In this chapter, we summarize the discussion in the previous chapters and refocus
them in relation to the concept of trust context, especially in networks. As
a complex high-order process, the study of trust requires a multi-disciplinary
approach. Throughout this brief, we illustrated the role trust plays in many different
types of networks from different points of view. On one hand, a computational
system may trust the inputs provided by its users to provide services that form an
institutional context. On the other hand, a person may trust social and computational
institutions and other trustees to perform actions and to provide inputs that impact
the functioning of different institutions. The use of the term “trust” in all these
applications have some common elements.
To review our initial discussion of the term “trust”, we note that trust evaluation
involves a dependence relation: the trustor is dependent on a trustee (or trustees) to
accomplish a certain task. The trustor is uncertain whether her trust is well-placed
before making a decision, and her decision to trust makes her vulnerable. Under
this general umbrella, we find many variations of how trust is modeled. Ultimately,
it is these differences that define the trust context, the specific conditions under
which trust evaluations differ. In this chapter, we take a broader view of trust context
and describe some of the modeling challenges that networked applications need to
We outlined many different aspects of the trust context in the previous chapters.
The first distinction we made was with respect to the trustor. If the trustor is a person,
then we worry about human cognition and possible social contexts that she belongs
to. For computational agents, the model of trust may depend on the underlying
application. In many networked applications, this distinction is not so clear. People
use computational tools to mediate their interactions, and computational agents rely
on human input for their function. As a result, the more interesting problem to
study is how these two come together as a network. What are the properties of this
network? What types of dynamics can result in responses to changes in the network
state? We discuss this in Sect. 5.1 .
We also drew a distinction between trusting information and trusting actions.
When trusting information, information has already been given and the trustor can