A Compact Argumentation System for Agent System Specification

We present a non-monotonic logic tailored for specifying compact autonomous agent systems. The language is a consistent instantiation of a logic based argumentation system extended with Brooks’ subsumption concept and varying degree of belief. Particulary, we present a practical implementation of the language by developing a meta-encoding method that translates logical specifications into compact general logic programs. The language allows n-ary predicate literals with the usual first-order term definitions. We show that the space complexity of the resulting general logic program is linear to the size of the original theory.