Uncertainty in Place-Based GIS:A Review and a Research Framework
TIAN Yuan,GAO Yong,LIU Yu
2008-01-01
Abstract: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.