Information Technology Reference
In-Depth Information
We now take the point of view of the trustee in the dependence network : so we present a
cognitive theory of trust as a capital. That is to say that if somebody is (potentially) strongly
useful to other agents, but is not trusted, its negotiation power is not high.
As showed in Chapter 3 we call Degree of Trust of the agent Ag j for the agent Ag i about the
task
τ k ( DoT(Ag j Ag i τ k ) ):
DoA Bj
ik
DoW Bj
ik
Ag i k ) Bj
DoT ( Ag j ,
=
(10.7)
In the same way we can also define the self-trust of the agent Ag i about the task
τ k :
DoA Bi
ik
DoW Bi
ik
ST ( Ag i k )
=
(10.8)
We call the Objective Trust Capital of Ag i
Agt about a potential delegable task
τ k the function:
l
l
DoA Bj
ik
DoW Bj
ik
Ag i k ) Bj
OTC ( Ag i k )
=
=
DoT ( Ag j ,
(10.9)
j = 1
j = 1
Where l is the number of agents (included in the dependence network) who need to delegate
the task
τ k . Note that we are calling as objective trust capital the sum of the trustworthiness
that the other agents in the DN attribute to Ag i rather than the capital Ag i could deserve on the
basis of his own objective relationships: in other words, it is referred to the partial (subjective)
points of view of the other agents.
In words, the cumulated trust capital of an agent Ag i with respect to a specific delegable
task
τ k , is the sum (all the agents need that specific task in the network dependence) of the
corresponding abilities and willingness believed by each potentially dependent agent. 4
We call the Subjective Trust Capital of Ag i
Agt for a potential delegable task
τ k the function:
l Bi
l Bi
DoA Bi Bj
ik
Bi Bj
Ag i k ) Bi Bj
STC ( Ag i k )
=
DoW ik
=
DoT ( Ag j ,
(10.10)
j = 1
j = 1
Where the apex B i B j means 'as Ag i believes is believed by Ag j '. Subjectivity means that both
the network dependence and the believed abilities and willingness are believed by (the point of
view of) the agent Ag i . The subjectivity consists in the fact that both the network dependence
and the believed abilities and willingness are believed by (the point of view of) the agent Ag i
(see Figure 10.8A and 10.8B).
Starting from Trust Capital we would like to evaluate its usable part . In this sense, we
introduce the Subjective Usable Trust Capital of Ag i
Agt for a potential delegable task
τ k as:
l Bi
Ag i k ) Bi Bj
DoT ( Ag j ,
SUTC ( Ag i k )
=
(10.11)
1
+
p kj
Bi
j = 1
Where the apex B i B j means 'what Ag i believes is believed by Ag j ' and where p kj Bi is (following
Ag i 's belief about the beliefs of Ag j ) the number of other agents in the dependence network
4 We might consider in an even more objective way the capital based on the real trustworthiness of Ag i (with respect
to the task) rather than based on the DoT(Ag j Ag i τ k ) of the various agents Ag j . But the trust capital is not interesting
if it is not in the mind of the potential 'users'.
 
Search WWH ::




Custom Search