Skip to Main content Skip to Navigation

Infrastructure adaptable pour les entrepots de données

Abstract : A data warehouse is a collection of historical data exploited by decision-support applications. It centralizes data of interest for a group of users so data acccess is fast, inexpensive and effective. In this work, we are interested in the evolution of data warehouses supporting multidimennsional data analysis applications. We propose an infrastructure to build warehouse managers that enable the creation and the evolution of data warehouses. A warehouse manager enables the creation of the schema of a data warcehouse and provides the means to modify this schema. It also offerd tools to build the warehouse from data coming from multiple heterogeneous and autonomous sources. It is basedd on a multidimensional data model that allows to manipulate the warehouse schema at an abstract level and to integrate heterogeneous data. We have specified a data model that is used for delining multidimensional schema. We propose a set of schema evolution primitives which guarantee that such a schema remains consistent between successive evolutions. We propose MDL, a language associated with our model, which offers expressions to create and evolve multidimensional schemata. To validate our approach, we implemented a prototype warehouse manager. It ensure the creation and the evolution of a relational data warehouse. This manager, developed in Java, integrates data in XML format coming from heterogeneous sources.
Document type :
Complete list of metadata
Contributor : Thèses Imag <>
Submitted on : Thursday, September 29, 2005 - 5:58:59 PM
Last modification on : Friday, November 6, 2020 - 3:52:42 AM
Long-term archiving on: : Friday, April 2, 2010 - 10:48:36 PM


  • HAL Id : tel-00010335, version 1




Edgard-Iván Benitez Guerrero. Infrastructure adaptable pour les entrepots de données. Interface homme-machine [cs.HC]. Université Joseph-Fourier - Grenoble I, 2002. Français. ⟨tel-00010335⟩



Record views


Files downloads