基于场所的GIS中的不确定性问题及分类框架

基于场所的GIS直接表达人类地理空间知识的管理和加工过程,而不确定性是人类智能的基本特点,因此GIS的智能化需要研究其中的不确定性问题。与传统的GIS相比,基于场所的GIS中的不确定性问题更为丰富,既包括随机性,也包括含糊性,而不确定性的主体既可以是地理要素、场所和空间关系,也包括命题和规则。该文介绍该领域相关的研究成果,基于不确定性主体、类型、表达手段及相关的活动4个视角,建立了基于场所的GIS中所涉及的不确定性框架,从而为相关的不确定性建模提供指导。

Uncertainty in Place-Based GIS:A Review and a Research Framework

A Place-Based GIS(PB-GIS) directly represents human geospatial knowledge.Since uncertainty is an inherent feature in both human intelligence and artificial intelligence,uncertainty in PB-GIS should be modeled.Compared with conventional coordinate-based GIS,a PB-GIS takes into account more types of uncertainty,which generally includes randomness and vagueness.This paper reviews and summarizes relevant research on modeling uncertainty in PB-GIS.Most research establishes qualitative or semi-quantitative models.A framework for managing different types of uncertainty is proposed.It represents uncertainty from the following four aspects:subjects,types,mathematical approaches to representation,and relevant activities.Firstly,the subjects of uncertainty include geographical features,places,spatial relationships,propositions,rules,and laws.Secondly,from the classification perspective,the uncertainty in PB-GIS may be vagueness,randomness,roughness,and ambiguity.Thirdly,each type of uncertainty can be modeled by corresponding representation approaches,such as fuzzy sets,rough sets,and probabilities.Lastly,the uncertainty in PB-GIS is always associated with different activities in mapping from the real world to information systems.These four aspects provide a guide to model uncertainty in place-based geographical information systems.