I
AE
S
I
nte
rna
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
5
,
No
.
1
,
Ma
r
ch
2
0
1
6
,
p
p
.
41
~
44
I
SS
N:
2252
-
8938
41
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
jo
u
r
n
a
l.c
o
m/o
n
lin
e/in
d
ex
.
p
h
p
/I
J
AI
T
w
itt
er Tw
eet
Cla
ss
ifier
Ash
w
in V
De
p
a
rte
m
e
n
t
o
f
In
f
o
r
m
a
ti
o
n
T
e
c
h
n
o
l
o
g
y
,
S
RM
Un
iv
e
rsit
y
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Dec
4
,
2
0
1
5
R
ev
i
s
ed
Feb
7
,
2
0
1
6
A
cc
ep
ted
Feb
2
5
,
2
0
1
6
T
h
is
p
a
p
e
r
a
d
d
re
ss
e
s
th
e
ta
sk
o
f
b
u
il
d
in
g
a
c
las
si
f
ier
th
a
t
w
o
u
ld
c
a
te
g
o
rise
tw
e
e
ts
in
T
w
it
ter.
M
icro
b
lo
g
g
in
g
n
o
w
a
d
a
y
s
h
a
s
b
e
c
o
m
e
a
to
o
l
o
f
c
o
m
m
u
n
ica
ti
o
n
f
o
r
In
tern
e
t
u
se
rs.
T
h
e
y
sh
a
re
o
p
in
i
o
n
o
n
d
if
f
e
re
n
t
a
sp
e
c
ts
o
f
li
f
e
.
A
s
th
e
p
o
p
u
larity
o
f
th
e
m
i
c
ro
b
l
o
g
g
in
g
sites
in
c
re
a
se
s
th
e
c
lo
se
r
w
e
g
e
t
to
th
e
e
ra
o
f
In
f
o
rm
a
ti
o
n
Ex
p
lo
si
o
n
.
T
w
it
ter
is
th
e
se
c
o
n
d
m
o
st
u
se
d
m
icro
b
lo
g
g
in
g
site
w
h
ich
h
a
n
d
l
e
s
m
o
re
th
a
n
5
0
0
m
il
li
o
n
tw
e
e
ts
twe
e
ted
e
v
e
r
y
d
a
y
w
h
ich
tran
sla
tes
to
m
in
d
b
o
g
g
li
n
g
5
,
7
0
0
tw
e
e
ts
p
e
r
se
c
o
n
d
.
De
sp
i
te
th
e
h
u
m
o
n
g
o
u
s
u
sa
g
e
o
f
t
w
it
ter
th
e
re
isn
’t
a
n
y
sp
e
c
i
f
ic
c
las
si
f
ier
f
o
r
th
e
s
e
tw
e
e
ts
th
a
t
a
re
twe
e
ted
o
n
t
h
is
sit
e
.
T
h
is
re
se
a
rc
h
a
tt
e
m
p
ts
to
se
g
re
g
a
te t
w
e
e
t
s
a
n
d
c
las
sify
th
e
m
to
c
a
teg
o
ries
li
k
e
S
p
o
rts,
Ne
w
s,
En
terta
in
m
e
n
t,
T
e
c
h
n
o
lo
g
y
,
M
u
sic
,
T
V
,
M
e
m
e
,
e
tc
.
Na
ï
v
e
B
a
y
e
s,
a
m
a
c
h
in
e
lea
rn
in
g
a
lg
o
rit
h
m
is
u
se
d
f
o
r
b
u
il
d
in
g
a
c
las
si
f
ier
w
h
ich
c
las
si
f
ies
th
e
twe
e
ts
w
h
e
n
train
e
d
w
it
h
th
e
t
w
it
ter
c
o
rp
u
s.
W
it
h
th
is
k
in
d
o
f
c
la
ss
i
f
ier
th
e
u
se
r
m
a
y
sim
p
l
y
s
k
i
m
th
e
twe
e
ts
w
it
h
o
u
t
g
o
in
g
th
ro
u
g
h
th
e
te
d
io
u
s
w
o
rk
o
f
sk
i
m
m
in
g
th
e
n
e
w
s
f
e
e
d
.
K
ey
w
o
r
d
:
Mic
r
o
b
lo
g
g
in
g
T
w
ee
t
s
Ste
m
m
i
n
g
Ma
ch
i
n
e
lear
n
i
n
g
al
g
o
r
it
h
m
Co
p
y
rig
h
t
©
2
0
1
6
In
stit
u
te o
f
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
rin
g
a
n
d
S
c
ien
c
e
.
Al
l
rig
h
ts
re
se
rv
e
d
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Ash
w
i
n
V
,
Dep
t.
o
f
I
n
f
o
r
m
at
io
n
T
ec
h
n
o
l
o
g
y
,
SR
M
U
n
i
v
er
s
it
y
,
Kattan
k
u
lath
u
r
C
h
e
n
n
ai,
I
n
d
ia
.
E
m
ail:
9
4
.
ash
w
i
n
v
ee
@
g
m
ail.
c
o
m
1.
I
NT
RO
D
UCT
I
O
N
Su
cc
es
s
f
u
l
m
icr
o
b
lo
g
g
in
g
s
er
v
ices
s
u
c
h
as
t
w
itter
h
av
e
co
m
e
to
an
i
n
te
g
r
al
p
ar
t
o
f
d
ai
l
y
li
f
e
o
f
m
illi
o
n
s
o
f
I
n
ter
n
et
u
s
er
s
.
I
n
t
er
est
in
m
i
n
i
n
g
in
t
w
itter
i
n
cr
ea
s
ed
w
ith
th
e
w
id
esp
r
ea
d
u
s
e
o
f
th
e
s
e
s
er
v
ices.
B
ec
au
s
e
t
h
e
y
b
ec
o
m
e
th
e
w
ar
e
h
o
u
s
e
o
f
p
eo
p
les
o
p
in
io
n
o
n
c
u
r
r
en
t
is
s
u
e
s
.
T
w
itter
is
th
e
m
o
s
t
p
o
p
u
lar
a
n
d
th
e
s
ec
o
n
d
m
o
s
t
u
s
ed
s
o
cial
n
et
w
o
r
k
i
n
g
s
ite.
S
in
ce
it
s
la
u
n
ch
i
n
2
0
0
6
,
th
e
p
o
p
u
lar
it
y
o
f
its
u
s
e
h
as
b
ee
n
d
r
asti
ca
ll
y
in
cr
ea
s
in
g
,
t
h
at
m
o
r
e
th
an
1
0
0
m
illi
o
n
u
s
er
p
o
s
t
3
4
0
m
illi
o
n
t
w
ee
t
s
a
d
a
y
i
n
2
0
1
2
.
A
s
o
f
m
ar
c
h
2
0
1
6
,
T
w
itter
h
a
s
m
o
r
e
th
a
n
3
1
0
m
ill
io
n
m
o
n
t
h
l
y
ac
tiv
e
u
s
er
.
I
n
t
w
itter
,
u
s
er
s
ar
e
allo
w
ed
t
o
cr
ea
te
s
tatu
s
m
e
s
s
a
g
es
w
h
i
ch
ar
e
ca
lled
t
w
ee
ts
.
T
w
ee
ts
ar
e
o
f
1
4
0
ch
ar
ac
ter
s
w
h
ic
h
ar
e
w
r
itte
n
b
y
r
eg
is
ter
ed
t
w
itter
u
s
er
s
ab
o
u
t
th
eir
lif
e,
o
p
in
io
n
s
o
n
v
a
r
iet
y
o
f
to
p
ics
an
d
d
is
cu
s
s
io
n
s
o
n
c
u
r
r
en
t
i
s
s
u
es.
As
m
o
r
e
an
d
m
o
r
e
u
s
er
s
s
h
a
r
e
th
eir
o
p
in
io
n
o
n
s
ev
er
al
f
i
eld
s
,
th
eir
v
ie
w
o
n
cu
r
r
en
t is
s
u
es,
m
icr
o
b
lo
g
g
i
n
g
s
ites
b
ec
o
m
e
t
h
e
ass
e
ts
o
f
u
s
er
’
s
o
p
in
io
n
an
d
s
e
n
ti
m
e
n
ts
.
I
n
th
i
s
r
esea
r
ch
w
e
u
s
e
a
d
ataset
f
o
r
m
ed
b
y
co
llecti
n
g
t
w
i
tt
er
t
w
ee
ts
.
T
h
ese
t
w
ee
ts
ar
e
co
llected
an
d
ar
e
f
ed
to
th
e
class
if
ier
w
h
i
ch
clas
s
if
ies
t
h
e
m
in
to
s
ev
er
al
ca
teg
o
r
ies
li
k
e
Sp
o
r
ts
,
Ne
w
s
,
E
n
ter
tai
n
m
en
t,
P
o
liti
cs,
an
d
Me
m
e.
W
it
h
t
h
e
h
elp
o
f
t
h
ese
ca
teg
o
r
ies
u
s
er
m
a
y
j
u
s
t
h
av
e
to
ch
o
o
s
e
t
h
e
ca
teg
o
r
y
o
f
h
i
s
/
h
er
in
ter
est
s
.
2.
RE
L
AT
E
D
WO
RK
W
ith
t
h
e
in
cr
ea
s
e
in
p
o
p
u
lar
ity
o
f
s
o
cial
n
e
t
w
o
r
k
s
a
n
d
b
lo
g
s
,
an
al
y
s
is
a
n
d
m
i
n
in
g
h
as b
ec
o
m
e
a
f
ield
o
f
i
n
ter
est
f
o
r
m
an
y
r
esear
c
h
er
s
.
T
h
er
e
ar
e
a
lo
t
o
f
r
esear
ch
es
d
o
n
e
in
t
w
it
ter
s
u
c
h
as
u
s
er
cla
s
s
i
f
icatio
n
,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
IJ
-
AI
Vo
l.
5
,
No
.
1
,
Ma
r
ch
2
0
1
6
:
41
–
44
42
T
r
en
d
d
etec
tio
n
,
Sen
ti
m
e
n
t
cl
ass
i
f
icatio
n
,
tr
en
d
d
etec
tio
n
et
c.
B
u
t
th
i
s
r
esear
ch
m
ai
n
l
y
ca
r
r
ies
o
u
t
th
e
ta
s
k
o
f
s
eg
r
e
g
atin
g
t
w
ee
t
s
i
n
to
s
ev
er
a
l c
ateg
o
r
ies f
o
r
u
s
er
’
s
co
n
v
e
n
i
en
ce
.
I
n
o
v
er
v
ie
w
in
g
L
.
H
u
ag
,
R
.
B
h
a
y
an
i
an
d
A
.
Go
v
in
d
,
2
0
0
9
r
e
s
ea
r
ch
,
t
h
e
y
ad
d
r
ess
ed
t
h
e
cla
s
s
i
f
icatio
n
o
f
t
w
ee
ts
b
ased
o
n
s
e
n
ti
m
e
n
t.
T
h
ese
t
w
ee
ts
ar
e
cla
s
s
i
f
ied
i
n
t
o
eith
er
p
o
s
iti
v
e
o
r
n
e
g
ati
v
e
with
r
esp
ec
t to
q
u
er
y
ter
m
.
T
h
e
y
h
a
v
e
al
s
o
u
s
ed
th
e
e
m
o
tico
n
s
o
cc
u
r
r
in
g
in
t
w
ee
ts
as
“
s
ilv
er
”
lab
els i.e
.
,
lab
els
w
it
h
m
o
r
e
u
n
ce
r
tai
n
s
tatu
s
t
h
an
t
h
e
o
n
e
s
f
o
u
n
d
i
n
u
s
u
a
l
“g
o
ld
”
s
ta
n
d
ar
d
s
f
o
r
t
w
ee
t
s
en
ti
m
en
t
a
n
al
y
s
i
s
.
T
h
is
r
esear
c
h
m
a
in
l
y
b
en
ef
it
s
co
n
s
u
m
er
s
w
h
o
w
an
t
to
k
n
o
w
ab
o
u
t
t
h
e
s
e
n
ti
m
e
n
t
o
f
p
r
o
d
u
cts
b
ef
o
r
e
p
u
r
c
h
ase,
o
r
co
m
p
a
n
ies
th
a
t
w
a
n
t
to
m
o
n
i
to
r
th
e
p
u
b
lic
s
e
n
ti
m
e
n
t
o
f
th
eir
b
r
an
d
.
W
h
ile
J
.
R
ea
d
in
h
is
r
esear
ch
(
R
ea
d
,
2
0
0
5
)
h
e
h
as
u
s
ed
e
m
o
tico
n
s
s
u
c
h
as
“
:
-
)
”
an
d
“
:
-
(
”
w
h
ic
h
s
er
v
e
as
n
o
i
s
y
lab
els
an
d
f
o
r
m
ed
a
tr
ain
i
n
g
s
et
f
o
r
s
en
t
i
m
e
n
t
class
i
f
icatio
n
.
Fo
r
t
h
is
p
u
r
p
o
s
e,
th
e
a
u
t
h
o
r
co
llected
te
x
ts
c
o
n
tain
i
n
g
e
m
o
tico
n
s
a
n
d
d
iv
i
d
ed
th
e
d
ata
s
e
t
i
n
to
p
o
s
itiv
e
an
d
n
eg
at
iv
e
s
a
m
p
le
s
San
k
ar
an
ar
a
y
a
n
a
n
et
al.
2
0
0
9
,
b
u
ilt
a
n
e
w
s
p
r
o
ce
s
s
i
n
g
s
y
s
te
m
f
o
r
id
en
ti
f
y
i
n
g
t
h
e
t
w
ee
t
s
th
a
t
co
r
r
esp
o
n
d
s
to
late
b
r
ea
k
in
g
n
e
w
s
.
T
h
e
y
co
llected
th
e
t
wee
ts
an
d
th
e
y
r
e
m
o
v
ed
n
o
is
e
s
f
r
o
m
th
e
m
.
T
h
e
y
c
lu
s
ter
ed
t
h
e
t
w
ee
ts
u
s
i
n
g
t
h
e
cl
u
s
ter
i
n
g
al
g
o
r
ith
m
ca
lle
d
lead
er
-
f
o
llo
w
er
cl
u
s
ter
i
n
g
,
w
h
ic
h
allo
w
s
f
o
r
clu
s
ter
i
n
g
i
n
b
o
th
co
n
ten
t
a
n
d
ti
m
e.
T
h
e
o
th
er
is
s
u
e
w
h
ic
h
i
s
ad
d
r
ess
ed
in
t
h
i
s
r
esear
ch
is
i
d
en
tify
i
n
g
r
elev
a
n
t
lo
ca
tio
n
s
as
s
o
ciate
d
w
it
h
t
h
e
t
w
ee
t
s
.
P
a
n
g
a
n
d
L
ee
,
2
0
0
2
r
esear
ch
ed
th
e
p
er
f
o
r
m
a
n
ce
o
f
v
ar
io
u
s
m
ac
h
i
n
e
lear
n
in
g
tech
n
iq
u
es
lik
e
Nai
v
e
B
a
y
es,
Ma
x
i
m
u
m
e
n
tr
o
p
y
a
n
d
SVM
i
n
t
h
e
s
p
ec
i
fi
c
d
o
m
ai
n
o
f
m
o
v
ie
r
ev
ie
w
s
.
B
y
t
h
i
s
th
e
y
w
er
e
ab
le
to
ac
h
i
ev
e
an
ac
c
u
r
ac
y
o
f
8
2
%.
Srir
a
m
et
al.
2
0
1
0
,
th
eir
w
o
r
k
is
m
o
r
e
r
ele
v
a
n
t
to
o
u
r
s
.
T
h
ey
clas
s
i
f
ied
t
w
ee
t
s
i
n
to
p
r
ed
ef
in
ed
s
et
o
f
class
es
s
u
c
h
as
n
e
w
s
,
e
v
e
n
ts
,
o
p
in
io
n
,
d
ea
ls
an
d
p
r
iv
ate
m
es
s
ag
e
s
w
it
h
t
h
e
u
s
e
o
f
i
n
f
o
r
m
at
io
n
ab
o
u
t
t
h
e
au
th
o
r
a
n
d
also
f
ea
tu
r
e
w
h
ic
h
ar
e
ex
tr
ac
ted
f
r
o
m
t
w
ee
t
s
s
u
ch
as
“@
u
s
er
n
a
m
e”
,
s
h
o
r
te
n
i
n
g
o
f
w
o
r
d
s
,
s
la
n
g
,
etc.
T
h
e
y
class
i
f
ied
t
w
ee
ts
i
n
o
r
d
er
to
im
p
r
o
v
e
in
f
o
r
m
atio
n
f
ilter
i
n
g
.
T
h
eir
f
ea
tu
r
e
o
u
tp
er
f
o
r
m
ed
th
e
b
ag
.
O
f
w
o
r
d
m
o
d
el
ap
p
r
o
ac
h
in
th
e
cl
ass
i
f
icatio
n
o
f
t
w
ee
t
s
.
Dif
f
er
en
t
f
r
o
m
t
h
e
ab
o
v
e
w
o
r
k
s
,
th
i
s
r
esear
ch
d
ef
i
n
e
s
a
s
y
s
te
m
t
h
at
w
o
u
l
d
clas
s
i
f
y
all
th
e
t
w
ee
t
s
ir
r
esp
ec
tiv
e
o
f
th
e
tr
e
n
d
,
s
en
ti
m
en
t,
u
s
er
i
n
to
ca
teg
o
r
ies
li
k
e
Ne
w
s
,
Me
m
e,
Sp
o
r
t,
E
n
ter
tain
m
e
n
t
w
h
ic
h
w
o
u
ld
m
ak
e
t
w
itter
f
ar
m
o
r
e
co
n
v
e
n
i
en
t to
u
s
e
a
n
d
also
s
a
v
es a
m
p
l
e
a
m
o
u
n
t o
f
ti
m
e
f
o
r
th
e
u
s
er
.
3.
SYST
E
M
DE
SI
G
N
Ou
r
s
y
s
te
m
h
er
e
i
s
s
e
g
r
eg
ated
in
to
d
if
f
er
en
t
m
o
d
u
le
s
.
3
.
1
.
T
w
ee
t
Ret
riev
a
l
M
o
du
le
T
h
e
b
asic r
eq
u
ir
em
e
n
t
f
o
r
th
e
r
etr
iev
al
o
f
T
w
ee
t
s
is
t
h
e
T
w
it
ter
A
P
I
.
R
eg
i
s
tr
atio
n
f
o
r
th
e
AP
I
is
d
o
n
e
u
s
i
n
g
an
e
x
is
ti
n
g
T
w
itter
ac
c
o
u
n
t.
O
n
ce
r
eg
is
ter
ed
,
th
e
u
s
e
r
is
p
r
o
v
id
ed
w
it
h
a
C
o
n
s
u
m
e
r
Key
,
a
C
o
n
s
u
m
er
Secr
et
Ke
y
,
an
A
cc
es
s
T
o
k
en
a
n
d
an
A
cc
es
s
T
o
k
en
Secr
et
u
s
in
g
w
h
ic
h
th
e
t
w
ee
ts
ar
e
r
etr
iev
ed
f
r
o
m
t
h
e
u
s
er
’
s
ti
m
eli
n
e.
3
.
2
.
T
ex
t
P
ro
ce
s
s
ing
M
o
du
le
T
o
an
al
y
s
e
an
d
class
if
y
tex
t,
th
er
e
ar
e
ce
r
tain
p
r
e
-
r
eq
u
is
ite
ac
tio
n
s
th
at
m
u
s
t
b
e
p
er
f
o
r
m
ed
.
T
h
e
r
etr
iev
ed
t
w
ee
ts
ar
e
w
r
itte
n
o
n
to
a
T
ex
t
Fil
e.
E
ac
h
t
w
ee
t
m
u
s
t
b
e
w
r
itte
n
o
n
to
a
d
if
f
e
r
en
t
tex
t
f
ile,
all
o
f
w
h
ic
h
m
u
s
t
b
e
i
n
t
h
e
s
a
m
e
d
ir
ec
to
r
y
.
T
h
ese
te
x
t
f
i
les ar
e
th
e
Do
cu
m
en
ts
th
a
t
w
i
ll b
e
u
s
ed
.
T
h
e
d
o
cu
m
e
n
ts
ar
e
f
ir
s
t
p
u
t
th
r
o
u
g
h
t
h
e
p
r
o
ce
s
s
o
f
clea
n
in
g
.
C
lea
n
i
n
g
r
ef
er
s
t
o
th
e
r
e
m
o
v
al
o
f
an
y
a
n
d
all
p
u
n
ct
u
atio
n
m
ar
k
s
f
r
o
m
w
it
h
i
n
t
h
e
d
o
cu
m
e
n
t.
O
n
ce
th
e
d
o
cu
m
e
n
ts
ar
e
clea
n
e
d
,
th
e
n
ex
t
p
r
o
ce
s
s
is
to
r
e
m
o
v
e
th
e
Sto
p
W
o
r
d
s
.
Sto
p
W
o
r
d
s
ar
e
w
o
r
d
s
lik
e
a
r
ticles,
p
r
o
n
o
u
n
s
,
p
r
ep
o
s
itio
n
s
an
d
co
n
j
u
n
ctio
n
s
w
h
ic
h
w
o
u
ld
n
o
t
af
f
ec
t
th
e
ev
en
t
u
al
cla
s
s
i
f
icatio
n
.
T
h
en
,
s
te
m
m
i
n
g
w
h
ic
h
is
t
h
e
p
r
o
c
ess
o
f
ex
tr
ac
tin
g
th
e
r
o
o
t
w
o
r
d
f
r
o
m
ea
ch
o
f
th
e
w
o
r
d
s
in
t
h
e
d
o
cu
m
en
t
is
ca
r
r
ied
o
u
t.
Ste
m
m
i
n
g
r
et
u
r
n
s
a
s
et
o
f
r
o
o
t
w
o
r
d
s
th
at
ca
n
t
h
en
b
e
f
ed
i
n
to
th
e
cla
s
s
i
f
ier
.
3
.
3
.
Co
nv
er
s
io
n
M
o
du
le
T
h
e
co
n
v
er
s
io
n
t
h
at
h
as
to
o
cc
u
r
is
to
g
et
a
n
AR
FF
Fi
le
to
b
e
f
ed
in
to
W
ek
a
i.e
.
,
W
aik
ato
E
n
v
ir
o
n
m
e
n
t
f
o
r
K
n
o
w
led
g
e
A
n
al
y
s
is
w
h
ich
is
a
Ma
c
h
i
n
e
L
ea
r
n
in
g
to
o
l
d
e
v
elo
p
ed
at
th
e
U
n
iv
er
s
it
y
o
f
W
aik
ato
in
Ne
w
Z
ea
la
n
d
.
T
h
e
w
h
o
le
d
ir
ec
to
r
y
o
f
Do
cu
m
en
ts
is
co
n
v
er
ted
i
n
to
o
n
e
AR
FF
Fil
e
u
s
in
g
t
h
e
W
ek
a’
s
C
o
r
e.
C
o
n
v
er
ter
s
L
ib
r
ar
y
.
T
h
e
co
n
v
er
s
io
n
is
d
o
n
e
u
s
in
g
th
e
TextDi
r
ec
to
r
yLo
a
d
er
m
et
h
o
d
th
at
d
er
iv
e
s
f
r
o
m
t
h
e
TextTo
A
r
ff
cla
s
s
i
n
W
ek
a.
T
h
e
co
m
m
an
d
w
ek
a
.
co
r
e.
co
n
ve
r
ters
.
Text
Dir
ec
to
r
y
Lo
a
d
er
d
ir
<Dir
ec
to
r
y
P
a
th
>
>
<A
R
F
F
F
ile
P
a
th
>
m
u
s
t
b
e
en
ter
ed
i
n
W
e
k
a’
s
C
o
m
m
a
n
d
L
in
e
I
n
ter
f
ac
e
w
h
ich
ca
r
r
ies
o
u
t
t
h
e
co
n
v
er
s
io
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
-
AI
I
SS
N:
2252
-
8938
T
w
it
ter T
w
e
e
t C
lass
if
ier
(
Ashw
in V
)
43
3
.
4
Cla
s
s
if
ier
M
o
du
le
W
e
b
u
ild
a
cla
s
s
i
f
ier
u
s
in
g
t
h
e
Naï
v
e
B
a
y
es
clas
s
i
f
ier
.
N
aıv
e
B
a
y
e
s
cla
s
s
i
fier
i
s
b
ased
o
n
B
a
y
e
s
’
th
eo
r
e
m
.
(
1
)
W
h
er
e
d
is
th
e
Do
cu
m
en
t,
an
d
c
is
th
e
C
ate
g
o
r
y
.
T
h
e
b
est cla
s
s
in
NB
clas
s
if
ica
tio
n
is
t
h
e
m
o
s
t lik
e
l
y
o
r
ma
ximu
m
a
p
o
s
teri
o
r
i
(
M
A
P
)
C
M
AP
class
:
(
2
)
Do
cu
m
e
n
t d
is
r
ep
r
esen
ted
as
f
ea
t
u
r
e
d
1
,
d
2
…,
d
n
.
T
h
e
class
if
ier
w
o
r
k
s
a
f
ter
th
e
A
ttrib
u
te
s
(
C
las
s
i
f
icatio
n
Fe
atu
r
e)
ar
e
m
e
n
tio
n
ed
o
n
th
e
AR
FF
F
ile.
Of
t
h
e
d
o
cu
m
en
ts
i
n
th
e
w
h
o
le
s
et,
7
0
%
ar
e
u
s
ed
as
th
e
tr
ain
i
n
g
s
et,
i.e
.
,
th
ese
7
0
%
o
f
th
e
t
w
ee
t
s
ar
e
m
a
n
u
al
l
y
lab
elled
as
Sp
o
r
ts
,
New
s
,
E
n
te
r
tain
m
en
t,
T
ec
h
n
o
lo
g
y
,
Mu
s
ic
,
T
V,
Me
m
e,
etc.
an
d
th
e
r
e
m
ain
i
n
g
3
0
%
is
u
s
ed
as th
e
T
est s
et.
3
.
5
.
User
M
o
d
ule
T
h
is
ca
n
b
e
ca
lled
th
e
f
r
o
n
t
-
e
n
d
o
f
th
e
ap
p
licatio
n
.
T
h
is
i
n
c
lu
d
es
t
h
e
Gr
ap
h
ic
User
I
n
ter
f
ac
e
(
GUI
)
.
T
h
is
p
r
o
v
id
es
th
e
lin
k
b
et
w
ee
n
th
e
ap
p
licatio
n
a
n
d
th
e
u
s
er
.
T
h
is
th
e
m
o
d
u
le
w
h
er
e
th
e
u
s
er
w
ill
b
e
d
is
p
la
y
ed
w
it
h
t
w
ee
t
s
t
h
at
ar
e
ca
teg
o
r
is
e
d
as Sp
o
r
t
s
,
E
n
ter
tain
m
en
t,
New
s
,
Me
m
e,
P
r
iv
ate
Me
s
s
ag
e.
4.
RE
SU
L
T
T
h
is
g
r
ap
h
s
h
o
w
s
t
h
e
ac
c
u
r
ac
ies
p
er
class
es
u
s
i
n
g
N
aïv
e
B
a
y
es
c
lass
if
ica
tio
n
.
T
h
is
g
r
ap
h
s
u
m
m
ar
izes
th
e
r
es
u
lt
o
f
o
u
r
c
lass
i
f
ier
.
T
h
e
ac
c
u
r
ac
ies
in
cla
s
s
i
f
icatio
n
i
n
to
ca
t
g
o
r
ies
ar
e
p
r
ett
y
h
i
g
h
w
h
ic
h
i
s
v
er
y
g
o
o
d
.
T
h
e
class
i
f
icatio
n
o
f
t
w
ee
ts
i
n
to
ca
te
g
o
r
ies
ar
e
o
n
l
y
b
ased
o
n
t
h
e
w
o
r
d
s
co
n
t
ain
ed
in
t
h
e
t
w
ee
t
s
th
at
w
e
u
s
ed
i
n
t
h
e
tr
ain
i
n
g
s
e
t
.
T
h
ese
w
o
r
d
s
m
a
y
co
n
tr
ad
ict
in
m
o
r
e
t
h
an
o
n
e
ca
te
g
o
r
ies
wh
ich
w
ill
a
f
f
ec
t
t
h
e
ac
cu
r
ac
y
.
B
u
t i
n
o
u
r
ca
s
e
t
h
at
i
m
p
ac
t
h
as b
ee
n
v
er
y
lo
w
ex
c
ep
t f
o
r
th
e
ca
teg
o
r
y
Ot
h
er
s
b
e
ca
u
s
e
th
e
a
m
b
i
g
u
it
y
in
t
w
ee
ts
t
h
at
f
al
l
u
n
d
er
t
h
e
o
t
h
er
s
ca
te
g
o
r
y
i
s
v
er
y
h
i
g
h
a
n
d
s
o
th
e
y
f
all
in
to
s
o
m
e
o
th
er
c
ateg
o
r
y
r
es
u
lti
n
g
i
n
th
is
lo
w
ac
c
u
r
ac
y
.
T
h
er
e
ar
e
s
ev
er
al
ap
p
r
o
ac
h
es lik
e
8
F,
B
OW
,
9
F th
at
ca
n
b
e
u
s
ed
to
in
cr
ea
s
e
th
e
ac
c
u
r
ac
y
.
Fig
u
r
e
1
.
A
cc
u
r
ac
ies
p
er
ca
teg
o
r
y
u
s
i
n
g
Naïv
e
B
a
y
es
5.
F
UT
UR
E
E
NH
AN
CE
M
E
N
T
Miss
io
n
lear
n
in
g
tec
h
n
iq
u
e
s
p
er
f
o
r
m
w
ell
f
o
r
clas
s
i
f
icatio
n
o
f
t
w
ee
ts
.
W
e
b
elie
v
e
t
h
at
ac
cu
r
ac
y
co
u
ld
s
till
b
e
i
m
p
r
o
v
ed
.
Mo
r
e
ad
v
an
ce
d
co
m
p
ar
is
o
n
ap
p
r
o
ac
h
es
ca
n
b
e
tak
en
i
n
to
co
n
s
id
er
atio
n
s
u
c
h
as
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
IJ
-
AI
Vo
l.
5
,
No
.
1
,
Ma
r
ch
2
0
1
6
:
41
–
44
44
clu
s
ter
i
n
g
.
A
n
d
also
u
s
i
n
g
m
a
ch
in
e
lear
n
i
n
g
a
lg
o
r
it
h
m
s
li
k
e
SVM,
Ma
x
i
m
u
m
en
tr
o
p
y
ca
n
b
e
u
s
ed
in
ad
d
itio
n
to
th
e
Naïv
e
B
a
y
e
s
clas
s
i
f
ier
to
im
p
r
o
v
e
t
h
e
ef
f
ic
ien
c
y
in
cla
s
s
i
f
icatio
n
.
As
m
e
n
tio
n
ed
ea
r
lier
,
t
w
itt
er
h
as
5
0
0
m
illi
o
n
t
w
ee
t
s
ev
er
y
d
a
y
,
h
e
n
ce
in
n
ea
r
f
u
tu
r
e
t
h
i
s
m
i
n
d
b
o
g
g
li
n
g
n
u
m
b
er
is
b
o
u
n
d
to
in
cr
ea
s
e
an
d
s
u
b
s
eq
u
en
t c
ateg
o
r
ies n
ee
d
to
b
e
ac
co
r
d
in
g
l
y
i
m
p
le
m
e
n
ted
.
6.
CO
NCLU
SI
O
N
As
d
is
c
u
s
s
ed
ea
r
lier
m
icr
o
b
lo
g
g
i
n
g
n
o
w
ad
a
y
s
h
as
b
e
co
m
e
o
n
e
o
f
t
h
e
m
aj
o
r
m
o
d
e
s
o
f
co
m
m
u
n
icatio
n
a
m
o
n
g
i
n
ter
n
et
u
s
er
s
.
A
r
ec
en
t
r
esear
ch
h
a
s
id
en
ti
fi
ed
it
as
o
n
li
n
e
w
o
r
d
-
of
-
m
o
u
t
h
b
r
an
d
in
g
(
J
an
s
en
et
al.
,
2
0
0
9
)
.
T
h
e
lar
g
e
a
m
o
u
n
t
o
f
in
f
o
r
m
a
tio
n
co
n
t
ain
ed
in
m
icr
o
b
lo
g
g
i
n
g
w
eb
-
s
ites
is
w
h
at
m
a
k
e
s
th
e
m
a
n
at
tr
ac
tiv
e
s
o
u
r
ce
o
f
d
ata
f
o
r
m
i
n
i
n
g
an
d
an
al
y
s
i
s
.
A
lt
h
o
u
g
h
t
w
itter
m
e
s
s
ag
es
h
a
v
e
u
n
iq
u
e
ch
ar
ac
ter
is
tic
co
m
p
ar
ed
to
o
t
h
er
co
r
p
o
r
a,
m
ac
h
in
e
lear
n
i
n
g
al
g
o
r
ith
m
h
av
e
s
h
o
w
n
to
c
lass
i
f
y
t
w
ee
t
s
w
it
h
s
i
m
ilar
p
er
f
o
r
m
an
ce
.
I
n
th
i
s
p
ap
er
w
e
u
s
e
t
w
itter
f
ee
d
s
as
o
u
r
d
ata
s
et
an
d
ca
t
eg
o
r
ies
t
h
e
m
b
ased
o
n
th
eir
n
atu
r
e
a
n
d
s
ig
n
i
f
ica
n
ce
o
f
t
w
ee
t.
Fu
r
t
h
er
r
esear
ch
ed
is
n
ee
d
ed
to
c
o
n
tin
u
e
to
i
m
p
r
o
v
e
t
h
e
ac
cu
r
ac
y
i
n
d
if
f
icu
lt
d
o
m
ain
.
Ma
ch
i
n
e
lear
n
i
n
g
al
g
o
r
it
h
m
c
an
ac
h
ie
v
e
h
ig
h
ac
cu
r
ac
y
f
o
r
c
lass
i
f
y
in
g
w
h
en
u
s
in
g
t
h
is
m
et
h
o
d
.
Fig
u
r
e
2
.
S
y
s
te
m
De
s
ig
n
RE
F
E
R
E
NC
E
S
[1
]
J S
a
n
k
a
ra
n
a
ra
y
a
n
a
n
,
H S
a
m
e
t,
BE
T
e
it
ler,
M
D L
ieb
e
r
m
a
n
,
J
S
p
e
rl
in
g
.
T
w
it
ters
tan
d
:
n
e
w
s in
tw
e
e
ts
.
2
0
0
9
.
[2
]
B
S
riram
,
D
F
u
h
ry
,
E
De
m
ir,
H
F
e
rh
a
to
sm
a
n
o
g
lu
,
M
De
m
irb
a
s
.
S
h
o
rt
tex
t
c
las
sifi
c
a
ti
o
n
in
twit
ter
to
im
p
ro
v
e
in
f
o
rm
a
ti
o
n
fi
lt
e
rin
g
.
2
0
1
0
.
[3
]
A
G
o
,
R
Bh
a
y
a
n
i,
L
Hu
a
n
g
.
Tw
it
t
e
r
se
n
ti
m
e
n
t
c
las
si
fi
c
a
t
io
n
u
sin
g
d
istan
t
su
p
e
rv
isio
n
.
2
0
0
9
.
[4
]
M
Bu
sh
,
I
L
e
e
,
T
W
u
.
NL
P
-
b
a
se
d
a
p
p
ro
a
c
h
to
T
w
it
ter Us
e
r
Clas
si
f
i
c
a
ti
o
n
.
2
0
1
0
.
[5
]
J
Re
a
d
.
Us
in
g
e
m
o
ti
c
o
n
s to
re
d
u
c
e
d
e
p
e
n
d
e
n
c
y
in
m
a
c
h
in
e
lea
rn
in
g
tec
h
n
iq
u
e
s f
o
r
se
n
ti
m
e
n
t
c
las
sifi
c
a
ti
o
n
.
2
0
0
5
.
[6
]
Ka
th
y
L
e
e
,
Dia
n
a
P
a
lse
ti
a
,
Ra
m
a
n
a
th
a
n
Na
ra
y
a
n
a
n
,
M
d
.
M
o
sto
f
a
A
li
P
a
tw
a
r
y
,
A
n
k
it
Ag
ra
w
a
l,
a
n
d
A
lo
k
Ch
o
u
d
h
a
ry
.
Tw
it
ter T
re
n
d
in
g
T
o
p
ic Cl
a
ss
if
ica
ti
o
n
.
2
0
0
9
.
[7
]
J Be
n
h
a
rd
u
s
.
S
trea
m
in
g
T
re
n
d
De
tec
ti
o
n
i
n
T
w
it
ter
.
2
0
1
0
.
[8
]
B
P
a
n
g
,
L
L
e
e
,
S
V
a
it
h
y
a
n
a
th
a
n
.
S
e
n
ti
m
e
n
t
c
las
sifi
c
a
ti
o
n
u
si
n
g
m
a
c
h
in
e
lea
rn
i
n
g
tec
h
n
i
q
u
e
s
.
2
0
0
2
.
[9
]
S
k
a
b
ir
a
n
d
M
T
a
h
m
i
m
.
L
o
c
a
li
z
e
d
tw
it
ter o
p
i
n
i
o
n
m
in
in
g
u
si
n
g
se
n
ti
m
e
n
t
a
n
a
ly
sis
.
2
0
0
9
.
[1
0
]
T
e
d
P
e
d
e
rse
n
.
A
sim
p
le
a
p
p
ro
a
c
h
to
b
u
i
ld
i
n
g
e
n
se
m
b
les
o
f
n
a
iv
e
b
a
y
e
sia
n
c
las
si
fi
e
rs
fo
r
w
o
rd
se
n
se
d
isa
m
b
ig
u
a
ti
o
n
.
2
0
0
0
.
[1
1
]
YS
Ye
g
in
G
e
n
c
,
JV
Nic
k
e
rso
n
.
Disc
o
v
e
rin
g
c
o
n
tex
t:
Cla
ss
if
y
in
g
twee
ts
th
ro
u
g
h
a
se
ma
n
ti
c
tra
n
sfo
rm
b
a
se
d
o
n
W
ikip
e
d
ia
.
P
ro
c
e
e
d
in
g
s
o
f
HCI In
tern
a
ti
o
n
a
l
.
2
0
1
1
.
[1
2
]
T
M
M
it
c
h
e
ll
.
M
a
c
h
i
n
e
L
e
a
rn
in
g
.
M
c
G
ra
w
-
Hill
,
Ne
w
Yo
rk
.
19
97.
[1
3
]
W
e
k
a
3
:
Da
ta M
in
i
n
g
S
o
f
tw
a
re
in
Ja
v
a
,
h
tt
p
:/
/w
ww
.
c
s.
w
a
i
k
a
to
.
a
c
.
n
z
/m
l/
w
e
k
a
/.
[1
4
]
Na
ïv
e
Ba
y
e
s cla
ss
i
f
ier.
Re
tri
e
v
e
d
f
ro
m
h
tt
p
s://
w
e
b
.
sta
n
f
o
rd
.
e
d
u
/cla
ss
/cs
1
2
4
/l
e
c
/n
a
iv
e
b
a
y
e
s.p
d
f
[1
5
]
Na
iv
e
B
a
y
e
s
tex
t
c
l
a
ss
i
f
ica
ti
o
n
.
Re
tri
e
v
e
d
f
ro
m
h
tt
p
:/
/
n
lp
.
sta
n
f
o
rd
.
e
d
u
/IR
-
b
o
o
k
/h
tm
l/
h
tm
led
it
io
n
/n
a
iv
e
-
b
a
y
e
s
-
tex
t
-
c
las
si
f
ica
ti
o
n
-
1
.
h
tm
l.
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
-
AI
I
SS
N:
2252
-
8938
T
w
it
ter T
w
e
e
t C
lass
if
ier
(
Ashw
in V
)
45
B
I
B
L
I
O
G
R
AP
H
Y
O
F
AUT
H
O
R
A
sh
w
in
V
is
f
ro
m
Ch
e
n
n
a
i
b
o
r
n
o
n
0
5
/
1
0
/1
9
9
4
Ha
s
c
o
m
p
lete
d
Un
d
e
rg
ra
d
u
a
te
B.
T
e
c
h
d
e
g
re
e
in
In
f
o
rm
a
ti
o
n
T
e
c
h
n
o
lo
g
y
a
t
S
RM
Un
iv
e
rsit
y
Ch
e
n
n
a
i,
In
d
ia
in
th
e
y
e
a
r
o
f
2
0
1
6
.
A
n
d
c
u
rre
n
tl
y
w
o
rk
in
g
a
t
E
m
e
rio
T
e
c
h
n
o
l
o
g
ies
,
G
u
in
d
y
,
Ch
e
n
n
a
i,
I
n
d
ia.
T
h
is
is
th
e
A
u
th
o
r’s f
irst
re
se
a
rc
h
w
o
rk
.
Evaluation Warning : The document was created with Spire.PDF for Python.