An argumentation framework based on conditional priorities

Abstract

We propose a framework to allow an agent to cope with inconsistent beliefs and to handle conflicting inferences. Our approach is based on a well-established line of research on assumption-based argumentation frameworks and defeasible reasoning. We propose a language to allow defeasible assumptions and context-sensitive priorities to be explicitly expressed and reasoned about by the agent. Our work reveals some interesting problems to conditional priority-based argumentation and establishes the fundamental properties of these frameworks. We also establish a sufficient condition for a conditional priority-based argumentation to have a unique stable extension based on the notion of stratification. © 2008 Springer Berlin Heidelberg.