Extension de l'algorithme Apriori et des règles d'association aux cas des données symboliques diagrammes et intervalles

This paper deals with the extension of the Apriori algorithm and of the association rules to the symbolic histogram and interval-valued data. We suggest a method that will enable us to discover rules at the level of the concepts. This extension requires new definitions for the support and the confidence in order to take advantage of the symbolic structure of the data. The market basket data example is developed throughout the paper. Thus, instead of mining rules between different items of some transactions recorded in a retail organization like in the classical case, we discover rules at the level of the customers in order to study their purchase behavior.