Exploration temporelle de données archéologiques imprécises : graphe d'antériorité

Cyril de Runz (1) , Éric Desjardin (1)

1 CRESTIC - Centre de Recherche en Sciences et Technologies de l'Information et de la Communication - EA 3804

In this paper, we propose a new temporal data mining method considering a set of archaeological objects which are temporally represented with fuzzy numbers. 0ur method uses an index which quanti es the anteriority between two fuzzy numbers for the construction of a weighted oriented graph. The vertices of the graph correspond to the archaeological objects. Using this anteriority graph, we determine the potential of anteriority, of posteriority and the temporal position of each object. The information obtained designs the temporal relations between objects and allows us to extract the most anterior object and the most posterior object. We apply this approach to the roman streets discovered in Reims.

Domaines: Intelligence artificielle [cs.AI]

Cyril de Runz, Éric Desjardin. Exploration temporelle de données archéologiques imprécises : graphe d'antériorité. Atelier Fouille de données complexes, conférence Extraction et Gestion des Connaissances, 2009, Strasbourg, France. pp.23--34. ⟨hal-00584806⟩