Information Technology Reference
In-Depth Information
Figure 5. The final decision maker module rules
and 0 for no trust. For example a vendor's Website with 0.75 trust factor is considered high and should
be trusted.
The inference rules are subject to the user's choice based on criteria of the risk and the gain as
defined by Tao and Thoen (2001). Fuzzy inference is a process to assess the trust index in five steps:
(1) register the initial values of each variable as defined by the information extraction system, (2) use
the membership functions to generate membership degrees for each variable related to each module,
(3) apply the fuzzy rule set defined for each module onto the output space (trust index) through fuzzy
'and' and 'or' operations, (4) aggregate the outputs from each rules related to each module, and (5)
derive the trust index through a defuzzification process using the centroid method. These same steps
will also be used for the decision maker module to generate the trust factor. From Figure 4, we can see
that the trust index increases with the increase of the contributing attribute of all trust indices values
and decrease when the decrease of all the attribute. Figure 5 shows a sample of the IF-THEN rules for
the final decision maker module.
the construction
of the rules Base
The decision to trust or not to trust EC as a shopping medium is up to consumers' evaluation, which can
be based on many factors such as price, convenience, selection of choice, and the information available
on the merchant's Website like those defined in our model. It is widely accepted that if the economic
gain is greater than the risk involved then the transaction is reasonably viable. Based on this assump-
tion, Tao and Thoen (2001) formalized the process as:
G
b
=
P
b
L
b
where
G
b
is the gain entering the EC
transaction,
P
b
is the risk that the consumer takes for trusting the EC merchants and
L
b
is the loss the
Search WWH ::
Custom Search