Multi-Agents Systems for Cartographic Generalization: Feedback from Past and On-going Research

Abstract

Cartographic generalization is a highly local and contextual process where decisions are taken locally to better adjust the transformations used to the local geography. Thus, cartographic generalization fits well with the multi-agents paradigm that promotes decentralized and autonomous decision-making. The past years of research in cartographic generalization showed several successful attempts to use multi-agents systems, and this paper provides a feedback on these attempts. We extracted a core modeling of a multi-agents system for generalization and highlighted its main components. Previous propositions of multi-agents generalization processes are described in relation to this core modeling, and feedbacks from experimentations with these processes are discussed to define a research agenda in multi-agents modeling for generalization.