Preserving Privacy in a (Timed) Concurrent Language for Argumentation

Modelling the behaviour of multiple agents concurrently interacting and reasoning in a dynamic environment is a challenging task. It necessitates tools capable of effectively capturing various forms of interaction, such as persuasion and deliberation while aiding agents in decision-making or consensus-building. In [1], we extended a language for modelling concurrent interactions between agents (tcla), which allowed us to specify agents equipped with a local argument memory and to reason with private knowledge. Furthermore, an initial formalisation of a multi-agent decision problem that preserves privacy has been provided. To illustrate the language’s capabilities, in this paper, we give a complete formalisation of a privacy-preserving multi-agent decision problem, and we demonstrate how it can be employed to define a general (correct and complete) translation function that generates a tcla program from a multi-agent decision-making process. Additionally, we present an application example that models a privacy-preserving multi-agent decision-making process to showcase an instance of our general translation function.