4.2. WEAKLY SUPERVISED FAKE NEWS DETECTION 63
F
F 2 R
N d
G
G
a
i
a
j
l
2
K
d.F
i
; F
j
/ D kF
i
F
j
k
2
:
G
N N O
Belief Propagation
G G
ŒI C ˛D c
0
Ob
h
D
h
;
h
b
h
O
I D D
ii
D
P
j
O
ij
D
ij
D 0
i ¤ j
4.2.2 A TENSOR DECOMPOSITION UNSUPERVISED APPROACH
64 4. CHALLENGING PROBLEMS OF FAKE NEWS DETECTION
A D fa
1
; a
2
; ; a
N
g a
i
† d D j†j
X
A1
A1 A1 A2 A3 A3 C5 C5A2
A2 A3
B1 B2B3
C1C2 C3 C4C5
B4
0.8
0.7
0.6
0.5
0.8
0.7
0.6
0.5
n
1
n
2
n
3
n
4
n
5
n
6
n
7
n
8
n
9
0.8
0.7
0.6
0.5
0.8
0.7
0.6
0.5
Co-cluster
Rank = 4
Rank = 5
Rank = 3
Terms
Documents
–
.
.
.
.
.
.
.
.
.
Terms
Tensor Ensemble Co-Clustering
X 2 R
N d d
a X
F
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.