Zheng-Yang Liu, Arthur M. Farley
Structural aggregation is an inherent aspect of our ability in reasoning about the real-world. In this paper, we present our investigation of structural aggregation to simplify domain models and suppress irrelevant details of complex physical systems. We address the role of clustering, cluster-orientation and creation of black boxes in qualitative causal reasoning. We describe algorithms that automate structural aggregation to achieve efficiency and clarity in planning and explanation of physical system behaviors.