A Model of Trust Derivation from Evidence for Use in Recommendation Systems

In this paper we present a technique for modeling trust relationships that can be used in environments where agents take part in a recommendation system. In such an environment, agents collaborate with each other with the common aim of providing accurate recommendations to each other. In the model we use techniques taken from collaborative filtering to express trust properties as beliefs. We also compare our model with existing techniques that can be used to map types of behaviors to trust values. Finally, we describe the requirements of a protocol that could be used for the deployment of such a model in a real distributed environment.