Information Technology Reference
In-Depth Information
trusted. In the context of interactions between a service provider (trustee) and customers
(trustors), some companies (e.g., e-bay and amazon) provide means for their customers
to provide their feedback on the quality of the services they receive, under the form of a
rating chosen out of a finite set of discrete values. This leads to a multi-valued domain
of trust, where each trust rating represents the level of trustworthiness of the trustor
agent as viewed by the trustee agent. While most of the web applications ask users to
provide their feedbacks within such a multi-valued rating domain, most studies [1], [2],
[3] and [4] are restricted to binary domains. Hence, our motivation for developing a
multi-valued trust domain where each agent can be evaluated within a multi-valued set
of ratings.
An agent may ask its advisors to provide information on a target agent who is un-
known to him. The advisors are not necessarily truthful (e.g., competition among market
shares, medical records when buying a life insurance) and therefore may manipulate
their information before reporting it. In addition, the advisor agents' trustworthy be-
haviour may differ from one interaction to another, leading to some uncertainty about
the advisors' trustworthiness and the accuracy of information revealed by them.
Possibility distribution is a flexible tool for modelling an agent's trust considering
uncertainties which arises form either lack of sufficient information about an agent's
trust or the variability in the agent's degree of trustworthiness. Possibility theory was
first introduced by [5] and further developed by Dubois and Prade [6]. It was utilized,
e.g., to model reliability [7]. We use possibility distributions to represent the trust of
an agent with respect to its uncertainties. Further, we propose an approach on merging
the possibility distributions of an explorer agent's trust in it's advisors with the reported
possibility distributions by the advisors on a target agent's trust. The resulted possibility
distribution represents an estimation of the explorer's trust in the target agent. The rest
of the paper is structured as follows: Section 2 describes the related works. Section 3
provides a detailed description of the problem domain. Section 4 reviews some fusion
rules for merging possibility distributions. In Section 6, we propose our approach on
merging two different sets of the possibility distributions in order to estimate the target
agent's trust. Finally, in Section 7, extensive experimental evaluations are presented to
validate the proposed merging approach.
2
Related Work
Considerable research has been accomplished in multi-agent systems providing models
of trust and reputation, a detailed overview of which is provided in [8]. In reputation
models, an aggregation of opinions of members towards an individual member which
is usually shared among those members is maintained. Starting with [9], the reputation
of an agent can be evaluated and updated by agents over time. However, it is implicitly
assumed that the agent's trust is a fixed unknown value at each specific time which does
not capture the uncertainties in an agent's trust. Regret [10] is another reputation model
which describes different dimensions of reputation (e.g. “individual dimension”, “social
dimension”). However, in this model the manipulation of information and how it can be
handled is not addressed. Some trust models try to capture different dimensions of trust.
In [11] a multi-dimensional trust containing elements like success, cost, timelines and
