A Prolog-based Approach to Self-Evaluated, Declarative and Zero-Knowledge Verifiable Policies

Abstract

In most domains where declarative policies are employed, it is typically the executor of the policy who performs policy evaluation, and not the subjects of policies. However, this approach has evident drawbacks from the trust, transparency and privacy aspects, especially when the subjects are natural persons. Building on recent developments in noninteractive zero-knowledge proofs and the technologies and standards supporting Self-Sovereign Indentity solutions, in this paper, we propose Self-Evaluated Policies, which move policy evaluation to the subject and leave the executor in a (zero-knowledge) proof-checking role. We present an SSI-based system model, propose a proof-tree-checking computational model for zero-knowledge proofs over the evaluations of Prolog-based policies, and describe a Circombased prototype.