POI3D : généralisation d'une méthode de visualisation 2D

David da Costa (1) , Gilles Venturini (1)

1 LIFAT - Laboratoire d'Informatique Fondamentale et Appliquée de Tours

We present in this paper an important extension of a visualization method based on points of interest in the context of visual data mining. We introduce a new method to select the POIs and most importantly a generalization to 3D in order to remove the ambiguities present in the 2D version. The set of points of interest is selected with a local optimization method with the aim of maximizing the quality of the visualization. We show that the generalization to 3D solves several problems of the 2D method and that it allows us to deal with large data sets.

Domaines: Interface homme-machine [cs.HC]; Vision par ordinateur et reconnaissance de formes [cs.CV]

David da Costa, Gilles Venturini. POI3D : généralisation d'une méthode de visualisation 2D. 3èmes Rencontres Inter-Associations, Recherche et extraction d'information, Mar 2007, Toulouse, France. ⟨hal-01037872⟩