A multigranular object-oriented framework supporting spatio-temporal granularity conversions

Several application domains require handling spatio-temporal data. However, traditional Geographic Information Systems (GIS) and database models do not adequately support temporal aspects of spatial data. A crucial issue relates to the choice of the appropriate granularity. Unfortunately, while a formalisation of the concept of temporal granularity has been proposed and widely adopted, no consensus exists on the notion of spatial granularity. In this paper, we address these open problems, by proposing a formal definition of spatial granularity and by designing a spatio-temporal framework for the management of spatial and temporal information at different granularities. We present a spatio-temporal extension of the ODMG type system with specific types for defining multi-granular spatio-temporal properties. Granularity conversion functions are introduced to obtain attributes values at different spatial and temporal granularities.