what-when-how
In Depth Tutorials and Information
In Reference 25 the opinion expressed by node
j
about given predicate
p
and the
opinion of node
i
about
j
are denoted as
O
p
j
and
O
i
, respectively. hus, the opin-
ion of node
i
about
p
is calculated by the recommendation operator denoted by
�
[25]:
ij
j
j
j
j
O
=
O O
i
⊗ =
b b b d
i
,
i
,
d
i
+
u
i
+
b u
i
(11.23)
p
p
p
p
p
From the equation we can see that the result of recommendation is based on the
order of the opinions. he joint opinion of node
i
and node
j
about the given predi-
cate is calculated by the consensus operator denoted by
�
[25]:
ij
j
j
j
j
j
j
k u u k
=
⊕ =
+
+
(11.24)
O
O O
i
(
b u
i
b u
i
)
k d u
,(
i
d u
i
)
,
i
p
p
p
p
p
p
p
p
p
p
p
p
p
j
j
k
=
u
i
+
u
−
u u
i
(11.25)
p
p
Based on the small world and scale free theory the trust model in [25] only consid-
ers two parameters in STNs: clustering coefficient and centrality. Centrality deter-
mines the relative importance of a node in the STN. Many centrality measures can
be used to compute this parameter. he clustering coeicient of a node in STN
is used to represent the closeness of the node and neighbors as a clique. It can be
computed as follows:
{ }
−
e
k k
ik
=
∈
∈
C
)
,
v v
,
N e
,
E
(11.26)
i
i
k
i
ik
(
1
i
i
where:
v
i
is node
i
,
N
i
is the neighborhood of the node
i
,
e
ik
is the link between node
i
and node
k
, and
k
i
is the degree of the node
i
.
In [25] it also introduces a trust evaluation algorithm to establish and update
the trust relationship as follows: Assume that node
i
is a newcomer of a network.
We can use the algorithm shown below to compute node
i
's trust of node
j
in the
network
(
O
i
:
Get all opinions about node
j
.
k
)
=
{
}
k k
,
,...,
k
n
is a set of all the nodes that have
1
2
opinion about node
j
.
O
net
=
(
b
,
d
,
u
)
11.27)
net k
,
net k
,
net k
,
k
i
i
i
i
represents the position of node
k
i
in the whole network.
O
k
net
i
b
net k
i
,
=
CENTRALITY
(11.28)
d
= −
1
b
−
u
(11,29)
net k
,
k j
,
k j
,
i
i
i