Behaviour Modelling with Causally Related Actions

This chapter introduces the modelling of a behaviour by relating its (inter)actions by means of causality relations. A causality relation consists of a causality condition, attribute constraints and a probability constraint. The causality condition defines how the occurrence of a result (inter)action depends on the occurrence and non-occurrence of other (inter)actions. We first introduce the initial, independent, enabling, disabling and synchronous (inter)actions as basic forms of these dependencies, leading to a complete set of so called basic causality conditions. Next we discuss the forming of more complex causality conditions by composing them from basic causality conditions with conjunction and disjunction operators. We then augment a causality condition with attribute constraints that define how the attribute values of a result (inter)action depend on attribute values established in other (inter)actions that the result (inter)action depends on. We finally augment the causality condition with a probability constraint. While discussing these concepts we also introduce specification language elements to represent them. These concepts together allow the complete modelling of a causality relation between a result (inter)action and the (inter)actions it depends on. The causality relations of all result (inter)actions in a behaviour define that behaviour completely, however, in an unstructured, i.e., monolithic, way.