A Production Rule-Based Framework for Causal and Epistemic Reasoning
Action theories are an important field of knowledge representation for reasoning about change and causality in dynamic domains. In practical implementations agents often have incomplete knowledge about the environment and need to acquire information at runtime through sensing, the basic ontology of action theories needs to be extended with epistemic notions. This paper presents a production system that can perform online causal, temporal and epistemic reasoning based on the Event Calculus and on an epistemic extension of the latter. The framework implements the declarative semantics of the underlying logic theories in a forward-chaining rule-based system. This way, it combines the capacity of highly expressive formalisms to represent a multitude of commonsense phenomena with the efficiency of rule-based reasoning systems, which typically lack real semantics and high-level structures.
KeywordsOperational Semantic Frame Problem Ambient Assist Live Epistemic Reasoning Partial Observability
Unable to display preview. Download preview PDF.
- 1.Levesque, H., Reiter, R.: High-level Robotic Control: Beyond Planning. A Position Paper. In: AIII 1998 Spring Symposium: Integrating Robotics Research: Taking the Next Big Leap (1998)Google Scholar
- 2.Van Harmelen, F., Lifschitz, V., Porter, B.: Handbook of Knowledge Representation. Elsevier Science, San Diego (2007)Google Scholar
- 4.Thielscher, M.: Representing the Knowledge of a Robot. In: KR 2000: Principles of Knowledge Representation and Reasoning, pp. 109–120 (2000)Google Scholar
- 6.Patkos, T., Plexousakis, D.: Reasoning with Knowledge, Action and Time in Dynamic and Uncertain Domains. In: Proceedings of the 21st International Joint Conference on Artifical Intelligence, IJCAI 2009, pp. 885–890 (2009)Google Scholar
- 9.Mueller, E.: Commonsense Reasoning, 1st edn. Morgan Kaufmann (2006)Google Scholar
- 10.Patkos, T., Plexousakis, D.: Efficient epistemic reasoning in partially observable dynamic domains using hidden causal dependencies. In: NRAC 2011 - 9th International Workshop on Non-Monotonic Reasoning, Action and Change, pp. 55–62 (2011)Google Scholar
- 14.Chen, L., Nugent, C., Mulvenna, M., Finlay, D., Hong, X., Poland, M.: A Logical Framework for Behaviour Reasoning and Assistance in a Smart Home. International Journal of Assistive Robotics and Mechatronics 9 (2008)Google Scholar
- 15.Sadri, F.: Intention recognition with event calculus graphs. In: Web Intelligence and Intelligent Agent Technology, pp. 386–391 (2010)Google Scholar
- 16.Kunze, L., Dolha, M.E., Beetz, M.: Logic programming with simulation-based temporal projection for everyday robot object manipulation. In: IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2011), pp. 3172–3178 (2011)Google Scholar
- 17.Patkos, T., Plexousakis, D.: Epistemic reasoning for ambient intelligence. In: 3rd International Conference on Agents and Artificial Intelligence (ICAART), pp. 242–248 (2011)Google Scholar
- 19.Shanahan, M.: The ramification problem in the event calculus. In: 16th International Joint Conference on Artificial Intelligence (IJCAI), pp. 140–146 (1999)Google Scholar