Identity-based change: A foundation for spatio-temporal knowledge representation

Abstract

As efforts grow to develop spatio-temporal database systems and temporal geographical information systems that are capable of conveying how geographical phenomena change, it is important to distinguish the elements that are fundamental to scenarios of change. This paper presents a model based on the explicit description of change with respect to states of existence and non-existence for identifiable objects. Such changes are of concern when, for instance, modelling and reasoning about nations that are subsumed through conflict only to return once more at a later time, or about water bodies that fluctuate due to seasonal or climatic change. The basis for tracing these changes is the concept of object identity. Identity, distinct from an object's properties, values, or structure, is that unique characteristic that distinguishes one object from another. Based on a small set of primitives relating to the identity states of objects, we model the semantics associated with change and through a systematic derivation, a complete set of identity-based change operations evolves from the primitives. These operations are basic to the types of change commonly experienced by geographical phenomena and modelled by researchers studying spatio-temporal change. This approach highlights the minimum elements necessary for reasoning about change, namely, object identity, an ordering of identity states, and co-occurrence of identity states. © 2000 Taylor & Francis Group, LLC.