Confidence maps: a tool to evaluate data's relevance in spatial analysis.

Krištof Oštir (1, 2) , Ziga Kokalj (1) , Laure Nuninger (3, 4)

1 IAPS - Institute of Anthropological and Spatial Studies
2 LEA ModelTER - Laboratoire européen de modélisation des paysages et des territoires dans la longue durée
3 MSHE - Maison des Sciences de l'Homme et de l'Environnement Claude Nicolas Ledoux (UAR 3124)
4 LCE - Laboratoire Chrono-environnement (UMR 6249)

Inventory data used in archaeology are often incomplete and heterogeneous. In the framework of the ArchaeDyn program, a method has been proposed to evaluate heterogeneity in archaeological inventories. The purpose of this work is to create a validation tool to interpret the results. This tool is called a "confidence map" and is produced by combining representation and reliability maps. The first step consists in generating representation maps to describe the clustering of archaeological items. The second step is based on reliability maps. Data providers are asked to define and outline the level of reliability of their data. Finally, representation and reliability layers are combined using map algebra. The resulting maps allow for the comparison and analysis of data confidence

Mots clés: confidence maps; spatial analysis; modelling territorial intelligence; intelligence territoriale

Domaines: Géographie

Krištof Oštir, Ziga Kokalj, Laure Nuninger. Confidence maps: a tool to evaluate data's relevance in spatial analysis.. 6th International Conference of Territorial Intelligence "Tools and methods of Territorial Intelligence", Oct 2008, Besançon, France. ⟨halshs-00982906⟩