Abstract
This paper proposes a method for formalizing and analyzing a partial knowledge of cryptographic functions within the runs and systems framework of multi-agent systems. We consider a multi-agent system where the interpretation of the cryptographic functions may vary from run to run. In this new model, we formalize the anonymity and coercion-resistance of electronic voting protocols in the presence of active adversaries, present a method for verifying these properties, and apply it to a practical electronic voting protocol. We also discuss computational justification for the method.
