SAFFIET : un système d'extraction de règles d'associations spatiales et fonctionnelles dans les séries de données géographiques
Abstract
We start from the hypothesis that spatial dynamics and geographical object usage evolution
may be in part caused by the different previous evolution and the spatial configuration in which
they are situated. Therefore, in order to analyze and understand changes in the geographical
object functions over time, and to build a prospective then predictif model, we introduce the
SAFFIET system. It exploits data mining (through the search of frequent itemsets and association
rules) for extracting evolution rules that are guiding spatial dynamics.