IEEE Technology and Society Magazine - Summer 2013 - 36

computed altruism (e.g., expecting
the cooperative act will bring more
benefit to future interactions).
Outcome
When the trustee behaves as the
trustor expects, the trustee gains
more power such as competence and
more trust in the eyes of the trustor.
In addition, a positive reciprocating
relationship may be created, leading
to enhanced trust. Moreover, trust
can create social capital meaning
that social networking or relationships will generate productivity
in performance. All these positive
products can be fed back to ensure
or update the belief for the trustor to
form trust in the trustee. In contrast,
when the trustee does not behave
as expected, the trustor feels disappointment or regretfulness, which
may trigger retaliation to penalize
untrustworthiness or require forgiveness to renew the relationship.
Since the outcome of the trusting act
also affects the initial belief to form
trust in the trustee, the trust process
is naturally recursive and dynamic.
One key observation is that trust
creates trust while distrust generates
distrust based on the conjecture that
an entity responds as it is treated [4].
This work captures the applications of the positive circulation of
trust to build sustainable networks.
In terms of multidisciplinary perspectives (e.g., psychology, ecology,
or biology), altruistic or reciprocating behaviors are studied as the
motive to initiate trust relationships
in humans, animals, or insects [2],
[16]. Although the relationship itself
may be initiated involuntarily or
incidentally, reciprocation creates
the circulation of the positive relationship. When two entities are in a
certain situation for a long time, it is
observed that they may benefit more
when they cooperate than when they
do not [2]. In addition, altruistic
behavior can trigger the reciprocal
altruistic relationship, the so called
"reciprocal altruism." In this relationship, some critical factors affect
the reciprocally altruistic behavior
36

such as duration of relationship
(length of lifetime), mobility rate or
pattern (disperse rate), and benefit of
mutual dependence [4]. The literature summates that there may exist
an optimal group size of animals or
insects to build a reciprocally altruistic system [4]. Friendship or kinship may be the main determinants
to increase altruistic behaviors. In
addition, when an entity reciprocates
the help it received in the past, it
can reciprocate offspring of the past
reciprocator even if the reciprocator
does not exist [16].
Based on the insights from various disciplines as above, optimal
network structures may exist to
promote reciprocating behavior to
foster trust relationships among entities and build a system level atmosphere of trust. These relationships
can affect decisions for altruistic or
reciprocal behaviors that promote
trust. At the same time, reciprocation can be inherited by another
entity that can be regarded as a child
(or other close relationship) of the
initial reciprocating entity when the
initial reciprocator does not exist.
This further facilitates potential
of the longevity of reciprocating
behaviors and trust relationships. In
addition, since an entity's resource is
limited to maintain its relationships,
the optimal group size or number of
friends (neighbors) may exist for an
entity to maximize its trust relationships. Mobility and durability of the
relationship are critical factors that
affect the decision making of each
entity of reciprocating or altruistic
acts. In the literature, as an application of reciprocation among nodes
in a network, social ties are used
as criteria to provide services to
particular nodes in delay tolerant
networks [8].
In our daily lives, we mostly
make decisions based on what we
gain from a relationship, commitment, cooperation, or collaboration
as payoff/utility. The decision making process about whether or not to
trust or cooperate can be discussed
from a game theoretic perspective.

Now we propose a trust-based decision making model assuming that a
rational entity may want to maximize its payoff (i.e., a net gain based
on the utility function of an entity)
in a decision on whether to trust or
not and whether to cooperate or not.

Emergence, Propagation,
and Sustainability of Trust
From the economic perspective,
decision making is a game. Each
entity in the system has the goal
to win and produce net gains by
making wise decisions as a selfinterested agent. rewards and
penalties are natural ways to
enforce or reinforce entities' behaviors to maximize payoffs of the system as well as those of individual
entities. We look into the benefit of
intelligent decision making on system sustainability in the long term
while also maintaining acceptable
performance in the short term.
This suggests the possibility of a
tradeoff, between short-term payoff (net gain) and long-term payoff.
The underlying idea of this work
is based on dyadic relationships
and the principle of payoffs upon
decisions of each entity. Modifying
the payoff theory from [10], [13] to
model the relationship between a
trustor and a trustee in their decision
making, we show Fig. 3 as the conceptual framework that promotes
trust emergence, propagation, and
sustainability based on the payoff
computation of each entity where
two players participate in a repeated
game. Note that in networks, payoffs
can be interpreted as any benefit an
entity may receive by the decision it
makes such as accessibility to confidential information, resources, or
energy savings. In each session of
the game, trustor A can make its
decision on whether to trust or not,
and trustee B can make its decision
on whether to cooperate or not.
When A has two options to choose,
whether to trust or not, there exists
the payoff of each decision. When A
predicts its payoff as more beneficial
by not trusting than trusting (i.e.,

IEEE TECHNOLOGY AND SOCIETY MAGAZINE

SuMMEr 2013

Table of Contents for the Digital Edition of IEEE Technology and Society Magazine - Summer 2013