A Taxonomy of Computational Models for Trust Computing in Decision-Making Procedures

Computational trust models are soft-security mechanisms that foster positive behaviours in a community, where users are inclined to be trustworthy to increase their chances of having a higher number of interactions. Even though computational trust is becoming more and more important, the literature on computational trust models is still a jungle of different formalisations, in which it is evident the lack of cohesion both from a linguistic and from a theoretical point-of-view. The paper will focus on those two issues, with the aim of untangling the jungle of computational trust. A meta-analysis of existing surveys on computational trust models will be provided, obtaining a novel taxonomy for those models; the taxonomy will be used to identify some classical examples of computational trust models, through which singling out the core features of a computational notion of trust. Through the whole process, emphasis will be put on theoretical aspects, abstracting away from subtleties of specific applications. The aim of the paper is two-fold: to provide a guide for newcomers to the subject, who will benefit from a linguistic clarification of the technical language of the subject, and to provide a solid starting point for all who wish to pursue research on computation trust.