Geocognostics — A new framework for spatial information theory

Abstract

The combination of cognitive and geometric approaches to space provides an integrated framework for a broad range of topics in spatial information theory. This framework, called geocognostics, organises space-time events into two basic representational structures, called views and trajectories. These structures are defined first in cognitive terms, but may be re-interpreted geometrically (in space-time) and hence serve as a common framework for both formal geometric representations of spatial information and formal representations of cognitive processes. Within this framework, views are collections of perceptual experiences recorded by an agent (and may be re-interpreted as collections of discrete space-time events), while trajectories are sequences of perceptual experiences of similar type (and may be re-interpreted as the space-time locus of persistent point entities). By combining these two low level representations into a series of hybrid representational structures, it is possible to deduce a great variety of conditions concerning perceptual experiences and, by extension, spatial data. To date, the new framework appears to have bearing on the theory of spatial and spatio-temporal data structures, the theory of interpolation, error, and spatial statistics, the processes of planning and way-finding, the design of computer interfaces for geometrical/geographical data and the larger question of human-computer interaction, the development of cognitively-based applications such as natural language interfaces, and many other sectors. In this paper, we describe the broad outlines of this framework and elaborate on some of the consequences of adopting the framework.