A Survey of Data Warehouse Model Evolution

A Survey of Data Warehouse Model Evolution

Cécile Favre (University of Lyon (ERIC Lyon 2), France), Fadila Bentayeb (University of Lyon (ERIC Lyon 2), France) and Omar Boussaid (University of Lyon (ERIC Lyon 2), France)
DOI: 10.4018/978-1-60566-242-8.ch015
OnDemand PDF Download:
$37.50

Abstract

A data warehouse allows the integration of heterogeneous data sources for analysis purposes. One of the key points for the success of the data warehousing process is the design of the model according to the available data sources and the analysis needs (Nabli, Soussi, Feki, Ben-Abdallah & Gargouri, 2005). However, as the business environment evolves, several changes in the content and structure of the underlying data sources may occur. In addition to these changes, analysis needs may also evolve, requiring an adaptation to the existing data warehouse’s model. In this chapter, we provide an overall view of the state of the art in data warehouse model evolution. We present a set of comparison criteria and compare the various works. Moreover, we discuss the future trends in data warehouse model evolution.
Chapter Preview
Top

Introduction

A data warehouse allows the integration of heterogeneous data sources for analysis purposes. One of the key points for the success of the data warehousing process is the design of the model according to the available data sources and the analysis needs (Nabli, Soussi, Feki, Ben-Abdallah & Gargouri, 2005).

However, as the business environment evolves, several changes in the content and structure of the underlying data sources may occur. In addition to these changes, analysis needs may also evolve, requiring an adaptation to the existing data warehouse’s model.

In this chapter, we provide an overall view of the state of the art in data warehouse model evolution. We present a set of comparison criteria and compare the various works. Moreover, we discuss the future trends in data warehouse model evolution.

Key Terms in this Chapter

Data Warehouse Schema: Designs the structuration of the data in the data warehouse; measures representing facts are analyzed according to dimensions.

Data Warehouse: Collection of historical data, built by gathering and integrating data from several data sources, structured in a multidimensional way to support decisional queries.

Model Updating: Making the same model evolve without keeping track of its evolution history; thus, the model corresponds to its current version.

Temporal Validity Label: A temporal validity label corresponds to a timestamp used to denote the valid time.

Analysis in a Consistent Time: Results are provided by taking into account the moment when a fact exists in the reality.

Temporal Modeling: Providing temporal extensions to keep track of the model history.

Model Versioning: Building several versions of a model where each new version corresponds to a schema evolution or an evolution of data valid for a given period.

Data Warehouse Model: The data warehouse model includes its schema and its data.

Complete Chapter List

Search this Book:
Reset