Multiple Inheritance in AgentSpeak(L)-Style Programming Languages

Agent-Oriented Programming (AOP) is a high-level pro-gramming paradigm for implementing intelligent distrib-uted systems. While a number of AOP languages have been proposed in the literature, many of them focus on the provision of support for intelligent decision making rather than addressing language design concerns such modularity and reuse. To address this imbalance, this paper presents an abstract model of multiple inheritance for AgentSpeak(L) style languages which decomposes agent programs into a set of inter-related agent classes and defines a run-time apparatus for rule selection on the relationship between those classes. To demonstrate our approach, a case study is presented that introduces a new AgentSpeak(L)-based language entitled ASTRA and its use in an illustrative implementation of an agent-based chat system is then presented.