Skip to Main content Skip to Navigation
Theses

Évolution de schémas dans les entrepôts de données : mise à jour de hiérarchies de dimension pour la personnalisation des analyses

Abstract : This thesis has been done in collaboration with the bank LCL-Le Credit Lyonnais. It deals with the evolution of analysis needs in data warehouses. Indeed the latter beeing a fundamental component in decision support systems it is important to place users at the center of the evolution process to better reflect individual analysis needs. Our proposal consists in guiding data warehouse schema evolution through the use of users' knowledge. These knowledge are used to create new analysis axis. Our solution is based on the definition of a formal model describing an evolving data warehouse. The model is composed of "if-then" rules that we call aggregation rules that support the personnalisation process. This process is also supported by an architecture composed of four steps: the knowledge acquisition gathers users' knowledge in the form of aggregation rules; integration of these aggregation rules in the data warehouse; the data warehouse schema evolution, and on-line analysis based on the new schema. The implementation of our proposal is done through a relational execution model that manages the overall evolution process. We also interested in the performance of our evolving data warehouse model. The performance evaluation is usually based on a workload (a set of queries), and since we are in an evolving context, we proposed an incremental update of the workload. To validate our contributions, we have implemented the WEDriK plateform (data Warehouse Evolution Driven by Knowledge).
Document type :
Theses
Complete list of metadatas

https://tel.archives-ouvertes.fr/tel-00269037
Contributor : Cécile Favre <>
Submitted on : Tuesday, April 1, 2008 - 10:48:36 PM
Last modification on : Tuesday, November 19, 2019 - 2:06:29 AM
Long-term archiving on: : Friday, May 21, 2010 - 1:07:19 AM

Identifiers

  • HAL Id : tel-00269037, version 1

Collections

Citation

Cécile Favre. Évolution de schémas dans les entrepôts de données : mise à jour de hiérarchies de dimension pour la personnalisation des analyses. Interface homme-machine [cs.HC]. Université Lumière - Lyon II, 2007. Français. ⟨tel-00269037⟩

Share

Metrics

Record views

1045

Files downloads

8551