TOWARDS AN INTEGRATION OF TIME AND CAUSATION IN A HYBRID KNOWLEDGE REPRESENTATION FORMALISM

The article focuses on the relations between time and causation and proposes a general framework in which a strict integration is achieved both in the representation formalism and in the reasoning process. By taking advantage of the basic capabilities of a Hybrid Knowledge Representation formalism (in particular, we use BACK), we provide an explicit and integrated representation of the basic entities in the temporal and causal ontology, and propose a taxonomy in which different types of causal relations found in the literature are classified, depending on the temporal relations they impose between causes and effects. Moreover, we develop a specialized causal-temporal reasoner which operates on causal nets built upon the basic causal-temporal representation. Such a reasoner is flexible (in the sense that different definitions of causation may be considered in the same reasoning process) and integrated (since the mutual effects of time and causation are coped with), and takes advantage of a specialized temporal reasoner.