Graph-Based Trust Model for Evaluating Trust Using Subjective Logic

Before using a digital system, it is necessary to evaluate it according to different parameters. Lately trust emerged as a momentous aspect of evaluation. Evaluating trust in a system is a complex issue that becomes more challenging when systems use distributed architectures. In a previous work, we proposed SOCIOTRUST, a trust model that is based on probability theory to evaluate trust in a system for an activity. In SOCIOTRUST, trust values are considered as the probability, by which a trustor believes that a trustee behaves as expected. A limitation of using traditional probability is that users cannot express their uncertainties about some actors of their activity. In real situations, not everyone is in possession of all the necessary information to provide a dogmatic opinion about something or someone. Subjective logic thus emerged to facilitate the expression of trust as a subjective opinion with degrees of uncertainty. In this paper, we propose SUBJECTIVETRUST, a graph-based trust model to evaluate trust in a system for an activity using subjective logic. The distinctive features of our proposal are (i) user’s uncertainties are taken into account in trust evaluation and (ii) besides taking into account the trust in the different entities the user depends on to perform an activity, it takes into consideration the architecture of the system to determine its trust level.