Utilisation d'outils de Visual Data Mining pour l'exploration d'un ensemble de règles d'association

Gwenael Bothorel (1, 2) , Mathieu Serrurier (3, 4) , Christophe Hurter (1, 5)

1 DGAC - Direction Générale de l'Aviation Civile
2 IRIT-APO - Algorithmes Parallèles et Optimisation
3 IRIT-ADRIA - Argumentation, Décision, Raisonnement, Incertitude et Apprentissage
4 ADRIA Développement
5 IRIT - Institut de recherche en informatique de Toulouse

Data Mining aims at extracting maximum of knowledge from huge databases. It is realized by an automatic process or by data visual exploration with interactive tools. Automatic data mining extracts all the patterns which match a set of metrics. The limit of such algorithms is the amount of extracted data which can be larger than the initial data volume. In this article, we focus on association rules extraction with Apriori algorithm. After the description of a characterization model of a set of association rules, we propose to explore the results of a Data Mining algorithm with an interactive visual tool. There are two advantages. First it will visualize the results of the algorithms from different points of view (metrics, rules attributes). Then it allows us to select easily inside large set of rules the most relevant ones.

Mots clés: aeronautical data; fouille de données; règles d'association; sémiologie graphique; données aéronautiques; association rules mining; visual semiology

Domaines: Interface homme-machine [cs.HC]

Gwenael Bothorel, Mathieu Serrurier, Christophe Hurter. Utilisation d'outils de Visual Data Mining pour l'exploration d'un ensemble de règles d'association. IHM 2011, 23ème Conférence Francophone sur l'Interaction Homme-Machine, Oct 2011, Nice, France. pp.Article N° 12, ⟨10.1145/2044354.2044369⟩. ⟨hal-01022276⟩