Belief Update in AgentSpeak-DL

In previous work (Moreira et al, DALT 2005) we proposed an extension for the belief base of AgentSpeak agents based on Description Logic (DL), aiming at enabling agent oriented programming to cope with recently proposed technologies for the Semantic Web. In such an extension an agent belief base contains the definition of complex concepts, besides specific factual knowledge. The foreseen advantages are: (i) more expressive queries to the belief base; (ii) a refined notion of belief update, which considers consistency of a belief addition; (iii) flexibility in plan searching allowed by subsumption relation between concepts; and (iv) knowledge sharing in a semantic web context (based on OWL). Following this proposal an extension of the well know Agent Speak interpreter, Jason, was presented by K lapiscak and Bordini in DALT 2008. Among the interesting open issues is how to deal with the addition of beliefs which violates ontology consistency. In this work discuss this problem related to ABox updating in the context of AgentSpeak-DL.