
=
20
mi
nutes
,
while
the
other
s
pe
c
if
ied
the
opti
mal
ti
me
in
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
1
,
F
e
br
ua
r
y
2020
:
124
-
132
126
be
twe
e
n
10
to
60
mi
nu
tes
.
Ha
ve
r
s
ine
Dis
tanc
e
(
HD
)
f
or
mul
a
,
p
r
e
s
e
nted
in
[
12]
,
is
us
e
d
in
thi
s
r
e
s
e
a
r
c
h
to
c
a
lcula
te
the
dis
tanc
e
be
twe
e
n
two
point
s
on
a
s
ph
e
r
e
.
T
he
f
or
mul
a
is
given
by
the
f
oll
owing
e
qua
ti
on:
Dis
tanc
e
=
2
s
in
−
1
(
√
2
(
∅
−
∅
2
)
+
(
∅
)
(
∅
)
2
(
−
2
)
)
(
1)
whe
r
e
(
R
)
r
e
pr
e
s
e
nts
e
a
r
t
h
r
a
dius
,
∅
a
nd
φ
a
r
e
c
or
r
e
s
pondingl
y
the
latit
ude
s
a
nd
longi
tudes
o
f
po
int
s
(
i,
j)
,
r
e
s
pe
c
ti
ve
ly.
F
igur
e
1
.
S
tay
point
de
tec
ti
on
a
lgor
it
hm
3.
2.
2.
E
xt
r
ac
t
p
oin
t
of
in
t
e
r
e
s
t
P
OI
is
a
n
im
por
tant
ve
nue
/l
oc
a
ti
on
in
the
phys
ica
l
wor
ld,
s
uc
h
a
s
a
s
hopping
mall
or
a
thea
tr
e
,
lake
.
Ge
ne
r
a
ll
y,
P
OI
be
longs
to
one
or
mor
e
c
a
tegor
ies
li
ke
e
duc
a
ti
on,
e
nter
tainment,
a
r
ts
,
f
ood
a
nd
dini
ng,
gove
r
nment,
he
a
lt
h
&
be
a
uty,
home
&
f
a
mi
ly
,
s
ho
ppin
g,
s
por
ts
,
a
nd
na
tu
r
e
[
18]
.
A
f
ter
c
ounti
ng
s
tay
point
s
in
the
pr
e
vious
pha
s
e
,
we
s
hould
now
be
a
ble
to
dis
c
ove
r
loca
ti
ons
whe
r
e
pe
ople
s
pe
nd
a
lot
of
ti
me
f
r
e
que
ntl
y
in
their
s
ur
r
oundings
.
T
o
f
ind
s
uc
h
plac
e
s
,
the
f
oll
owing
s
teps
a
r
e
a
ppli
e
d:
−
I
nter
e
s
ti
ng
p
oint
s
a
r
e
c
ol
lec
ted
us
ing
the
de
ns
it
y
-
ba
s
e
d
c
lus
ter
ing
a
lgor
it
hm
to
f
ind
g
r
oups
c
ontaini
ng
a
t
lea
s
t
k
point
s
withi
n
them.
T
he
s
e
c
lus
ter
s
r
e
pr
e
s
e
nt
the
r
e
gions
that
a
r
e
f
r
e
que
ntl
y
vis
it
e
d
.
T
he
r
e
f
or
e
,
ve
r
y
li
ke
ly
to
be
r
e
gion
o
f
int
e
r
e
s
t.
−
E
a
c
h
r
e
gion
is
r
e
pr
e
s
e
nted
by
c
e
nter
point
,
wh
ich
is
a
point
o
f
int
e
r
e
s
t.
T
he
s
e
loca
ti
ons
c
a
n
be
a
r
e
s
taur
a
nt,
a
s
hopping
c
e
nter
,
a
univer
s
it
y
buil
ding
,
or
a
tou
r
is
t
a
tt
r
a
c
ti
on
.
DB
S
C
AN
a
lgor
it
hm
of
[
19]
is
a
dopted
in
thi
s
wor
k.
T
his
a
lgor
it
h
m
c
ompos
e
s
a
gr
oup
of
point
s
a
nd
c
lus
ter
s
togethe
r
a
s
we
ll
a
s
the
point
s
that
a
r
e
pa
c
ke
d
s
tr
ongly
withi
n
a
given
thr
e
s
hold
dis
tanc
e
in
s
pa
c
e
a
nd
mar
ks
point
s
a
s
ou
tl
ier
s
that
li
e
a
lone
in
low
de
ns
it
y
r
e
gions
.
DB
S
C
AN
r
e
qui
r
e
s
two
pa
r
a
mete
r
s
;
the
f
ir
s
t
one
is
e
ps
il
on,
whic
h
is
the
maximum
d
is
tanc
e
be
twe
e
n
two
s
a
mpl
e
s
to
be
c
ons
ider
e
d
in
the
s
a
me
ne
igh
bor
hood
while
the
s
e
c
ond
one
is
the
mi
nim
u
m
numbe
r
o
f
point
s
,
r
e
qui
r
e
d
to
f
o
r
m
a
de
ns
e
r
e
gion.
T
o
e
s
ti
m
a
te
thes
e
two
pa
r
a
mete
r
,
the
a
utho
r
s
of
[
20]
pr
opos
e
d
a
h
e
ur
is
ti
c
to
de
ter
mi
ne
them
with
r
e
ga
r
ds
to
the
“
t
hinnes
t”
c
lus
ter
in
the
da
taba
s
e
.
I
n
their
e
xpe
r
im
e
nt
,
the
a
u
thor
s
indi
c
a
te
that
the
opti
mal
va
lue
of
f
o
r
k
>
4
.
T
hus
,
in
thi
s
wor
k,
we
s
e
t
k
=
M
ini
mum
P
oint
s
in
c
lus
ter
=
4.
3.
3
.
F
in
d
n
e
ar
e
s
t
p
lace
s
/p
r
e
d
ict
ion
T
he
f
indi
ng
of
the
ne
a
r
e
s
t
tour
is
m
plac
e
s
f
or
the
t
our
is
t
is
the
ne
xt
s
tep
in
pr
opos
e
d
s
y
s
tem.
T
his
is
to
e
a
s
e
the
pr
e
diction
of
the
r
e
c
omm
e
nde
d
plac
e
s
f
or
the
tour
is
t
that
c
a
n
s
a
ti
s
f
y
his
/her
r
e
que
s
ts
.
T
he
e
va
luation
of
ne
a
r
e
s
t
plac
e
s
is
pe
r
f
or
med
us
ing
KN
N
method.
KN
N
s
e
a
r
c
h
is
one
of
the
mos
t
f
unda
menta
l
pr
oblems
,
whic
h
ha
s
be
e
n
e
xtens
ively
s
tudi
e
d
in
va
r
ious
f
ields
o
f
c
omput
e
r
s
c
ienc
e
,
s
uc
h
a
s
da
ta
mi
ning,
inf
o
r
mation
r
e
tr
ieva
l,
a
nd
s
pa
ti
a
l
da
taba
s
e
s
[
16]
.
Us
ing
the
us
e
r
GPS
pos
it
ion
a
nd
P
OI
s
,
t
he
KN
N
que
r
y
c
a
n
f
ind
the
c
los
e
s
t
P
OI
f
r
om
that
tour
is
t
(
s
malles
t
dis
tanc
e
f
r
om
the
tour
is
t)
.
I
n
s
pa
ti
a
l
da
taba
s
e
s
,
the
KN
N
que
r
y
c
a
n
be
us
e
d
in
f
indi
ng
the
ne
a
r
e
s
t
P
OI
,
s
uc
h
a
s
a
r
e
s
taur
a
nt
to
a
tour
is
t’
s
c
ur
r
e
nt
loc
a
ti
on.
I
n
thi
s
wor
k,
Ge
opa
nda
s
is
us
e
d
a
s
a
n
ope
n
s
our
c
e
ge
os
pa
ti
a
l
da
ta
pr
oc
e
s
s
ing
method
a
s
a
n
a
p
pli
c
a
ti
on
in
P
ython
langua
ge
[
21
]
.
As
a
r
e
s
ult
,
the
s
ys
tem
is
now
a
ble
to
pr
e
dict
the
na
mes
of
the
tou
r
is
t's
f
a
vor
it
e
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
A
de
v
e
loped
GP
S
tr
ajec
tor
ies
data
manage
me
nt
s
y
s
tem
for
pr
e
dicting
tour
is
ts
'
P
OI
(
R
ula
A
mjad
Ha
mid
)
127
plac
e
s
thr
ough
the
pr
e
f
e
r
r
e
d
type
of
tour
is
m
plac
e
s
(
pr
e
vious
ly
e
xtr
a
c
ted)
a
nd
the
ne
a
r
e
s
t
tour
is
m
pla
c
e
s
f
or
m
him
/her
.
F
igur
e
2
il
lus
tr
a
tes
the
wor
king
s
teps
of
the
p
r
opos
e
d
s
ys
tem
f
or
pr
e
dicting
the
r
e
c
omm
e
nde
d
tour
is
m
plac
e
s
f
or
int
e
r
e
s
ti
ng
tour
is
ts
us
ing
the
inf
or
mation
o
f
GPS
a
nd
P
OI
f
o
r
them.
I
t
is
c
lea
r
ly
s
h
own
that
the
im
por
tanc
e
of
e
va
luating
the
P
OI
f
o
r
tour
is
ts
in
int
e
r
e
s
ti
ng
a
r
e
a
to
p
r
e
dict
the
tour
is
m
plac
e
c
a
n
be
a
tt
e
nde
d.
T
he
tr
a
jec
tor
y
da
ta
f
or
e
a
c
h
us
e
r
is
c
oll
e
c
ted.
T
his
da
ta
is
c
lea
ne
d
up
to
r
e
move
a
ny
pos
s
ib
le
nois
e
.
T
his
is
to
e
xtr
a
c
t
the
P
OI
f
or
them
,
indi
vidually
.
T
he
r
e
a
l
Ge
oloca
ti
on
of
the
unde
r
lyi
ng
us
e
r
s
(
tou
r
is
ts
)
is
c
oll
e
c
ted
f
r
om
thei
r
s
mar
t
phone
s
to
e
va
luate
the
ne
a
r
e
s
t
dis
tanc
e
be
twe
e
n
them
a
nd
P
O
I
,
whic
h
r
e
c
omm
e
nde
d
a
s
tour
is
m
plac
e
s
.
F
igur
e
2
.
F
low
diagr
a
m
o
f
p
r
opos
e
d
s
ys
tem
4.
RE
S
UL
T
S
Af
ter
a
pplyi
ng
r
e
pr
oc
e
s
s
ing
ope
r
a
ti
on
on
the
c
ons
ider
e
d
da
tas
e
ts
,
s
top
point
s
a
lgo
r
it
hm
wa
s
a
ppli
e
d
on
r
e
f
ined
us
e
r
s
'
t
r
a
jec
tor
ies
.
T
he
number
of
e
xt
r
a
c
ted
s
top
point
s
is
e
qua
l
to
13320
f
o
r
181
us
e
r
s
(
t
our
is
ts
)
.
F
igur
e
3
s
hows
pa
r
t
of
e
va
luate
d
s
top
point
s
a
f
ter
mi
ning
a
ll
the
tr
a
jec
tor
ies
with
e
xtr
a
c
ted
f
e
a
tur
e
(
us
e
r
id,
longi
tude,
lat
it
ude
o
f
point
,
a
r
r
ivi
ng
ti
me,
lea
ving
ti
me
a
nd
tot
a
l
s
pe
nding
ti
me)
.
T
a
ble
1
r
e
pr
e
s
e
nts
t
he
us
e
r
s
with
their
c
lus
ter
e
d
point
s
of
P
O
I
that
a
r
e
de
tec
te
d
by
DB
S
C
AN
a
lgor
it
hm.
F
igu
r
e
4
s
hows
the
e
xe
c
uti
on
of
the
c
lus
ter
ing
a
lgor
it
hm
f
o
r
one
tour
is
t
us
ing
S
kle
a
r
n.
c
lus
ter
in
python
3
[
22]
with
c
ha
r
ts
that
s
hows
Numbe
r
of
point
s
in
e
a
c
h
c
lus
ter
with
c
e
nter
pos
it
ion
of
c
l
us
ter
a
nd
e
s
ti
mate
d
nu
mber
of
c
lus
ter
e
d
point
s
of
int
e
r
e
s
t
a
nd
the
nois
e
point
s
.
F
igur
e
3.
E
x
tr
a
c
ted
s
tay
point
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
1
,
F
e
br
ua
r
y
2020
:
124
-
132
128
T
a
ble
1
.
Numbe
r
o
f
s
top
poin
ts
a
nd
number
of
c
lus
ter
s
/P
OI
U
se
r
id
N
o
o
f
tr
a
j
e
ct
o
r
i
e
s
N
o
o
f
st
o
p
p
o
i
n
ts
N
u
m
b
e
r
o
f
cl
u
st
e
r
s/
P
OI
0
171
186
5
1
71
39
4
2
175
135
5
3
322
553
13
4
395
587
13
5
86
63
4
12
77
66
4
13
144
89
7
17
391
361
16
22
146
238
11
23
34
42
6
24
101
77
10
30
296
514
11
35
74
330
10
38
110
163
6
39
227
158
11
42
150
36
5
52
104
98
9
84
215
100
8
92
157
56
4
104
115
67
5
119
45
75
5
126
263
85
7
144
610
83
10
163
809
145
17
167
385
197
9
(
a
)
(
b)
F
igur
e
4.
E
x
tr
a
c
ted
P
OI
f
o
r
one
us
e
r
(
a
)
Numbe
r
o
f
point
s
in
e
a
c
h
c
lus
ter
with
c
e
nter
pos
it
ion
o
f
C
lus
ter
/P
OI
(
b)
E
s
ti
mate
d
number
of
c
lus
ter
s
T
he
e
xtr
a
c
ted
P
OI
of
us
e
r
s
a
r
e
p
r
ojec
ted
on
dyna
mi
c
map
(
f
r
o
m
Google
)
with
ge
ona
mes
of
P
OI
f
o
r
B
e
ij
ing
a
s
s
hown
in
F
igur
e
5.
B
e
ij
ing
P
OI
we
r
e
c
oll
e
c
ted
f
r
om
we
bs
it
e
in
[
23]
.
F
igu
r
e
6
s
hows
the
P
OI
f
o
r
e
a
c
h
tour
is
t
a
nd
the
type
o
f
e
a
c
h
plac
e
.
B
y
Appl
ying
KN
N
the
dis
tanc
e
s
be
twe
e
n
tour
is
t
GPS
po
int
s
a
nd
B
e
ij
ing
P
OI
a
r
e
c
a
lcula
ted
to
f
ind
K
ne
a
r
e
s
t
int
e
r
e
s
ted
plac
e
s
f
r
om
tou
r
is
t
to
f
inally
p
r
e
dict
the
p
r
e
f
e
r
r
e
d
plac
e
s
.
T
he
f
oll
owing
a
s
s
umpt
ions
a
r
e
im
pos
e
d
wh
e
n
c
a
lcula
ti
ng
a
dis
tanc
e
f
or
r
e
gion
o
f
int
e
r
e
s
t
:
−
C
ir
c
ular
r
e
gion
o
f
in
ter
e
s
t:
the
c
e
nter
of
r
e
gion
r
e
p
r
e
s
e
nts
P
OI
−
T
he
R
a
dius
e
s
of
r
e
gions
a
r
e
dif
f
e
r
e
nt
f
r
om
e
a
c
h
other
.
As
s
umi
ng
lake
a
r
e
a
to
be
dif
f
e
r
e
nt
in
s
ize
c
ompar
e
d
with
mall
.
T
o
incr
e
a
s
e
the
va
li
d
it
y
of
the
pr
opos
e
d
s
ys
tem
a
n
e
s
s
e
nti
a
l
ne
e
d
f
or
f
u
r
ther
tes
t.
A
number
of
dono
r
tour
is
ts
a
r
e
s
e
lec
ted
to
e
xtr
a
c
t
th
e
ir
GPS
s
top
poi
nts
withi
n
E
r
bil
c
i
ty
(
one
of
the
mos
t
f
a
mous
c
it
i
e
s
in
I
r
a
q
with
it
s
va
r
ied
tour
is
t
p
lac
e
s
)
.
I
n
o
r
de
r
to
pr
e
dict
their
p
r
e
f
e
r
r
e
d
tour
is
m
type,
the
dis
tanc
e
s
of
thes
e
point
s
f
r
om
f
a
mous
tou
r
is
t
plac
e
s
(
c
a
n
be
c
ons
ider
e
d
a
s
R
OI
)
a
r
e
c
a
lcula
t
e
d.
A
da
taba
s
e
is
c
r
e
a
ted
f
or
t
he
mos
t
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
A
de
v
e
loped
GP
S
tr
ajec
tor
ies
data
manage
me
nt
s
y
s
tem
for
pr
e
dicting
tour
is
ts
'
P
OI
(
R
ula
A
mjad
Ha
mid
)
129
we
ll
-
known
tour
is
m
loca
ti
ons
in
I
r
a
q
wi
th
their
po
s
it
ions
(
latit
ude
,
longi
tude
)
a
nd
the
type
of
int
e
r
e
s
t
f
or
e
a
c
h
plac
e
a
s
s
hown
in
F
igur
e
7.
F
or
e
xa
mpl
e
,
the
dis
t
a
nc
e
us
ing
KN
N
is
c
a
lcula
ted
by
a
s
s
umi
ng
the
r
a
dius
f
r
om
the
c
e
nter
of
E
r
bil
C
a
s
tl
e
=
200
mete
r
s
while
S
e
r
s
a
nk
R
e
s
or
t
=
500
mete
r
s
a
nd
F
a
mi
ly
M
a
ll
50
mete
r
s
.
T
he
r
e
s
ult
s
a
r
e
s
hown
in
F
igur
e
8.
T
a
ble
2
r
e
pr
e
s
e
nts
the
s
top
point
s
of
two
us
e
r
s
with
their
dis
tanc
e
s
f
r
om
the
c
e
nter
point
of
r
e
gion.
F
igur
e
5
.
Dyna
mi
c
google
map
s
hown
B
e
ij
ing
P
O
I
a
nd
us
e
r
P
OI
F
igur
e
6.
P
OI
types
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
1
,
F
e
br
ua
r
y
2020
:
124
-
132
130
F
igur
e
7.
I
r
a
qi
P
O
I
with
types
of
int
e
r
e
s
t
F
igur
e
8.
Dyna
mi
c
google
map
of
tour
is
ts
GPS
poi
nts
T
a
ble
2
.
C
lus
ter
ing
pe
r
f
o
r
manc
e
mea
s
ur
e
U
s
e
r
i
d
S
il
houe
tt
e
C
oe
f
f
ic
ie
nt
U
s
e
r
i
d
S
il
houe
tt
e
C
oe
f
f
ic
ie
nt
0
0.165
36
0.3
1
0.364
38
0.4
2
0.318
52
0.12
3
0.03
68
0.5
5
0.3
84
0.1
6
0.114
85
0.1
7
0.415
92
0.2
9
0.638
104
0.1
12
0.351
112
0.037
13
0.332
119
0.6
14
0.4
126
0.083
15
0.382
157
0.274
17
0.419
159
0.724
18
0.04
165
0.484
23
0.2
167
0.228
24
0.1
22
0.458
30
0.6
179
0.087
35
0.9
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
A
de
v
e
loped
GP
S
tr
ajec
tor
ies
data
manage
me
nt
s
y
s
tem
for
pr
e
dicting
tour
is
ts
'
P
OI
(
R
ula
A
mjad
Ha
mid
)
131
5.
CL
UST
E
R
P
E
RF
ORM
AN
CE
M
E
ASUR
E
I
n
thi
s
wor
k
S
il
houe
tt
e
I
nde
x
[
24]
is
a
ppli
e
d
to
mea
s
ur
e
c
lus
ter
ing
pe
r
f
or
manc
e
.
T
his
mea
s
ur
e
is
a
dopted
ba
s
e
d
on
it
s
a
c
c
ur
a
c
y,
popular
i
ty
a
nd
s
im
pli
c
it
y
of
im
pleme
ntation
.
S
il
houe
tt
e
index
give
s
a
n
idea
a
bout
the
s
a
mpl
e
s
s
im
il
a
r
it
y
with
other
s
a
mpl
e
s
withi
n
the
s
a
me
c
lus
ter
(
c
ohe
s
ion)
a
nd
dis
s
im
il
a
r
it
y
with
other
s
a
mpl
e
s
in
othe
r
c
lus
ter
s
(
s
e
pa
r
a
ti
on)
.
I
t
r
a
nge
s
(
f
r
om
−
1
to
+
1)
,
whe
r
e
the
h
igher
va
lue
m
e
a
ns
it
is
withi
n
c
lus
ter
s
im
il
a
r
it
y
a
nd
the
lowe
r
va
lue
mea
ns
it
is
the
int
r
a
-
c
lus
ter
s
im
il
a
r
it
y
[
24
]
.
T
a
ble
2
s
hows
the
S
il
houe
tt
e
of
e
a
c
h
us
e
r
a
f
ter
a
pplyi
ng
DB
S
C
A
N
c
lus
ter
ing
on
Ge
oli
f
e
da
tas
e
t.
M
os
t
of
c
oe
f
f
icie
nts
ha
ve
the
va
lue
(
be
twe
e
n
0.
1
a
nd
0.
9
)
,
whic
h
r
e
pr
e
s
e
nts
pe
r
f
e
c
t
c
lus
ter
ing
a
nd
o
ther
les
s
than
z
e
r
o
whic
h
r
e
pr
e
s
e
nts
wor
s
t
c
lus
t
e
r
ing.
6.
E
VA
L
UA
T
I
ON
I
n
thi
s
wor
k
the
e
xpe
r
im
e
ntal
r
e
s
ult
s
a
r
e
e
va
luate
d
us
ing
pr
e
c
is
ion,
r
e
c
a
ll
a
s
:
(
2)
the
r
e
c
a
ll
va
lue
of
DB
S
C
AN
method
is
a
bout
(
0
.
5
91489)
,
while
the
pr
e
c
is
ion
is
a
bout
(
0.
371658
)
.
As
a
be
nc
h
mar
k,
thes
e
va
lues
a
r
e
c
ompar
e
d
with
the
r
e
s
ult
s
in
[
25]
,
whe
r
e
the
be
s
t
r
e
c
a
ll
va
lue
wa
s
a
bout
(
0.
36)
.
T
his
mea
ns
DB
S
C
AN
jus
t
dis
c
ove
r
e
d
(
36%
)
of
the
c
or
r
e
c
t
s
tops
with
p
r
e
c
is
ion
a
t
(
0
.
5)
.
T
he
r
e
f
or
e
,
the
pr
opos
e
d
wor
k
in
thi
s
pa
pe
r
outper
f
o
r
ms
the
p
r
e
vious
wor
k
r
e
s
ult
s
in
te
r
m
of
r
e
c
a
ll
.
7.
CONC
L
USI
ON
A
tour
is
m
p
lac
e
s
pr
e
diction
a
nd
r
e
c
omm
e
nda
ti
on
wa
s
pr
opos
e
d.
I
n
thi
s
wo
r
k,
DB
S
C
AN
a
nd
ne
a
r
e
s
t
ne
ighbor
we
r
e
us
e
d
to
e
xt
r
a
c
t
a
nd
pr
e
dict
the
types
of
tour
is
m
plac
e
s
pr
e
f
e
r
r
e
d
by
the
tour
is
ts
us
ing
the
vis
it
e
d
plac
e
s
by
them.
S
uc
h
in
f
or
mation
wa
s
us
e
d
in
th
e
tour
is
m
r
e
c
omm
e
nda
ti
on
s
ys
tems
or
b
y
tour
is
m
a
ge
nc
ies
to
know
plac
e
s
of
tou
r
is
t
a
tt
r
a
c
ti
ons
.
C
lus
ter
ing
a
lgo
r
it
hm
wa
s
e
va
luate
d
us
ing
S
i
lhouette
C
oe
f
f
icie
nt.
T
he
s
ys
tem
e
va
luation
s
howe
d
the
out
pe
r
f
or
manc
e
of
the
p
r
opos
e
d
s
ys
tem
ove
r
pr
e
vious
one
s
in
ter
m
o
f
r
e
c
a
ll
.
T
he
s
ys
tem
c
ould
be
e
xpa
nde
d
us
ing
thes
e
r
e
s
ult
s
in
tour
is
t
r
e
c
omm
e
nda
ti
on
s
ys
tems
to
pr
ovide
s
ugge
s
ti
ons
to
the
tour
is
t
de
pe
nding
on
the
type
of
plac
e
s
he
/s
he
pr
e
f
e
r
s
.
RE
F
E
RE
NC
E
S
[1
]
S.
K
an
t
o
l
a,
M.
U
u
s
i
t
al
o
,
V
.
N
i
v
a
l
a,
an
d
S.
T
u
u
l
en
t
i
e,
"
T
o
u
r
i
s
m
re
s
o
r
t
u
s
ers
’
p
ar
t
i
c
i
p
a
t
i
o
n
i
n
p
l
a
n
n
i
n
g
:
T
es
t
i
n
g
t
h
e
p
u
b
l
i
c
p
ar
t
i
c
i
p
at
i
o
n
g
e
o
g
ra
p
h
i
c
i
n
f
o
rmat
i
o
n
s
y
s
t
em
met
h
o
d
i
n
L
ev
i
,
F
i
n
n
i
s
h
L
ap
l
an
d
,
"
To
u
r
.
M
a
n
a
g
.
P
er
s
p
ect
.
,
v
o
l
.
2
7
,
p
p
.
2
2
–
3
2
,
2
0
1
8
.
[2
]
F.
Pi
cci
al
l
i
an
d
A
.
Ch
i
an
e
s
e,
"
T
h
e
i
n
t
ern
e
t
o
f
t
h
i
n
g
s
s
u
p
p
o
rt
i
n
g
co
n
t
ex
t
-
aw
are
co
m
p
u
t
i
n
g
:
a
cu
l
t
u
ral
h
eri
t
ag
e
c
as
e
s
t
u
d
y
,
"
M
o
b
.
Net
w
o
r
k
s
A
p
p
l
.
,
v
o
l
.
2
2
,
p
p
.
3
3
2
–
3
4
3
,
2
0
1
7
.
[3
]
S.
J
.
Mi
ah
,
H
.
Q
.
V
u
,
J
.
G
ammack
,
an
d
M.
McG
rat
h
,
“A
Bi
g
D
a
t
a
A
n
a
l
y
t
i
cs
Met
h
o
d
fo
r
T
o
u
r
i
s
t
Beh
av
i
o
u
r
A
n
a
l
y
s
i
s
,
”
I
n
f
.
M
a
n
a
g
.
,
v
o
l
.
5
4
,
n
o
.
6
,
p
p
.
7
7
1
–
7
8
5
,
2
0
1
7
.
[4
]
E
.
O
s
p
i
n
a,
F.
Mo
ren
o
,
an
d
I.
A
.
U
ri
b
e,
"
U
s
i
n
g
cri
t
eri
a
reco
n
s
t
r
u
ct
i
o
n
fo
r
l
o
w
-
s
am
p
l
i
n
g
t
ra
j
ect
o
ri
e
s
as
a
t
o
o
l
fo
r
an
al
y
t
i
cs
,
"
P
r
o
ced
i
a
Co
m
p
u
t
.
S
ci
.
,
v
o
l
.
5
1
,
p
p
.
3
6
6
–
3
7
3
,
2
0
1
5
.
[5
]
J
.
Z
h
an
g
,
T
.
W
u
,
an
d
Z
.
Fan
,
"
Res
earch
o
n
Prec
i
s
i
o
n
Mark
et
i
n
g
M
o
d
e
l
o
f
T
o
u
ri
s
m
In
d
u
s
t
ry
Ba
s
ed
o
n
U
s
e
r’s
Mo
b
i
l
e
Beh
a
v
i
o
r
T
ra
j
ect
o
ry
,
"
M
o
b
.
In
f
.
S
y
s
t
.
,
v
o
l
.
2
0
1
9
,
p
p
.
1
-
1
4
,
2
0
1
9
.
[6
]
W
.
Z
h
en
g
et
a
l
.
,
"
U
n
d
er
s
t
a
n
d
i
n
g
t
h
e
t
o
u
r
i
s
t
mo
b
i
l
i
t
y
u
s
i
n
g
G
PS:
H
o
w
s
i
m
i
l
ar
are
t
h
e
t
o
u
r
i
s
t
s
?,
"
To
u
r
.
M
a
n
a
g
.
,
v
o
l
.
7
1
,
p
p
.
5
4
–
6
6
,
2
0
1
9
.
[7
]
Z
.
Z
h
u
,
L
.
Sh
o
u
,
an
d
K
.
Ch
en
,
"
G
et
i
n
t
o
t
h
e
s
p
i
r
i
t
o
f
a
l
o
cat
i
o
n
b
y
mi
n
i
n
g
u
s
er
-
g
e
n
erat
e
d
t
rav
e
l
o
g
u
e
s
,
"
Neu
r
o
co
m
p
u
t
i
n
g
,
v
o
l
.
2
0
4
,
p
p
.
6
1
–
6
9
,
2
0
1
6
.
[8
]
R.
A
mj
ad
an
d
M.
S.
Cro
o
ck
,
"
D
o
mi
n
at
e
d
d
es
t
i
n
at
i
o
n
s
o
f
t
o
u
ri
s
t
i
n
s
i
d
e
i
raq
u
s
i
n
g
p
er
s
o
n
al
i
n
f
o
rmat
i
o
n
an
d
freq
u
e
n
cy
o
f
t
rav
e
l
,
"
T
E
LKO
M
NIK
A
Tel
eco
m
m
u
n
i
ca
t
i
o
n
C
o
m
p
u
t
i
n
g
E
l
ect
r
o
n
i
c
s
a
n
d
Co
n
t
r
o
l
,
v
o
l
.
1
7
,
n
o
.
4
,
p
p
.
1
7
2
3
-
1
7
3
0
,
2
0
1
9
.
[9
]
C.
K
.
K
e,
M.
Y
.
W
u
,
W.
C
.
H
o
,
S.
C.
L
ai
,
an
d
L
.
T
.
H
u
a
n
g
,
"
In
t
el
l
i
g
en
t
Po
i
n
t
-
of
-
In
t
eres
t
Reco
mmen
d
at
i
o
n
fo
r
T
o
u
ri
s
m
Pl
an
n
i
n
g
v
i
a
D
e
n
s
i
t
y
-
b
a
s
ed
C
l
u
s
t
er
i
n
g
an
d
G
en
et
i
c
A
l
g
o
ri
t
h
m,
"
P
A
CI
S
2
0
1
8
P
r
o
cee
d
i
n
g
s
,
2
0
1
8
.
[1
0
]
J
.
M.
L
u
o
,
H
.
Q
.
V
u
,
G
.
L
i
,
an
d
R.
L
aw
,
"
T
o
u
ri
s
t
b
e
h
av
i
o
r
an
a
l
y
s
i
s
i
n
g
ami
n
g
d
e
s
t
i
n
a
t
i
o
n
s
b
a
s
ed
o
n
v
e
n
u
e
ch
ec
k
-
in
d
at
a,
"
J.
Tr
a
vel
To
u
r
.
M
a
r
k.
,
v
o
l
.
3
6
,
n
o
.
1
,
p
p
.
1
0
7
–
1
1
8
,
2
0
1
9
.
[1
1
]
J
.
L
i
,
L
.
X
u
,
L
.
T
an
g
,
S.
W
an
g
,
an
d
L
.
L
i
,
"
Bi
g
d
at
a
i
n
t
o
u
r
i
s
m
res
earc
h
:
A
l
i
t
era
t
u
re
rev
i
ew
,
"
To
u
r
.
M
a
n
a
g
.
,
v
o
l
.
6
8
,
p
p
.
3
0
1
–
3
2
3
,
2
0
1
8
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
1
,
F
e
br
ua
r
y
2020
:
124
-
132
132
[1
2
]
Y
.
Z
h
en
g
,
H
.
Fu
,
X
.
X
i
e,
W
.
-
Y
.
Ma,
an
d
Q
.
L
i
,
"
G
eo
l
i
fe
G
PS
t
raj
ec
t
o
r
y
d
a
t
as
e
t
-
u
s
er
g
u
i
d
e,
"
M
i
c
r
o
s
o
f
t
R
es
.
,
[O
n
l
i
n
e],
A
v
ai
l
ab
l
e
o
n
l
i
n
e
h
t
t
p
s
/
/
w
w
w
.
mi
cr
o
s
o
ft
.
co
m/
e
n
u
s
/
re
s
earch
/
p
u
b
l
i
ca
t
i
o
n
/
g
e
o
l
i
fe
-
g
p
s
-
t
raj
ec
t
o
r
y
-
d
a
t
a
s
et
-
u
s
er
-
g
u
i
d
e,
2
0
1
1
.
[1
3
]
V
.
G
an
t
i
an
d
A
.
D
a
s
Sarma,
"
D
a
t
a
cl
ea
n
i
n
g
:
A
p
rac
t
i
cal
p
er
s
p
ec
t
i
v
e,
"
S
y
n
t
h
.
Lect
.
D
a
t
a
M
a
n
a
g
.
,
v
o
l
.
5
,
n
o
.
3
,
p
p
.
1
–
8
5
,
2
0
1
3
.
[1
4
]
S.
Ph
i
t
h
ak
k
i
t
n
u
k
o
o
n
,
T
.
H
o
ra
n
o
n
t
,
A
.
W
i
t
a
y
an
g
k
u
r
n
,
R.
Si
ri
,
Y
.
Sek
i
mo
t
o
,
a
n
d
R.
S
h
i
b
as
a
k
i
,
"
U
n
d
er
s
t
an
d
i
n
g
t
o
u
ri
s
t
b
e
h
av
i
o
r
u
s
i
n
g
l
arg
e
-
s
cal
e
m
o
b
i
l
e
s
e
n
s
i
n
g
ap
p
ro
ac
h
:
A
ca
s
e
s
t
u
d
y
o
f
mo
b
i
l
e
p
h
o
n
e
u
s
er
s
i
n
J
ap
an
,
"
P
er
v
a
s
i
ve
M
o
b
.
Co
m
p
u
t
.
,
v
o
l
.
1
8
,
p
p
.
1
8
–
3
9
,
2
0
1
5
.
[1
5
]
T
.
Mu
n
H
en
g
,
"
Mi
n
i
n
g
U
s
er
S
i
mi
l
ari
t
y
Ba
s
ed
o
n
L
o
ca
t
i
o
n
H
i
s
t
o
r
y
,
"
E
co
n
.
P
l
a
n
.
I
n
d
.
P
o
l
i
cy
G
l
o
b
.
E
c
o
n
.
Co
n
ce
p
t
s
,
E
xp
.
P
r
o
s
p
ec
t
.
,
n
o
.
c,
p
p
.
2
9
–
4
2
,
2
0
1
5
.
[1
6
]
D
.
A
.
Pei
x
o
t
o
,
"
Mi
n
i
n
g
T
ra
j
ect
o
ry
D
at
a,
"
p
p
.
1
–
2
3
,
2
0
1
3
.
[1
7
]
Y
.
Z
h
en
g
,
L
.
Z
h
an
g
,
X
.
X
i
e,
an
d
W
.
Y
.
Ma,
"
Mi
n
i
n
g
i
n
t
ere
s
t
i
n
g
l
o
c
at
i
o
n
s
an
d
t
rav
e
l
s
e
q
u
e
n
ces
fr
o
m
G
PS
t
raj
ec
t
o
r
i
es
,
"
P
r
o
c.
1
8
th
In
t
.
Co
n
f
.
W
o
r
l
d
wi
d
e
we
b
-
W
W
W
’0
9
,
2
0
0
9
.
[1
8
]
W
.
L
u
an
,
G
.
L
i
u
,
C.
J
i
an
g
,
an
d
L
.
Q
i
,
"
Part
i
t
i
o
n
-
b
as
e
d
co
l
l
ab
o
rat
i
v
e
t
e
n
s
o
r
fact
o
r
i
zat
i
o
n
fo
r
PO
I
reco
mmen
d
at
i
o
n
,
"
IE
E
E
/
C
A
A
J.
A
u
t
o
m
.
S
i
n
.
,
v
o
l
.
4
,
n
o
.
3
,
p
p
.
4
3
7
–
4
4
6
,
2
0
1
7
.
[1
9
]
M.
E
s
t
er,
H
.
P.
K
ri
e
g
el
,
J
.
San
d
er,
X
.
X
u
,
an
d
o
t
h
er
s
,
"
A
d
e
n
s
i
t
y
-
b
a
s
ed
a
l
g
o
ri
t
h
m
f
o
r
d
i
s
co
v
eri
n
g
c
l
u
s
t
er
s
i
n
l
a
rg
e
s
p
a
t
i
a
l
d
at
a
b
as
e
s
w
i
t
h
n
o
i
s
e,
"
i
n
KD
D
-
9
6
P
r
o
ceed
i
n
g
s
,
v
o
l
.
9
6
,
n
o
.
3
4
,
p
p
.
2
2
6
–
2
3
1
,
1
9
9
6
.
[2
0
]
J
.
San
d
er,
M.
E
s
t
er,
H
.
P.
K
r
i
eg
e
l
,
an
d
X
.
X
u
,
"
D
e
n
s
i
t
y
-
b
as
e
d
cl
u
s
t
eri
n
g
i
n
s
p
a
t
i
a
l
d
at
a
b
as
e
s
:
T
h
e
a
l
g
o
r
i
t
h
m
g
d
b
s
ca
n
an
d
i
t
s
ap
p
l
i
ca
t
i
o
n
s
,
"
D
a
t
a
M
i
n
.
Kn
o
wl
.
D
i
s
c
o
v.
,
v
o
l
.
2
,
n
o
.
2
,
p
p
.
1
6
9
–
1
9
4
,
1
9
9
8
.
[2
1
]
K
.
J
o
r
d
ah
l
,
"
G
eo
Pan
d
as
:
Py
t
h
o
n
t
o
o
l
s
f
o
r
g
eo
g
ra
p
h
i
c
d
a
t
a
,
"
[O
n
l
i
n
e],
A
v
ai
l
ab
l
e:
h
t
t
p
s
/
/
g
i
t
h
u
b
.
co
m/
g
eo
p
an
d
as
/
g
e
o
p
a
n
d
a
s
,
2
0
1
4
.
[2
2
]
F.
Ped
reg
o
s
a
et
a
l
.
,
"
Sci
k
i
t
-
l
ear
n
:
Mach
i
n
e
l
earn
i
n
g
i
n
P
y
t
h
o
n
,
"
J.
M
a
c
h
.
Lea
r
n
.
R
es
.
,
v
o
l
.
1
2
,
p
p
.
2
8
2
5
–
2
8
3
0
,
2
0
1
1
.
[2
3
]
Reb
el
e,
T
h
o
mas
,
Fab
i
an
Su
c
h
an
e
k
,
J
o
h
an
n
es
H
o
ffart
,
J
o
an
n
a
Bi
eg
a,
E
r
d
a
l
K
u
zey
,
an
d
G
erh
ard
W
ei
k
u
m.
"
Y
A
G
O
:
A
mu
l
t
i
l
i
n
g
u
a
l
k
n
o
w
l
ed
g
e
b
as
e
fro
m
w
i
k
i
p
e
d
i
a,
w
o
rd
n
e
t
,
an
d
g
eo
n
ames
.
"
In
t
e
r
n
a
t
i
o
n
a
l
S
em
a
n
t
i
c
W
e
b
Co
n
f
er
e
n
ce
,
p
p
.
1
7
7
-
1
8
5
,
2
0
1
6
.
[2
4
]
P.
J
.
Ro
u
s
s
eeu
w
,
"
Si
l
h
o
u
e
t
t
e
s
:
a
g
rap
h
i
c
al
ai
d
t
o
t
h
e
i
n
t
erp
re
t
at
i
o
n
an
d
v
a
l
i
d
at
i
o
n
o
f
cl
u
s
t
er
an
a
l
y
s
i
s
,
"
J.
Co
m
p
u
t
.
A
p
p
l
.
M
a
t
h
.
,
v
o
l
.
2
0
,
p
p
.
5
3
–
6
5
,
1
9
8
7
.
[2
5
]
T
.
K
.
D
an
g
,
N
.
T
h
o
a
i
,
an
d
o
t
h
ers
,
"
H
y
b
ri
d
s
t
o
p
d
i
s
c
o
v
e
ry
i
n
t
raj
ec
t
o
r
y
reco
r
d
s
,
"
2
0
1
3
2
4
th
I
n
t
e
r
n
a
t
i
o
n
a
l
W
o
r
ks
h
o
p
o
n
D
a
t
a
b
a
s
e
a
n
d
E
xp
e
r
t
S
ys
t
em
s
A
p
p
l
i
ca
t
i
o
n
s
,
2
0
1
3
.
Evaluation Warning : The document was created with Spire.PDF for Python.