Modeling Moving Geospatial Objects from an Event‐based Perspective

Abstract

Abstract In this paper, the dynamic happenings or occurrents experienced by entities in a geospatial domain, are modeled as events and from an analysis of these events, we show how meaningful information about the movement of objects can be abstracted. The semantics of moving objects are captured based on different events, for example, specialized events, such as departure , arrival, or unexpected destination events. The paper presents an approach for abstracting patterns of movement from event sequences based on key attributes: object identity, event description, and event location. A set of eight possible event pattern types are distinguished and discussed in the paper. From these patterns, three notable patterns of events are presented, i.e. identifiable trends of movement that are worthy of further attention or perhaps intervention. Two of these patterns, repeating and collocating , capture particular kinds of movement within a traversal or transit of a domain by a single moving object, such as a zigzag movement or reversal of direction, while a reiterating pattern of events corresponds to a movement pattern associated with multiple objects and describes, for example, convergence in a region. The inclusion of events in the data model provides a foundation for distinguishing particular semantics of movement based on patterns of events, providing a basis for querying different kinds of events as well as providing a foundation for developing automated event notification systems.