Model-based cartographic generalisation with uncertainty

The aim of this paper is to outline a proposed project to holistically generalise spatial data using agents. Cartographic generalisation is a process that is fraught with uncertainty – for a particular spatial scale there are an infinite amount of combinations for the display (or non-display) of data in the map space. Each map element (e.g. objects such as roads or buildings can be map elements) is an agent, with the ability to self-diagnose for cartographic conflict and reason with uncertainty (using Dempster-Shafer theory) to choose how to display itself in conjunction with neighbouring objects. Synoptically, a legible map will have been created through the intelligent interaction of agents at the local scale. This paper will explore issues associated with the above process.