Abstract
M.Sc. (Computer Science)
Novel trust and reputation models are frequently proposed by the research community to suit
the needs of a specific environment. From the plethora of models that are available, it
becomes difficult to know which features can be combined in general-purpose models suitable
for commercial use. In order to address this problem, the focus of recent research on trust and
reputation systems has been on the identification of common features in order to enable reuse.
Organizations who need to use a reputation system within their application domain have to
custom build it, which may be challenging for novice reputation system developers.
This dissertation defines a strategy to develop a configurable SaaS reputation service that has
the ability to support common features, but at the same time accommodate the unique
requirements of a variety of online communities. A domain analysis reveals common features
that can be arranged and re-organized using variability modelling to enable a SaaS providers
to support the configuration of a SaaS reputation service.
The research conducted in this dissertation proposes a Reputation-as-a-Service model that
support configuration in order to accommodate the reputation requirements of a variety of
communities.